On Whether It Is and What It Is

Abstract

This dialogue, taking place between Prof. Whether and Prof. What, focuses on the nature of the relationship between ontology, conceived as the branch of philosophy concerned with the question of what entities exist, and metaphysics, conceived as the complementary part of philosophy that seeks to explain, of those entities, what they are. Most philosophers claim that it is not possible to address the first question without at the same time addressing the second, since knowing whether an entity exists requires knowing what that entity is. Prof. Whether argues against this popular position and offers a detailed analysis of the idea according to which it is possible to do ontology without engaging in metaphysics. Prof. Whether and Prof. What agree that, rather than being merely possible, in some cases it is, for methodological reasons, even preferable to start with a metaphysics-free ontology, postponing any inquiry concerning the nature of the entities included in the ontological inventory to a later stage. However, Prof. What notes that it is not always possible to do ontology without metaphysics, because there are certain kinds of entities, such as universals and possible worlds, that necessarily need a prior metaphysical characterization.

Prof. Whether: Do you think that aliens exist?

Prof. What: Well, I can’t answer if you don’t tell me what you mean by ‘aliens’.

Prof. Whether: By ‘aliens’ I just mean creatures inhabiting a planet that is not Earth.

Prof. What: But this is still not clear. Are you talking about living creatures or, more specifically, sentient creatures? Because I believe in the existence of the former, but not the latter. For example, I believe that there are some kinds of microorganisms somewhere in the universe, other than on Earth, but I don’t believe in the existence of small green humanoids with huge black eyes. Only if you tell me the nature of the thing you’re talking about will I be able to say whether it exists, or, for that matter, to profess my agnosticism.

Prof. Whether: Well, since some kinds of microorganisms located somewhere else in the universe are creatures from a planet that is not Earth, you didn’t actually need further clarification to answer my question. Anyway, you wanted me to tell you about ‘the nature of the thing’… Do you really think that it’s necessary to know this in order to say whether that thing exists?

Prof. What: Of course! Do you remember Meno’s paradox of inquiry? ‘And how are you going to search for this, Socrates, when you don’t have the faintest idea what it is? Which of the things that you don’t know will you suppose that it is, when you are searching for it? And even if you do come across it, how are you going to know that this is the thing you didn’t know?’ (Plato, 1985, 80d5–8). An ontological version of the paradox could be: how do you know if something exists if you don’t know what it is? Now, assume I don’t know the meaning of the word ‘table’…

Prof. Whether: Oh, please, use a less clichéd example.

Prof. What: OK, let’s go back to the aliens. If I don’t know the meaning of the word ‘alien’, or if I’m unsure about exactly what you mean by it, as I was before, and you ask me ‘Are there aliens?’, I just can’t answer your question, simply because I don’t know what you’re talking about. This seems obviously true to me.

Prof. Whether: I agree, it seems obviously true. But it isn’t true: in order to know whether there is something, you don’t need to know what it is.

Prof. What: Well… Do you remember Quine’s ontological question? It says… ‘What is there?’ (Quine, 1948, p. 21).

Prof. Whether: Oh, come on, that’s only a contingent feature of the English language.

Prof. What: Alright, alright, but now I want to know why you think we don’t need to know what something is in order to say whether it exists.

Prof. Whether: If you want to answer ontological questions (such as ‘Are there aliens?’), you don’t need to know what they are or even what these aliens look like. All you need to know is the attribution of certain properties that helps fix the reference to these things.

Prof. What: But when you answered my questions about what you meant by ‘aliens’, you did tell me something about aliens.

Prof. Whether: Yes, exactly: I told you something, but not more than that and, in particular, not their nature, namely what they are. It is enough if I tell you something about them that can help you understand what I mean by ‘aliens’. If you ask me a similar ontological question and you want to tell me more about the entities you have in mind, fine, but I don’t need it. One can even show that something exists without saying literally anything about it! For example, I could simply point at certain things, such as an armchair and a sofa, in my living room. I know that the existence of my furniture is not a particularly fascinating topic, but still, the point is that I’m showing you that certain things exist without saying anything about them.

Prof. What: I’m not worried that the existence of your armchair and sofa is philosophically uninteresting, but rather that in this case, unlike in the case of the aliens, we’re talking about instances of kinds of entities, rather than kinds of entities.

Prof. Whether: Yes, there’s this difference, but it doesn’t change anything: I’m still showing you that something exists without saying anything about it. Both statements of the form ‘There is an X’ and ‘There are Ys’ are ontological statements.

Prof. What: OK, but in the case of the furniture in your living room I know that this is an armchair and that thing over there is a sofa: I know to which kinds those entities belong. Doesn’t this say something about their nature?

Prof. Whether: I don’t think that just by knowing the kind to which an entity belongs implies that you know the nature of that entity. I may know that you’re a person, but… what is a person? Moreover, you can know that a certain thing exists even if you don’t even know to what kind that thing belongs. Suppose that, in my living room, there’s something we are totally unfamiliar with. We can resort to a bit of science fiction again here: let’s say that we find an entity with a humanoid shape in my living room, and that we’re not sure whether it is a living creature or a robot. Now that we have seen it, we both know that it exists, whatever it may be.

Prof. What: I see your point. But don’t you think that as soon as we have different opinions about that entity with a humanoid shape in your living room, we ontologically disagree? For example, you think it’s an extraterrestrial creature; I think it’s a robot. Therefore, you believe there is an extraterrestrial creature, whereas I believe there is a robot. These are two different ontological claims that depend on our different views of the nature of the object. Moving from a science fiction example to a philosophical one: let’s assume that we both agree that physical objects exist, but then it turns out that, say, I am a three-dimensionalist, whereas you are a four-dimensionalist. Thus, I believe that physical objects extend in three dimensions whereas you believe that they extend in four dimensions. It turns out that we don’t share the same ontology! I believe that we live in a world where there are spatial parts of physical objects, whereas you believe we live in a world where there are spatiotemporal parts of physical objects. Therefore, we have no ontological agreement.

Prof. Whether: Although our opinions radically differ, I still think that we ontologically agree on the fact that there is an entity with a humanoid shape in my living room, it just happens that we have different opinions about its nature. In general, competing views about the properties that an entity enjoys do not necessarily imply divergent ontological views. Think about two countries disputing the ownership of a strip of land at their border. Their disagreement about this disputed territory doesn’t imply a disagreement on the existence itself of the two countries (or of the strip of land itself!) Or, to give another example, we might disagree on whether the dust cover of a physical copy of a book is a proper part of that volume, but that doesn’t mean that we disagree on the existence itself of the physical copy of that book. If every different attribution of properties corresponded to a different entity, we would ontologically disagree about pretty much everything!

Prof. What: Hmm… And what do you think about the fact that sometimes we make negative ontological statements about certain entities by saying that we don’t know what they are? For example, if I don’t believe in the existence of linguistic entities such as meanings, I may express my disagreement by saying ‘I don’t know what meanings are supposed to be!’.

Prof. Whether: Yes, that happens sometimes, but that is not a way to say that you don’t know what the metaphysics of meanings is, but that you don’t know where to find these entities in your inventory of the world! Again, we need only an identifying description of the entities, that allows us to look for them, not a characterizing description that tells us what they are.¹

Prof. What: You might be starting to convince me…

Prof. Whether: And I’ll tell you more: not only do I think that it is possible to know that something exists without knowing what it is, but I also believe that this is always how things work. Let me finally use the terms ‘ontology’ and ‘metaphysics’. By ‘ontology’ here, I mean the inquiry into what there is, and by ‘metaphysics’ the inquiry into what it is. I know that, although it is a common way to distinguish the two domains of research, not everyone uses the terms in this way, but let’s stipulate here that this is what they mean: the issue we are discussing is philosophical, not terminological.

Prof. What: OK, we’re agreed on the meaning of ‘ontology’ and ‘metaphysics’.

Prof. Whether: Excellent. Let me now restate my thesis then: ontology always precedes metaphysics.

Prof. What: But now you just stated a new thesis.

Prof. Whether: You’re right, it’s a stronger one.

Prof. What: OK then, let me get this straight. One thesis states that it is possible to know that some entity exists without saying what the entity is.

Prof. Whether: Exactly.

Prof. What: Can we call this thesis the ‘Autonomy Thesis’? To the effect that ontology is independent of metaphysics.

Prof. Whether: Sounds good to me.

Prof. What: Perfect! You then stated the thesis that we have to first know whether something exists, and only after can we figure out what it is.

Prof. Whether: Yes, exactly.

Prof. What: I propose to call this thesis the ‘Priority Thesis’, to the effect that ontology is prior to metaphysics.

Prof. Whether: This sounds good too. Note that the Priority Thesis does not imply that metaphysics is a branch of ontology. Of course, the two domains of inquiry are strictly related, but I do not take that metaphysics is part of ontology (and vice versa): all I am stating is that ontology precedes metaphysics.²

Prof. What: So, the way I see it, the two theses are related in the following way: the Priority Thesis implies the Autonomy Thesis, but not the other way round.

Prof. Whether: This is correct.

Prof. What: Well, I agree now that the Autonomy Thesis is true, but I want to know more about the Priority Thesis before giving you my verdict…

Prof. Whether: Sure, but I think we should first agree on one meta-ontological issue, or else we cannot go any further.

Prof. What: Explain.

Prof. Whether: What we mean by ‘existence’ or ‘being’ has to be understood only in one sense.

Prof. What: So, by denying that existence has more than one meaning, are you endorsing what is known as ‘univocalism’ and denying what is known as ‘multivocalism?³

Prof. Whether: Yes, precisely: according to univocalism, there is only one sense of existence, whereas according to multivocalism, entities of different nature exist in different ways.

Prof. What: Why do you need to take a stance on this matter?

Prof. Whether: Because if different kinds of existence are respectively related to different metaphysical kinds of entities, then stating the existence of something would mean that it has been already metaphysically characterized.

Prof. What: We should talk about this more extensively on another occasion… but I’ll give you that this time, so let’s assume that univocalism is true.

Prof. Whether: Great. Oh, I almost forgot to tell you that I’m not the only one to have claimed that ontology has to be done before metaphysics.

Prof. What: Who else said it?

Prof. Whether: I am in good company: Aristotle writes in the Posterior Analytics that we first ask whether something exists or not (‘if a centaur or a god is or is not’) and, once we have ascertained its existence, we can ask what it is (‘what is a god? or what is a man?’).⁴

Prof. What: But in book Z of the Metaphysics, Aristotle writes that we can find out ‘whether there are any [substances] besides sensible substances’ and ‘whether there is a separable substance’, but that ‘we must first sketch the nature of substance’ (Aristotle, 1984, 1028b27–32). We can only say then that Aristotle’s Priority Thesis is limited to those cases where we deal with specific kinds of substances (‘the god’, ‘the man’), and not with the highest kinds of substances, namely sensible and non-sensible substances.

Prof. Whether: You’re right. Let me then talk about the Priority Thesis as stated by Saint Thomas Aquinas this time. In the first book of the Summa Theologica, Aquinas argues that the question about the ‘essence’ of a thing follows the question about the ‘existence’ of that thing.⁵ As you know, he attempts to prove the existence of God through the ‘five ways’, but he claims that concerning the question of the essence of God, he can only give a partial response because of the complexity of the question and the limits of human intelligence. While Aquinas focuses on God, he argues that for every kind of entity the ‘an sit’ (whether it is) always precedes the ‘quid sit’ (what it is): ‘When the existence of a thing [an sit] has been ascertained there remains the further question of the manner of its existence, in order that we may know what it is [quid sit]’ (Aquinas, 1952, I, q.3).⁶

Prof. What: As a matter of fact, Aquinas says something different in De ente et essentia, where he argues instead that we can ‘understand’ the essence of a certain kind of entity, such as ‘a man’ or ‘a phoenix’ without knowing whether that entity exists or not.⁷ I am not sure if and how we can make these different passages consistent, but it would be more prudent, for the time being, to attribute to Aquinas a version of the Priority Thesis restricted to God.

Prof. Whether: Alright, let’s have some regard for historians.

Prof. What: In any case, does the ‘an sit’ precede the ‘quid sit’ in a modal or a methodological sense?

Prof. Whether: What do you mean by that?

Prof. What: Well, the thing is, it’s not clear to me what you mean by the ‘we have to’ in expressions such as ‘we have to first know whether something exists, and only after can we figure out what it is’. Do you believe that it is something that we are necessarily committed to do when we do philosophy, or that it is something that we ought to prefer for methodological reasons, but in theory we can disregard?

Prof. Whether: I believe in the stronger modal thesis rather than in the weaker methodological one: I cannot see how we can do metaphysics of something if we do not know whether it exists or not. For example, let us assume that I ask you what you think about a certain person. I start describing them, telling you their name, their physical and psychological traits, but then I add: ‘However, I’m not sure whether this person exists or not’. Dumbfounded, you would certainly reply with something like ‘How on earth can I tell you what I think about this person if we’re not even sure whether they exist or not?!’.

Prof. What: But nothing stopped you from describing this person, even if you weren’t sure whether they existed or not! We can start by doing metaphysics. What prevents us from taking a set of metaphysical properties, such as being four-dimensional or being three-dimensional, etc., putting them together and trying to see if there is such a thing in the world? I understand it might turn out to be a ridiculously inconvenient task, but it is still possible.⁸

Prof. Whether: OK, you’re right, it is possible to do metaphysics before doing ontology. Thus, conceived in the methodological sense, our Priority Thesis states that it is always possible and preferable to do ontology before metaphysics. At this point, though, another clarification is in order. You just talked about metaphysical properties. Now, my question is: What distinguishes them from non-metaphysical properties? Or, in other words, what does doing metaphysics amount to?

Prof. What: This is a good question. I think it is not an easy task to delineate the scope of metaphysics, but I have something in mind, related to our linguistic competence. There are application conditions associated with the correct use of the words we use to refer to things, and these application conditions reflect common-sense truths about the world. For example, statues have certain properties such as being made of a certain material, having a certain purpose, and so on. By knowing the properties associated with statues, properties reflecting certain common-sense truths about the world, we can correctly use the term ‘statue’ and agree on what we are talking about. Now, when it comes to attributing properties that fall outside the range of properties that our common-sense truths tell us about, properties like being three-dimensional or four-dimensional, then we’re doing metaphysics.⁹

Prof. Whether: I have a worry with this view, but I can still endorse it if we amend it appropriately. Let me explain: I would rather talk about common-sense views instead of commons-sense truths, the reason being that we might be mistaken in holding certain common-sense views, but that does not necessarily prevent us from fixing the reference of our expressions.

Prof. What: That works.

Prof. Whether: OK, now we can be more specific what we mean by ‘inquiry into what it is’: by doing metaphysics, we attribute to the entity under investigation properties that are not contemplated in our common-sense descriptions of things. And we can say that by ‘metaphysical properties’ we mean those properties that do not figure in our common-sense views.

Prof. What: Isn’t the scope of metaphysics too broad if we define it this way though? Shouldn’t metaphysics be concerned with something more specific, like the fundamental nature of the world?¹⁰

Prof. Whether: That is certainly one way to define it, but I’d rather take a more liberal approach to what is metaphysics, as it is closer to the actual practice of doing metaphysics. If you think about it, metaphysics isn’t always concerned with fundamental entities. Take, for example, the metaphysics of the social world: arguably, this is not about the fundamental nature of the world, yet we consider it as proper metaphysics.

Prof. What: I guess that this way of defining it is a bit too inclusive, as it would consider as metaphysics all sorts of scientific endeavor, since sciences usually go beyond what common sense tells us about the world.

Prof. Whether: I don’t mind recruiting as part of metaphysics much of the sciences! Although, of course, we need to have a division of intellectual labor in place (and I’m aware that many scientists may not be too thrilled to be called ‘metaphysicians’…).

Prof. What: I’m also thinking that, on the other hand, one could argue that this view of metaphysics is not liberal enough. After all, you are denying that common-sense views of the world can be properly metaphysical.¹¹

Prof. Whether: This is true, but keep in mind that there is still an important place for common-sense views when we do ontology, as we can posit the existence of entities that are included in our common-sense ontology, such as ordinary objects. Thus, while we may not have a common-sense metaphysics, we can still have a metaphysics of common-sense entities.

Prof. What: As long as common-sense views may still play a role, I’m fine with that.

Prof. Whether: Now that we have found out what doing metaphysics amounts to, do you agree with me that the Priority Thesis is true, and that it is always possible and preferable to do ontology before metaphysics?

Prof. What: Although I agree that it is always preferable to do ontology before doing metaphysics, I also think that, unfortunately, that’s not always possible.

Prof. Whether: How so?

Prof. What: I believe that in some cases we have to characterize our entities in a metaphysical sense in order to be able to discuss them from an ontological point of view. And here I mean ‘we have to’ in a stronger sense than the methodological one.

Prof. Whether: Which kinds of entities are you thinking about?

Prof. What: The first kinds of entities that come to my mind are universals. If we want to talk about them, we have to characterize them metaphysically.

Prof. Whether: Why?

Prof. What: We could say that universals are abstract entities: this doesn’t seem like an attribution of a property implied by our common sense.

Prof. Whether: But we cannot say that only universals are abstract entities because this is also how philosophers often characterize other entities, for example numbers. Thus, this is not enough to distinguish universals from other kinds of entities.

Prof. What: OK, let me try with a different property. Universals are those entities that can be instantiated. The property of being instantiated characterizes universals in a specific way. Notice that this is only one metaphysical property, but that this is enough to restrict the scope of the Priority Thesis. I don’t need to add extra metaphysical properties like that of being abstract.

Prof. Whether: But realists argue for the existence of universals because they aim at explaining the phenomenon of predication. A sentence of the form subject-predicate is true if and only if the things in the world are like the sentence describes them. Realists believe that just as we need a referent for the subject in the sentence ‘Socrates is brave’ and that this referent is Socrates, we also need a referent for the predicate ‘brave’. We need to postulate the existence of something, ‘bravery’, which has a certain relation with Socrates, such that the sentence ‘Socrates is brave’ is true. We then infer that the relation that ties Socrates to bravery is a relation of instantiation: this is why, according to the realist, we can paraphrase ‘Socrates is brave’ with ‘Socrates instantiates bravery’.

Prof. What: I don’t get it: how is that different from what I said?

Prof. Whether: It is different because it seems to me that this argument in favor of the existence of universals doesn’t presuppose—it implies the metaphysical characterization of universals as instantiable entities. The realist starts with the goal of explaining predication, and then postulates the existence of entities that could be the referents of the predicates that are in a certain relation with the referents of the subject. Only after does the realist claim that this relation is that of instantiation. The realist doesn’t start with a metaphysical characterization of universals, but they first postulate their existence, and only then characterize them as having the property of being instantiated.

Prof. What: It seems to me that the metaphysical characterization of universals is part of the realist argument that comes together with the positing of their existence, rather than an implication of the argument. However, even if we admit that the metaphysical characterization of the universals is implied by the argument, there is still a problem: the ontological discussion of universals does not always stem from the problem of predication. When they do ontology, philosophers ask ‘Are there universals?’, not ‘How do we explain predication?’ Universals are already metaphysically characterized when the nominalist and the realist debate their existence.

Prof. Whether: Well, but you know that I’m a nominalist… I don’t have to deal with universals when I do ontology!

Prof. What: How can you do ontology without dealing with universals?

Prof. Whether: I consider doing ontology as listing things that exist, not listing both things that exist and things that do not exist. After all, Quine’s ontological question, the one you reminded me of at the beginning of this conversation, is ‘What is there?’ not ‘What is there and what is not there?’.¹²

Prof. What: That’s a nice way to describe how we go about our job, but even if I agreed with you on this, I’d still think that your view has a problem that stems from your commitment to nominalism. Your thesis holds only provided that we are nominalists about all those entities that need to be metaphysically characterized. But we’re concerned with meta-ontology here (and meta-metaphysics!), and so our account needs to be neutral about which ontology we endorse. Moreover, how would the nominalist reply to the realist’s claim that universals exist? A true nominalist won’t back out of the ontological debate by staying silent, and they would instead argue that there aren’t such things as universals.

Prof. Whether: But what if I told you that the ontological dispute over universals is only a ‘disguised’ metaphysical debate? After all, the realist believes that there are things like qualities and relations, namely all those entities that can be expressed by one-place predicates and many-place predicates, and they believe that these entities are universals.¹³ Since when it comes to qualities and relations, we don’t need to attribute any metaphysical property to characterize them in order to talk about them, ontology would still precede metaphysics. This also explains the difference between trope theory and realism. A trope theorist believes, like the realist, that qualities and relations exist, but unlike the realist, they claim that they are abstract particulars: Socrates is not brave in virtue of the universal bravery, but in virtue of that abstract particular which is Socrates’ bravery. Thus, we have a metaphysical disagreement, not an ontological one: the trope theorist claims that qualities and relations are particulars, therefore entities that do not enjoy the property of being instantiated; whereas for the realist, qualities and relations are universals, therefore entities that enjoy the property of being instantiated. The question ‘Are there universals?’ actually stands for the question ‘What are qualities and relations?’.

Prof. What: Again, I think that we face the problem of explaining how nominalists can express their ontological disagreement about universals. It is true that the nominalist would probably deny the existence of properties and relations in the first place, and that therefore they wouldn’t have any interest in their metaphysical characterization, but there are arguments whose purpose is precisely to refute the existence of what is metaphysically characterized as a universal. For example, nominalists often object that if we accept the relation of instantiation between a universal and the particulars, we also have to accept the multiple location of the universal, a feature they don’t find plausible, the reason being that we would then have to admit as meaningful and true seemingly absurd sentences such as ‘wisdom is four miles away from itself’ if there are two wise people four miles away from each other.¹⁴

Prof. Whether: However, would you agree that the question ‘Are there universals?’ can be a metaphysical question disguised as an ontological question?

Prof. What: Yes, I agree with you, but it can be a genuine ontological question as well. However, now I want to show you a case of a question that seems ontological and is ontological! And, more importantly, it requires us to do some metaphysics first.

Prof. Whether: Go on.

Prof. What: The question is… ‘Are there possible worlds?’ According to some philosophers, quantification over possible worlds must be taken seriously, but… what are possible worlds? We know that for David Lewis possible worlds are concrete and real worlds, but ones that are spatiotemporally and causally disconnected from ours (Lewis, 1986). But this is not the only way to characterize them. According to other modal realists like Alvin Plantinga, for example, possible worlds are maximal possible states of affairs (Plantinga, 1974).

Prof. Whether: I see what you’re getting at…

Prof. What: The example of modal realism shows us even more clearly than that of universals that sometimes metaphysics has to precede ontology. Otherwise, how can we argue the existence of such things as possible worlds without metaphysically characterizing them in any way? For both Lewis and Plantinga, possible worlds are entities about which the common speaker has no opinion whatsoever! It doesn’t seem like there’s a way to characterize possible worlds such that would appeal only to our common-sense views. Characterizing possible worlds as ‘those entities in virtue of which modal sentences are true or false’, which is something that both Lewis and Plantinga appear to agree on, doesn’t seem like an attribution of properties merely implied by common sense.

Prof. Whether: Here’s another way to describe possible worlds though: ‘ways in which our world could have been if things had been different from how they actually are’. This way of characterizing possible worlds does not require any metaphysical characterization, and we could say that it is a common-sense way to look at them. You should keep things easy if you can!

Prof. What: You are definitely right; we can refer to possible worlds by using this non-metaphysical definition. However, when philosophers like Lewis and Plantinga argue that possible worlds exist, they do so by attributing the property of being entities in virtue of which modal sentences are true or false. Thus, the ontological commitment takes place thanks to the attribution of a metaphysical property: philosophers posit the existence of possible worlds because, in their view, they are what makes modal sentences true or false. And if we maintain the broad characterizations of ‘ontology’ and ‘metaphysics’ we previously discussed, we could find similar examples in sciences, when the existence of entities with specific properties is hypothesized to explain certain phenomena (take the theorization of the Higgs boson as an example of this kind of approach).

Prof. Whether: Wait a minute, entities with very different natures can be entities in virtue of which modal sentences are true or false, so it is not true that by characterizing possible worlds this way we are saying what they are.

Prof. What: You are right, but only to the extent that this characterization does not give us a full metaphysical picture of what possible worlds are. But we are still doing metaphysics, since we are attributing a metaphysical property to the entity: remember that we said that doing metaphysics is attributing to the entity under investigation certain properties that are not part of our common-sense views.

Prof. Whether: Fair enough.

Prof. What: Great! Now, let me add another way in which metaphysics can precede ontology, namely when a metaphysical claim has ontological implications. For example, let’s suppose that we found an argument that convinced us of the existence of numbers. At this point, we can fulfill our metaphysical duty and try to figure out what these entities are. Let’s suppose now that we start characterizing numbers, and we say that numbers are abstract. If, by doing ontology, we didn’t previously admit the existence of abstract entities—and let’s not forget that our ontology proceeds piece by piece, and therefore it doesn’t force us to consider the existence of anything we can imagine—then these entities will be new in our ontology. If we have found out that certain entities are abstract, then we have found out that there are some abstract entities.

Prof. Whether: However, unlike with universals or possible worlds, the case here is different because we already established what exists, and only after we give a metaphysical characterization that has an ontological implication. This is a weaker sense in which metaphysics precedes ontology.

Prof. What: Right.

Prof. Whether: OK, then the Autonomy Thesis is true, but we need to revise the Priority Thesis: it is always preferable to do ontology before doing metaphysics, but this is not always possible. Great, now we can start doing ontology.

Prof. What: …or metaphysics if we need to!