The Atlas of Reality
Page 44
1. Common sense. As we've seen, it seems reasonable to say that many things, golden mountains and round squares among them, don't exist. Adopting a common-sense approach to the truth of these claims leads naturally to Anti-Actualism.
2. Intentional objects. Human beings can, apparently, bear de re attitudes like belief, though, desire, and intention toward merely possible or even impossible objects. An example of such an attitude would be that of seeking the Fountain of Youth. A de re attitude is an attitude directed toward a particular thing, in contrast to a de dicto attitude, which takes a property or a description as its object.
Saul Kripke, in his classic Naming and Necessity (Kripke 1980), builds a strong case for the existence of de re attitudes toward individual objects that are directed toward those objects not by way of some general description. Suppose, for example, that one is seeking O'Hare Airport in Chicago because one has a reservation on a flight that leaves from there. One has beliefs, thoughts, and desires that are directed toward a certain thing, namely O'Hare Airport. These attitudes cannot be identified with any corresponding attitudes toward a certain kind of thing. For example, in seeking O'Hare, one is not seeking the largest airport in the Midwest; at least, one isn't seeking it under that description. One might not know or care that O'Hare is the largest airport in the Midwest. Similarly, one is not seeking an airport that is commonly named ‘O’Hare'. Again, one doesn't really care what it's called. If, unbeknownst to one, the airport's name had been changed to ‘Daley Airport’, one's seeking would have the same object: one would still be trying to get there, whatever it's called.
Meinongians point out that such de re attitudes can also be directed toward non-existent or even impossible objects. Ponce de León was seeking the Fountain of Youth, ancient Greeks worshipped Zeus, and many would-be detectives admire and emulate Sherlock Holmes.
At any rate, Anti-Actualists can claim that we ought to posit non-existent objects to function as the objects of these intentional attitudes.
A simple Actualist response to this problem is to accept that we do have de re attitudes in such cases, and such attitudes do involve our being related to something as the object of the attitude. However, Actualists can deny that we need non-existent things to serve as those objects. Instead, they can press into service abstract entities of various kinds: properties, states of affairs or propositions. So, when Ponce de León was seeking the Fountain of Youth, he was in the seeking attitude toward the state of affairs in which he finds the Fountain of Youth or the property of being a finder of the Fountain of Youth or the proposition with the content that he himself finds the Fountain of Youth. All of these abstract entities, Actualists might claim, actually exist even if there is no Fountain of Youth.
However, this view does involve some cost. From a naive point of view, it seems that Ponce de León did not seek a property or a centered proposition or anything of the kind. He sought something that, were it to exist, would be a fountain, something concrete. Properties and propositions, in contrast, are necessarily abstract.
In addition, Actualists face a dilemma. They must either suppose that all of our de re attitudes are actually directed toward abstract objects, even when there is a real, concrete object available (like O'Hare Airport) or they must give two separate accounts of de re attitudes, one for cases in which an appropriate object exists and another for cases in which no such object exists.
Both approaches come with some cost. We should always prefer a unified account to a combination of disparate accounts. This is another aspect of Ockham's Razor. If Actualists try to give a unified account in terms of abstract objects like properties or concepts, they will then have to answer the objections to such descriptivist accounts raised by Kripke in Naming and Necessity. When one seeks O'Hare, one is in a particular relation to a concrete object. Similarly, if one sees Bill Clinton, one sees him, the former President, and this perceptual relation cannot be cashed out in terms of a relation to any description, such as seeing the forty-second President or seeing the man named ‘Bill Clinton’.
Probably the best Actualist account is one proposed by Alvin Plantinga in The Nature of Necessity (Plantinga 1974). Plantinga introduces the notion of a haecceity or thisness (Def 9.2). By way of reminder, a thisness or haecceity is a property that of necessity is possessed by at most one thing and could not be possessed by any other thing. The properties of being O'Hare Airport or being Bill Clinton are examples of haecceities. O'Hare Airport necessarily has the property of being O'Hare Airport (if it exists at all), and nothing other than O'Hare could have that property. Plantinga proposes that haecceities, like other properties, exist of necessity. O'Hare itself does not exist of necessity, but if it didn't exist, its haecceity, though uninstantiated, would still exist.
If there were such things as haecceities, then they could serve as the objects of all de re attitudes. When one seeks O'Hare Airport, one is really seeking to be in a place that instantiates the haecceity of being O'Hare Airport. When one sees Bill Clinton, one is seeing the object that instantiates the haecceity of being Bill Clinton. Similarly, when Ponce de León was seeking the Fountain of Youth, he was seeking to be in a place that instantiated the haecceity of being the Fountain of Youth. Although the Fountain of Youth does not exist, its haecceity might, and it might therefore serve as the object of Ponce de León's seeking attitude.
Other Actualists, like Robert M. Adams (1981), have expressed skepticism about whether there really are such things as uninstantiated haecceities. Adams has argued that the only haecceities that exist are the haecceities of actual objects, since Bill Clinton's haecceity contains Bill Clinton himself as a part or component. In Chapter 16, on de re modality, we take up again the issue of whether there are such things as haecceities, including uninstantiated ones.
Another Actualist response to this problem about intentional objects is to suppose that, strictly speaking, our intentional states do not relate us to things as their objects. Instead, we should re-interpret claims about intentional states using only adverbs. If Smith believes in unicorns, it is not the case that he stands in the believing relation to some unicorns or even to the set of unicorns or the property of being a unicorn. Instead, she believes unicornly. Similarly, Ponce de León was seeking in a Fountain-of-Youthish way.
Roderick Chisholm (1973a) pointed out a problem for this adverbial account. The following argument seems valid:
(29) Jones is thinking of unicorns.
(30) Jones thinking of something.
(31) Therefore, there are unicorns.
The adverbial account cannot explain the logical structure of these propositions in such a way as to give an account of the argument's validity that isn't entirely ad hoc. In other words, adverbialists must introduce a large number of new logical principles governing adverbial constructions. Possibilists, in contrast, can explain the argument's validity much more simply, using just the usual logical rules of the quantifiers. So, attributing Jones's thought to her should involve placing her in some relation to unicorns.
3. Fictional and illusory objects as persistent and available to many minds. We've already mentioned Peter Geach's imagined situation in which two medieval peasants, Hob and Nob, share a belief in a non-existent witch who has cursed their village (1967). It seems that there must exist in some sense an object that is believed by both Hob and Nob to exist and to be a witch in their village. There need not be a description or property that Hob and Nob agree upon that could be used to define the common object in their shared fantasy. Such shared objects seem to require exportation. Hob believes that there is a witch, Nob believes there is a witch, and the witch that Hob believes in is identical to the witch that Nob believes in. In order to make sense of this, there must be a non-existent witch believed in by both villagers.
It is true that Hob could define the witch in terms that refer to Nob's beliefs. For example, Hob could think of the witch as the witch that Nob believes cursed his cow last month. If Hob believes that that wit
ch also cursed his chickens today, there is a sense in which we can identify an object in Hob's imagined world with one in Nob's imagined world. However, it is quite possible that Hob and Nob could begin to refer to ‘the witch’ in their shared conversations after they have both forgotten who first thought of it. At that point, we cannot define ‘the witch’ in terms of any property or set of properties, whether ordinary or intentional.
A similar problem arises from the persistence through time and change of fictional and illusory objects. The same non-existent object can be the object of thought or perception at different times by the same person or by different people at the same time. These non-existent objects cannot simply be identified with sets of properties. Sherlock Holmes, for example, acquired more representational properties as Conan Doyle wrote new stories.
4. Non-existent objects as components of merely possible states of affairs. There could have been more things in existence than actually exist. There are possibilities according to which some things exist that do not exist in the actual world. If something exists according to a possibility, then it is in some sense a component or constituent of that possibility. The possibility consists of that very thing, standing in certain possible relations to other things. If that is right, then we have another reason to countenance non-existent objects.
Alvin Plantinga's (1974) Actualist response again involves haecceities. The possibilities apparently involving non-existent things really contain only uninstantiated haecceities. Again, critics of Plantinga, including Robert M. Adams, worry that haecceities involve non-existent things no less than possibilities do. Our haecceitiess (the properties of being Robert C. Koons and of being Timothy H. Pickavance) include RCK and THP themselves as constituents, say these objectors. Hence, uninstantiated haecceities must contain non-existent things. If Adams is right, then haecceities don't blunt this Anti-Actualist argument.
There is another problem with Plantinga's haecceity proposal. We can't make significant de re assertions about the relations between things and their haecceities. For example, we cannot say, in an informative way, that necessarily, if Socrates exists, his haecceity is instantiated. This becomes (on Plantinga's account) a mere logical tautology, instead of a substantive metaphysical proposition. It turns out to be nothing more than to say that necessarily, if Socrates' haecceity is instantiated, then it is instantiated, which isn't what we really want to say. We want to say something about the necessary relations between things and their haecceities, and yet this important feature of Plantinga's theory seems to be inexpressible if the theory is true.
Another Actualist response is what David Lewis (1986a) called the “ersatz” Actualist response. According to this response, a possible world is a kind of representation. So a possibility according to which non-actual things exist is not one that literally contains non-existent things or mysterious haecceities as constituents. Instead, it contains things that represent the possible existence of non-actual things.
This ersatz Actualist has a problem representing iterated de re possibilities, possibilities involving particular things. For example, there could exist a world undergoing eternal cycles without beginning or end, each cycle qualitatively like all the others, and each one containing a duplicate of Napoleon. If such a world, w0, had been actual, then each “Napoleon” in w0 could have been the very last “Napoleon” in another world. These further possibilities, or possible possibilities, should correspond to an infinite number of distinct worlds, each one with a different Napoleon from w0 as its last.
The problem for ersatz Actualists is that they have no way to distinguish the many different de re possibilities from one another. For example, there should be one possibility, wt, corresponding to this Napoleon's (picking out one of the infinite series from w0) being the last and another, different possibility wt+1 in which the next Napoleon is the last one. However, for the ersatz realist, these two possibilities would be the same. Both would contain an infinite series of qualitatively similar “Napoleon”-representations with an end but no beginning.
We can make the same point using a Max Black-style possible world wMB. Suppose that the original Big Bang in our past had had the power to produce two exactly similar and symmetrically arranged black holes. Suppose further that the two black holes had had a kind of free will, or some other unpredictable and undetermined power to act. Let's say, for example, that each of the two black holes might have destroyed itself; w1 is a world where one black hole destroys itself and w2 is a world where the other black hole destroys itself. Ersatz Actualists have a problem explaining how it is possible for w1 and w2 to be distinct. What could distinguish them? They involve the same intrinsic properties and the same spatiotemporal relations to all of the inhabitants of the actual world. Could ersatz Actualists say that there is just a brute distinctness between the two possible worlds? This doesn't seem right. The two possible worlds are distinct because the black holes are (or would be) distinct.
There is really only one way out for Actualists: give up the modal logical principle S4, which says that what is necessary is necessarily necessary, and what is possibly possible is possible. Without S4, ersatz Actualists could argue that there are not in fact two such possibilities (one in which one black hole destroys itself, and another in which the other does), although there could have been two such possibilities. If wMB had been actual, there would have been w1 and w2, but because it is not, there are no such possibilities. In other worlds, these two further possibilities are themselves only possibly existent, and so the things that exist according to them are only possibly possible, not actually possible. This is probably the best response for Actualists, although it comes with some significant theoretical cost, namely, the loss of the simplest and most powerful modal logic, S5, which includes S4 (see Chapter 16 for a further discussion of de re possibilities).
12.1.6 Defining existence
Throughout this chapter we've been discussing existence, and yet we have never said what existence itself is. There's a good reason for this omission. For Actualists, it is arguably impossible to define existence, since existence is not a property that some things lack and other things have. Instead, it is trivially true that absolutely everything exists.
In contrast, Anti-Actualists can sensibly ask the question of whether existence is definable. For Possibilists, it is actual existence that must be defined, since they hold that it is trivially true that everything has at least possible existence. We will look in detail at actuality in Chapters 14 and 15. Consequently, we will focus here on Meinongian definitions of ‘existence’.
Importantly, we cannot assume that absolutely every property or concept can be defined. Some notions are so fundamental that they cannot be defined. Existence might be one such notion. Nonetheless, some philosophers have proposed something like definitions of existence. For example, George Berkeley argued that esse est percipi, that to exist is to be perceived. To be completely accurate, we should say that for Berkeley there are two ways to exist: to perceive or to be perceived. If Berkeley were right, the very idea that something could exist while neither perceiving nor being perceived would be absurd. However, it does not seem absurd. It seems relatively easy to imagine such a thing, so by Imagination as Guide to Possibility (PEpist 1), it is reasonable to think it is possible to exist without perceiving or being perceived.
However, Berkeley had a response to this argument. He claimed that we cannot in fact imagine something that is unperceived, since the very act of imagining something involves imagining something to be perceived by one's self. Berkeley is clearly wrong here because he failed to distinguish between the act of imagining and the content of the act. Just as we see colors and do not see our own acts of seeing, so too we can imagine colored things without imagining our act of imagining them. I can imagine a tree falling in a lonely forest, making a great crashing sound, without imagining anyone's hearing the sound.
Samuel Alexander (1920) proposed a different definition of existence: to exist is to act
, to have some effect on something else or at least to have the power to act upon something. This is Alexander's Dictum, sometimes called the Eleatic Principle, from a passage in Plato's Sophist in which a character, the Eleatic Stranger, suggests that having causal power is the essential mark of being (Sophist 247d3).
The trouble with Alexander's Dictum is that it seems possible that something exist without active causal powers. We can easily imagine something, like an invisible or ghostly observer, that is affected by other things but forever unable to have any effect. So, we might consider modifying Alexander's Dictum to the claim that to exist is to have some causal power, either active or passive. Things with causal powers have something further in common, namely, some intrinsic nature or quality. We should, therefore, entertain the possibility that to exist is to have some intrinsic nature.
This suggestion accords nicely with Anti-Actualism's problem with truthmaking. Recall that Meinongians could solve this puzzle by assuming that non-existent objects have only modal and intentional properties. Such entities would have no intrinsic natures, and so would lack existence in that final sense.
12.2 Ontic Vagueness
In our discussion of Monism (Section 11.2.4), we raised a question from Jonathan Schaffer (2010): can there exist things with vague boundaries, whether spatial, mereological, or temporal? Something with a vague spatial boundary would have a somewhat indeterminate location. We couldn't say (truthfully) precisely where the thing is at any point in time. A mereologically vague boundary would involve some vagueness in the composition of a thing. There would be material bits that are neither definitely a part of the thing nor definitely not a part. A thing would have a vague temporal boundary if there is no precise point in time at which it begins or ceases to exist.