Jump to ContentJump to Main Navigation
On What MattersVolume Two$

Derek Parfit

Print publication date: 2011

Print ISBN-13: 9780199572816

Published to Oxford Scholarship Online: May 2015

DOI: 10.1093/acprof:osobl/9780199572816.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 25 February 2017

(p.719) Appendix J On What There Is

(p.719) Appendix J On What There Is

Source:
On What Matters
Publisher:
Oxford University Press

1

Rather than being Actualists, who believe:

There is nothing except what actually exists,

we ought, I have claimed, to be Possibilists, who believe:

There are some things that are merely possible.

I have also made some conceptual claims. According to

the Plural Senses View: There is one wide, general sense in which we can claim that there are certain things, or that such things exist. We can also use these words in other, narrower senses. For example, if we say that certain things exist in what I call the narrow actualist sense, we mean that these things are, at some time, actually existing concrete parts of the spatiotemporal world.

As Possibilists, we should claim:

There are in the wide sense some possible things that never exist in this actualist sense.

There are also, I have claimed, some abstract entities, such as some logical and normative truths, which exist in a distinctive, non-ontological sense. I shall here develop and defend these claims. In defending Possibilism, my main aims are to defend the Plural Senses View, and the implications of both views. Possibilism is the thin end of a wider wedge.

Some Actualists say:

Nothing actually exists except what actually exists.

(p.720) But Possibilists would accept this trivial claim. For Actualism to be a significant view, Actualists must claim:

Nothing exists in any sense except what actually exists.

Many Actualists would deny that there is any other sense in which things might exist. These people assume

the Single Sense View: The words ‘there are’ and ‘exist’ must have only the same single sense.

Plantinga, for example, writes that, when Possibilists say that

  1. (A) there is some entity that is merely possible and never actually exists,

this claim is ‘monumentally perplexing’, since (A) means that

  1. (B) there ‘is a thing such that there is no such thing’.

This remark is surprising, since Plantinga earlier wrote:

What might it mean to say that there are some individuals that do not exist? … Perhaps we can say something about what is not meant. It is not suggested, of course, that there exist some things that do not exist, ‘exist’ being taken the same way in each occurrence.

When Plantinga discusses (A), he seems to forget this earlier claim, since he assumes that Possibilists are making the contradictory suggestion that there exist some things that, in the same sense of ‘exist’, do not exist.

Few people give arguments for the Single Sense View. When he defends this view, for example, Quine merely writes:

There are philosophers who stoutly maintain that ‘exists’ said of numbers, classes, and the like and ‘exists’ said of material objects are two uses of an ambiguous term ‘exists’. What mainly baffles me is the stoutness of their maintenance. What can they possibly count as evidence? Why not view [‘exists’] as unambiguous but very general . . ?

(p.721) If there could be no evidence for the view that ‘exists’ can have two senses, there could also be no evidence against this view, and Quine should also be baffled by the stoutness with which some people defend the Single Sense View.

There is, however, some evidence for the Plural Senses View. Quine himself writes that, if some word or phrase ‘can be clearly true or false of one and the same thing’, that is ‘the nearest we have to a clear condition of ambiguity’. Quine’s example is the claim that

  1. (C) dark feathers are light.

Since dark feathers are light in weight but not in colour, (C) is in one sense true and in another sense false. As this example shows, the word ‘light’ is ambiguous, having two senses. Return next to

  1. (D) There was a palace designed by Wren to replace the burnt Palace of Whitehall.

This claim is also in one sense true and in another sense false. We can truly say

  1. (E) There was such a possible palace designed by Wren, but this palace was not built and never actually existed.

This example shows that, on Quine’s proposed criterion, the phrase ‘there was’ has at least two different senses. There was such a palace in the wide sense, but not in the narrow actualist sense.

Single Sense Theorists might reply:

  1. (F) What makes (D) ambiguous is not the phrase ‘there was’ but the word ‘palace’. This word has two senses, since it can mean either ‘possible palace’ or ‘actual palace’.

On this view, we should replace (E) with

  1. (G) There was such a possible palace designed by Wren, but since this palace was not built, there was no actual palace designed by Wren.

(p.722) But (F) is implausible, and (G) supports both Possibilism and the Plural Senses View. (G) tells us that there was a palace that was merely possible, because this palace never became actual. That is another way of claiming that there was such a possible palace, though this palace never existed in the actualist sense.

In the passage quoted above, Quine is defending the view that

  1. (H) material objects and abstract entities can both be claimed to exist in the same ‘very general’ sense.

It is often assumed that, if we accept (H), we thereby accept the Single Sense View. But that is not so. We could accept both (H) and the Plural Senses View. We could claim

  1. (I) There are, in the wide sense, both material objects and abstract entities.

But we would add:

  1. (J) As well as existing in this wide sense, many material objects also exist in the actualist sense, by being actual concrete parts of the spatio-temporal world. Abstract entities do not exist in this narrower sense, nor do material objects that are merely possible.

Quine’s remarks provide no argument against this view.

Other writers deny the distinction drawn by (J). Stalnaker writes that, as an Actualist, he believes that

  1. (K) ‘existing and actually existing are the same thing. There exists nothing that is not actual.’

This claim, Stalnaker remarks, should not be understood as ‘a restrictive metaphysical thesis’. This remark suggests that (K) does not conflict with any metaphysical view. That would be true if there was a difference (p.723) between the meanings of the phrase ‘there are’ and the word ‘exist’, so that Possibilists could accept (K) but add that

  1. (L) there are, in the wide sense, some things that are merely possible, and never actually exist.

Stalnaker also calls (K) ‘a trivial consequence of the meaning of the word “actual” ’. This remark again suggests that (K) and (L) do not conflict. It could not be a consequence of the meaning of ‘actual’ that the phrase ‘there are’ cannot be used in this wide sense.

When Stalnaker calls (K) a trivial consequence of the meaning of ‘actual’, his point may be that ‘actually’ can be used, like ‘truly’, in a way that reinforces any indicative statement or assertion, without adding anything to its meaning. Instead of saying ‘X’, we could always say ‘Actually X’. But this fact does not support Actualism. Possibilists could use this sense of ‘actually’, and restate (L) as

  1. (M) There actually are, in the wide sense, some things that are merely possible and never exist in the actualist sense.

Stalnaker’s Actualism seems, however, to be a metaphysical view, so he might reject both (L) and (M). When he claims that (K) is not a restrictive metaphysical thesis, Stalnaker may mean that, though (K) makes a metaphysical claim, there is no other intelligible or coherent metaphysical view. He may assume that

  1. (N) the words ‘there are’ and ‘exist’ must have only the same single sense, which means ‘actually exist’.

We can call (N) the Actualist Single Sense View. If ‘there are’ must mean ‘there actually exist’, Possibilists could not coherently claim that there are some things that are merely possible, and never actually exist. Though (N) could not follow from the meaning of the word ‘actually’, many Actualists assume (N).

Rather than merely assuming the Single Sense View, van Inwagen vigorously defends this view. Van Inwagen rejects the very idea of a (p.724) merely possible concrete object, such as a merely possible horse, or human being. ‘Like “round square”,’ he writes, ‘ “non-actual horse” is a contradiction in terms.’ But if we say

  • (O) There was a possible palace that was never actual,

that is not a contradiction. Van Inwagen might reply

  1. (P) Nothing that isn’t actual could be a palace.

But in the sense in which (P) is true, (P) means that nothing that isn’t actual could be an actual palace. This claim does not conflict with (O), which isn’t a claim about an actual palace.

Like Stalnaker, Van Inwagen seems to assume that

  1. (Q) since the word ‘actually’ adds nothing to the content of any assertion, we can truly assert that everything that exists actually exists.

But (Q) does not support the Single Sense View. If the word ‘exists’ can be used in both the wide and actualist senses, we could claim

  1. (R) Everything that exists in the wide sense actually exists in this sense, and everything that exists in the actualist sense actually exists in this sense.

What (Q) shows is that, in explaining the actualist sense, it is not enough to say

Something exists in the actualist sense if this thing actually exists.

That is why I claimed

Something exists in the narrow actualist sense if this thing is an actual concrete part of the spatio-temporal world.

Return now to the claim that

  1. (E) There was a possible palace designed by Wren, but this palace was not built so that it never actually existed.

(p.725) If the Single Sense View were true, (E) would mean

  1. (S) There actually existed such a possible palace, but this palace was not built so that, in the same sense of ‘exists’, this possible palace never actually existed.

This claim is a contradiction, which could not possibly be true. But (E) does not mean (S). If we use my definitions and the redundant sense of ‘actually’, (E) could be more fully stated as

  1. (T) There actually existed in the wide sense such a possible palace, but this palace was not built so that it never actually existed in the actualist sense.

Unlike the contradictory (S), this claim is coherent and might be true.

Van Inwagen might object that (E) and (T) are not coherent, because the words ‘there are’ and ‘exist’ cannot have two such different senses. There are, van Inwagen writes, ‘two clear and compelling arguments’ for the Single Sense View. According to one of these arguments:

When we say ‘There are some Xs’, we mean ‘The number of Xs is greater than zero’.

The phrase ‘The number of … is greater than zero’ has only one sense.

Therefore

The phrase ‘There are some Xs’ has only one sense.

But if ‘there are’ has two senses, A and B, we could coherently claim both

The number of Xs that there are, in sense A, is greater than zero,

and

The number of Ys that there are, in sense B, is greater than zero.

(p.726) For example:

The number of possible buildings that there are in the wide sense is greater than zero,

and

The number of actual buildings that there are in the actualist sense is greater than zero.

These claims conform to van Inwagen’s second premise, since they both use the phrase ‘The number of … is greater than zero’ in the same sense. But these claims use the phrase ‘there are’ in two different senses. So this argument does not show that this phrase has only one sense.

Van Inwagen also argues:

When we say ‘There exists an F’ what we mean is equivalent to ‘It is not the case that everything is not an F’.

The word ‘not’ has only one sense.

Therefore

The phrase ‘There exists an F’ has only one sense.

We could reply:

It is not the case that everything is not a possible building, nor is it the case that everything is not an actual building. As these facts might show, there is one wide sense in which there are both possible and actual buildings. But some of the possible buildings do not also exist in the actualist sense.

Van Inwagen suggests another argument, which takes the form of a funny story. He is discussing Meinong’s view that words like ‘there are’ and ‘exist’ can have one sense when they are applied to abstract entities, such as numbers or mythical beings, and can have another sense when they are applied to physical objects, such as buildings or rocks. On this view, we might say that there are some (p.727) abstract objects which do not, in the other sense, exist. Van Inwagen’s story goes:

One day my friend Wyman told me that there was a passage on page 253 of Volume IV of Meinong’s Collected Works in which Meinong admitted that his theory of objects was inconsistent. Four hours later, after considerable fruitless searching, I stamped into Wyman’s study and informed him with some heat that there was no such passage. ‘Ah’ said Wyman, ‘you’re wrong. There is such a passage. After all, you were looking for it: there was something that you were looking for. I think I can explain your error; although there is such a passage, it doesn’t exist. Your error lay in your failure to appreciate this distinction.’ I was indignant. My refusal to recognize a distinction between existence and being is simply my indignation, recollected in tranquillity and generalized.

Though this joke is funny, it does not apply to Meinong’s view. Since Wyman accepts Meinong’s view, he would not have claimed that there was such a passage, in the sense that applies to abstract entities. Such a passage in a printed book would not have been an abstract entity but a sequence of visible marks on a physical object. Wyman would have claimed only that no such passage existed. Van Inwagen’s indignation was not justified.

I shall now tell another story, about the view that Plantinga asserts and van Inwagen defends. My story goes:

As Plantinga leaves the room, he tells me that one actually existing state of affairs is that my wife is dead. I am struck with horror and grief. Four hours later, when he returns, Plantinga says: ‘Don’t worry. Though this state of affairs actually exists, it isn’t actual. Your wife is alive and well.’

After my needless hours of grief, my indignation would be justified. On Plantinga’s view, merely possible states of affairs actually exist, and they exist ‘just as serenely as your most solidly actual state of affairs’. But these actually existing states are not, Plantinga claims, actual, in the sense of being actualized, or obtaining. Given Plantinga’s definitions, (p.728) his claims are coherent. But when my wife is alive and well, it may be misleading to claim that one actually existing state of affairs is that my wife is dead. Unlike my wife’s actual state of being alive, her possible state of being dead is not real, or is at least less real, and can therefore be plausibly claimed to exist only in the wide sense. That is why it matters whether someone is actually or merely possibly dead. As Fine writes:

there is an ontological difference between actual objects and merely possible objects … We might call someone who takes actuality seriously an actualist.

Possibilists like me are also, in this sense, actualists. We believe that being actual is ontologically very different from being merely possible. That is why we claim that, though there is one wide sense in which there are both actual objects and objects that are merely possible, it is only the actual objects that also exist in the narrower, actualist sense.

There is also an ontological difference between concrete objects, such as rocks and stars, and abstract entities, such as numbers and logical truths. We can therefore defensibly claim that, though both kinds of entity exist in the same wide sense, these kinds of entity also exist in different, narrower senses. I shall return to this claim.

Van Inwagen’s arguments for the Single Sense View do not, I conclude, succeed. Nor, I believe, could any such argument succeed. Such arguments could at most show that everything that exists should be claimed to exist in the same wide sense. These arguments could not show that we cannot also intelligibly use other, narrower senses. To illustrate this point, we can turn from the concept being or existing to the concept doing. Consider:

accidentally killing someone, stumbling over a hidden stone, forgetting something, digesting food, growing older, contracting measles.

These can all be claimed to be things that, in a wide sense, we do. But we can also use the word ‘do’ in a narrower sense, which applies only to voluntary and intentional acts. The things just listed are not, in this sense, things that we do. No argument could show that the word ‘do’ cannot be intelligibly used in such different senses. Nor could (p.729) any argument show that the words ‘there are’ and ‘exist’ cannot have similarly different senses.

Some Actualists reject Possibilism with surprisingly extreme remarks. Plantinga calls this view ‘monumentally perplexing’, and Lycan calls it ‘literally gibberish or mere noise’. These people may be misled by the fact that

  1. (U) the word ‘actually’ adds nothing to the content of an assertion.

This fact may suggest that Possibilism is false, since everything that exists actually exists. But as Plantinga points out, the word ‘actually’ can be misleading. (U) could not show that the words ‘there are’ and ‘exist’ cannot be intelligibly used in different senses. (U) could show only that

  1. (V) when something exists in any of these senses, this thing actually exists in this sense.

It could still be true that

  1. (W) though something actually exists in one of these senses, this thing does not actually exist in some other sense.

In my example:

  1. (X) There actually was, in the wide sense, a possible palace designed by Wren to replace the Palace of Whitehall. This possible palace was not built and therefore never actually existed in the actualist sense.

(X) is not ‘monumentally perplexing’ or ‘literally gibberish’, but a clear and coherent claim.

2

According to

Possibilism: There are, in the wide sense, some things that are merely possible, and never actually exist.

(p.730) Since Actualists cannot appeal to the Single Sense View, they cannot reject Possibilism as incoherent, or self-contradictory. But they might claim that Possibilism is false. Actualists might say:

It could not be in any sense true that there are some things that are merely possible.

In discussing this view, we can first consider, not persisting things such as buildings or people, but acts and other events. Many Actualists ignore events. When these people deny that there are any merely possible entities, they often discuss farfetched examples. Quine suggests that, when people claim that there are such entities, their ‘main motive’ is to be able to make claims about mythical beings, such as the winged horse Pegasus. And Burgess and Rosen write:

Among wilder metaphysical entities are possibilia, unactualized possible worlds and the unactualized possible entities that inhabit them.

But such entities include anything that we could have done, such as the knock that we should have knocked before opening someone else’s bedroom door. There is nothing mythical or wild in such merely possible events.

Rather than merely ignoring events, some Actualists claim:

  1. (A) There are no events. There are only persisting things, such as people, rocks, and stars.

Some of these people argue:

  1. (B) We cannot justifiably believe that there are entities of some kind unless there are facts that we cannot adequately describe except in ways that refer to such entities.

  2. (C) Whenever someone makes some true claim which seems to refer to some event, we can restate this claim, or adequately redescribe the relevant facts, without referring to any event.

(p.731) Therefore

We cannot justifiably believe that there are any events.

We ought, I believe, to reject (B). Consider, for example, the facts that

  1. (D) there are some happily married couples, mountain ranges, and clusters of stars.

We could redescribe such facts in ways that referred only to the relations between various people, mountains, and stars. So (B) mistakenly implies that we could not justifiably believe (D).

Though we can truly claim that there are such happy couples, mountain ranges, and clusters of stars, we should admit that these composite entities are not fundamental. Actualists might similarly claim:

  1. (E) When we describe what fundamentally exists, we need not mention events. It is enough to refer only to persisting things.

If (E) seems plausible, that may be because in ordinary English we do not say that events exist. We would not, for example, say that the First World War came into existence in 1914, continued to exist for four years, and then ceased to exist. We would say instead that this war occurred during these four years. But there was a First World War, and a Second World War, and we can hope that there will not be a Third World War.

Nor are events less fundamental than persisting things. When some persisting thing hardly changes, as when some rock stays on the surface of the Moon for a million years, we might call this thing a very boring event. That claim would be a category mistake, since it is really this thing’s history that is very boring. But such claims provide a different and acceptable way of redescribing some parts of reality. Though the Sun is a persisting object, and the Great Fire of London was an event, we could think of the Sun in a different way, as a much greater and longer lasting fire. When things change, in contrast, we cannot redescribe these changes as persisting things. If I jump into some river and save your life, this act is not remotely like some rock or other unchanging persisting thing. If there were no events, because nothing ever happened, the (p.732) Universe would have no history, nor could we exist. We live lives, and each life is a series of events.

In living our lives, and thinking about what happens, we must also think about many possible events. When we are deciding what to do, we must choose between different possible acts, often by considering the possible outcomes of these acts. And there are other ways in which we should try to have true beliefs about many possible events. Without such beliefs, for example, we could not give causal explanations, since such explanations appeal to facts about what would have happened, if things had been in some ways different.

Actualists cannot defensibly deny that there are acts, and other events. To defend Actualism, however, these people must deny that there are any merely possible events. When we are deciding what to do, Possibilists like me believe,

  1. (F) there are, in the wide sense, different possible acts between which we choose. Since only one of these acts will be actual, the other acts are merely possible.

Actualists must claim:

  1. (G) There are only actual acts. It could not be in any sense true that there are some acts that are merely possible.

Since Actualists must reject (F), they must give a different account of what is involved when we decide what to do. Some Actualists claim

  1. (H) There actually exist the possibilities that we shall act in any of several ways, and we choose between these possibilities.

These Actualists might say that, unlike (F), (H) does not assert or imply that there is anything that is merely possible.

There is, I believe, no such difference between (F) and (H). Though claims that there actually exist these different possibilities, these Actualists must admit that only one of these possibilities will be actualized, in the sense of being what actually happens. When we are deciding what to do, we choose which of these possibilities will be what actually happens. The other possibilities will not actually happen, (p.733) but will remain mere possibilities. Though (F) claims that there are, in the wide sense, some acts that are merely possible, and (H) claims instead that there actually exist these unactualized possibilities, these are different ways of stating the same fact. If those who claim to be Actualists accept (H), these people are not, I shall argue, really Actualists.

Some other Actualists claim

  1. (I) There actually exist several ways in which we might act, and we choose in which of these ways we shall act.

Similar remarks apply. Of these ways in which we might act, only one will be the actual way in which we act. The unactualized ways in which we might act do not relevantly differ from what Possibilists call merely possible acts.

Other Actualists deny that there exist such abstract entities as possibilities or ways of acting. When these people describe what is involved in our making some decision, some of them claim

  1. (J) It might be true that we shall act in any of several ways, and we choose which of these things will be true.

Others claim

  1. (K) We think thoughts about the different ways in which we could act, and we choose which of these actual thoughts will guide our actual future act.

Similar remarks apply. Like (H) and (I), these claims imply that, when we are deciding what to do, we have different possible alternatives. When we choose which alternative will be actual, the other alternatives will be merely possible.

Actualists might reply that, when they say that

  1. (L) we could act in several different ways,

this claim does not imply that

  1. (M) there are such entities as different possible alternatives.

(p.734) But there is no relevant difference between (L) and (M). We do not state a different metaphysical view merely by using the verb ‘could’ and the adverb ‘in different ways’, rather than the adjective and noun ‘possible alternatives’. These alternatives are the different ways in which we could act. If Actualists make any of claims (H) to (L), they cannot defensibly deny that

  1. (N) there are, in the wide sense, such merely possible alternatives.

Since these people cannot deny (N), they should become Possibilists, who believe that

  • (O) there are, in the wide sense, some things that are merely possible.

It is irrelevant whether we can describe such cases without explicitly referring to such possible alternatives. As Church pointed out, misogynists might adequately describe the world without claiming that there any women, but that would not show that there are no women. We are asking whether, in our thoughts about our lives and other features of the world, it is enough to think only about what actually happens or will happen. And that is not enough. To make good decisions, or understand what causes what, we should try to form true beliefs about what might happen, or what would have happened. If there was no sense in which there are such merely possible events, we could not form such true beliefs.

Some Actualists would reply that we can form true beliefs about some things that don’t exist. One such belief is

  1. (P) Pegasus, the winged horse, doesn’t exist.

In believing (P), we need not be believing that there exists a winged horse, Pegasus, that doesn’t exist. To avoid the appearance of selfcontradiction, we could restate (P) as

  1. (Q) There is no such winged horse.

In the same way, these Actualists would say, we can have some true beliefs about merely possible events, even though there are no such events.

(p.735) This reply overlooks the difference between negative and positive beliefs. When we believe that certain things do not exist, this belief could be true even though there are no such things. But for us to have true positive beliefs about certain things, there must be such things. We could not truly believe that

  1. (R) some horses run faster than others

unless there are some horses. I have claimed that

  1. (S) when we are deciding what to do, we should try to form true positive beliefs about some events that are merely possible, such as beliefs about the possible outcomes of different possible acts, most of which will not be actual.

We could not form such true beliefs unless there are, in the wide sense, such merely possible events.

When we claim that there are such events, we do not mean that such events actually occur. As Possibilists, we distinguish between what is actual and what is merely possible. When we are in great pain, for example, our painful conscious state is actual and real. This pain is very different from the merely possible pain that someone who is not in pain might now be in. This great difference can make Actualism seem undeniably true. There may seem to be no sense in which there could be pain or suffering that is merely possible, rather than actual and real.

There is, I am arguing, such a sense. We ought, when we can, to prevent suffering. But we can prevent suffering only if there is a sense in which there is some possible suffering that we are preventing. That is why it could not be true that we have prevented the suffering of some rock. But though there is a sense in which there can be possible suffering that we prevent, this sense is very different from the thicker actualist sense in which there is actual suffering that we fail to prevent. That is why we should try to prevent suffering. Only actual suffering matters.

Suppose next that I let you die, though there was something that I could have done which would have saved your life. We can claim that

  1. (T) though there was this possible act, this act did not exist in the thicker sense of being actual.

(p.736) When we make such claims, we need not mean that, since determinism is false, some other act would have been causally possible. It is enough that I would have saved your life if I had chosen to act in this way. For Actualists to reject our view, they must claim that

  1. (U) it is never in any sense true that there was something else that we could have done, or any possible suffering that we could prevent, or anything else that might happen, but does not in fact happen.

This claim, I believe, is clearly false.

Of those who accept Actualism, most assume the Single Sense View. These people believe that, when we say that there are certain things, we must mean that these things actually exist. If that were true, Possibilists could not coherently claim that

  1. (V) there are some events that are merely possible, and never actually exist or occur.

As I have now argued, since this claim uses ‘there are’ in the wide sense and uses ‘exist’ in the actualist sense, (V) is coherent, true, and not in any way metaphysically misleading.

Of those who once defended Actualism, some would now reject (U) and accept (V). Though many people still claim to be Actualists, most of these people, I shall argue, misdescribe their real view. When such people cease to be Actualists, they could revise some of their arguments so that these arguments support a partly similar view. According to this view, which I call

Actualist Foundationalism: Though there are some things that are merely possible, and never actually exist, all truths about what is possible are in some way grounded on truths about what is actual.

It is of great importance whether this view is true. Many truths about what is possible are grounded, as these people claim, on truths about what is actual. This view could not, I believe, cover all such truths, but I shall not defend this belief here.

(p.737) 3

We can now turn from events to persisting things, which raise some different questions. In some cases, Possibilists could claim

  1. (A) There is a possible person who would become actual if a certain actual ovum and actual sperm cell were united and successfully implanted in some woman’s womb.

Actualists might reject (A), claiming instead

  1. (B) There actually exists a pair of reproductive cells which, if united, would later become an actual person.

Possibilists could reply that, though this pair of cells actually exists, this person does not actually exist, and that, if these cells are never united, this person will never actually exist. It would then be true, I believe, that there was a merely possible person.

Though such cases support Possibilism, they also support Actualist Foundationalism. This truth about this possible person is grounded on truths about these actual reproductive cells. Such cases are, in this respect, unusual, and misleading. Some writers suggest that

  1. (C) a possible person is something that is possibly a person.

In cases of the kind described by (A), we might claim that there is indeed such a thing. This pair of actually existing cells, we might say, is possibly a person. But many claims about possible people should not take this form. Suppose that, as members of some community, we are choosing between two energy policies which will have significant effects in the further future. We might then truly claim that

  1. (D) there are many possible future people whose well-being might be seriously affected by our choice between these policies.

But we should not, I believe, claim that there are or will be some actually existing entities which are possibly these people, or which (p.738) might become these people. Rather than claiming (C), we should claim that

  1. (E) a possible person is a person who might be, or become, actual.

Return to Jane, my imagined 14-year-old girl who intends to have a child. We might truly claim that

  1. (F) if Jane has a child now, she would give to this particular child a worse start in life than she could later give to any of the children whom she might have if she waited for several years before having children.

This claim is about a very large number of possible particular people, who are the many children some of whom Jane might have in the next ten or twenty years. We should not regard Jane’s possible children as actually existing entities that are not people but that might be people.

Such claims should take a simpler form. We should claim that

  1. (G) there are many possible children whom Jane might have,

in the sense that,

  1. (H) of the possible events that might occur, many would involve Jane’s having some particular child.

There are these possible people in the sense that there might later actually exist one or more of these people. If it is true that

  1. (I) Jane might have a certain child,

it is true that

  1. (J) there is this possible child whom Jane might have.

These are two ways of stating the same fact. When we claim that there are many possible children whom Jane might have, or that there are many possible future people whose well-being might be affected by our choice between two energy policies, we mean that there might in the future actually exist such people.

(p.739) Of those who claim to be Actualists, some would accept (I). Plantinga, for example, claims that, as an Actualist, he believes that ‘there are no things that do not exist’. But he also claims that

  1. (K) there could exist things that do not actually exist.

Suppose that some woman, Sarah, is being treated for infertility, and doctors have obtained from Sarah and her husband an actual ovum and sperm cell. Plantinga would then accept that

  1. (L) there could exist the child whom Sarah would have if this pair of cells were united and successfully implanted in Sarah’s womb.

What Plantinga rejects is only the claim that

  1. (M) there is this possible child whom Sarah could have.

Someone might now object:

There is no real disagreement here. If we would all agree that this child could exist, it is unimportant whether we claim that there is this possible child. And if we would all agree that we could act in different ways, it is unimportant whether we claim that there are different possible acts between which we must choose.

There is, indeed, no real disagreement here. But that is because, though Plantinga claims to be an Actualist, that is not really true. We can first return to Plantinga’s view about states of affairs. Possibilists like me claim

  1. (N) There are some possible states of affairs that are never actual.

Plantinga claims

  • (O) There actually exist some possible states of affairs that are never actual.

These claims do not state significantly different views. Like Possibilists, Plantinga claims that there are some states of affairs that are merely possible, since they are never actual.

(p.740) Plantinga might reply that he is defending Actualism. Unlike Possibilists, Plantinga claims that these merely possible states of affairs actually exist. But since we can use the word ‘actually’ in a sense that does not change the meaning of an assertion, Possibilists could restate (N) as

  1. (P) There actually are, in the wide sense, some possible states of affairs that are never actual.

As this restatement helps to show, (N) and (O) are not relevantly different claims. And Plantinga’s use of ‘actually exist’ may, as I have said, be misleading. If Plantinga claimed that one actually existing state of affairs is that the USA has declared war on China, we might take him to mean that the USA has actually declared war on China. Since Plantinga accepts that there are some states of affairs that are possible but are never actual, he is really a Possibilist about such states of affairs. We cannot defend Actualism by saying that such merely possible states of affairs actually exist.

We can now return to possible people. On Plantinga’s view, the word ‘actual’ has different meanings when applied to states of affairs and persisting things. While states of affairs can be actual in the sense that they obtain, persisting things can be actual in the different sense that they exist. Plantinga therefore claims that, though we should believe that

  1. (Q) there actually exist some states of affairs that are never actual in the sense that these states never obtain,

we cannot coherently believe that

  1. (R) there actually exist some persisting things that are never actual in the sense that these things never exist.

There cannot be such merely possible persisting things, Plantinga assumes, because claims like (R) imply that there ‘is a thing such that there is no such thing’. I have argued that, since Plantinga accepts (Q), he is really a Possibilist about states of affairs. But Plantinga might reply that, since he rejects (R), he is an Actualist about persisting things.

(p.741) This hybrid view is not, I believe, defensible. In the case described above, Plantinga would accept that

  1. (S) if some pair of actually existing cells were united and successfully implanted in Sarah’s womb, a certain child would be conceived and come into existence.

Suppose next that Sarah will never actually have this child. Plantinga would then accept that

  1. (T) there is a possible state of affairs in which this child would exist, but this state of affairs will never be actual, in the sense that it will never obtain.

This claim does not relevantly differ from the Possibilist claim that

  1. (U) there is a possible child whom Sarah might have, but this child will never actually exist.

These claims are both about some possible child, and tell us that this child will never actually exist. We cannot defensibly accept (T) but reject (U).

Plantinga would reject (U) because he believes that such claims involve a contradiction. The words ‘there are’ and ‘exist’, he assumes, both have only the same single sense. If that were true, and we use the word ‘actually’ in the sense that adds nothing, we could restate (U) as

  1. (V) There actually exists a possible child whom Sarah might have, but in the same sense of ‘actually exist’ this child will never actually exist.

This claim would indeed be a contradiction. As I have argued, however, we should reject the Single Sense View. If we use my definitions, and add the word ‘actually’, (U) could be more fully stated as

  1. (W) There actually is, in the wide sense, a possible child whom Sarah might have, but this child will never be actual, by existing in the actualist sense.

(p.742) Such claims are coherent, and in this imagined case (U) and (W) would be true.

Of those who claim to be Actualists, some would reject Plantinga’s view that there exist such abstract entities as merely possible states of affairs. If they considered Sarah’s possible child, many of these people would use one of the following phrases:

  • It might be true that such a child will exist,
  • There might exist such a child,
  • Such a child might exist,
  • Possibly: Such a child will exist,
  • It is possible that such a child will exist,
  • There could be such a child,

adding:

But there will actually be no such child.

But these are not ways of avoiding the Possibilist claim that there are some things that are merely possible. These are merely other ways of stating the fact that

  1. (U) there is a possible child whom Sarah might have, but this child will never actually exist.

The difference is only that, instead of using the non-modal verb ‘is’ and the modal adjective ‘possible’, these other claims use the modal verb ‘might’, or the modal adverb ‘possibly’, or the modal phrase ‘It is possible that’. These claims would all be about the possible event in which Sarah has this child, and would tell us that this possible event will not occur, so that this possible child will not actually exist. To defend Actualism, we would have to defend the claim that

  1. (W) it is in no sense true that there is such a possible event, and such a possible child whom Sarah might have.

And (W) would be clearly false. There would be, in the wide sense, both such a possible event and such a possible child.

There are other ways in which, in our abstract thinking, we can be misled by such grammatical differences. When we discuss normative (p.743) reasons, for example, we can say that certain facts are reasons to act in some way. This way of talking treats these facts as having the property of being reasons. We can also say that these facts give us reasons to act in this way, thereby treating reasons as entities which are distinct from the reason-giving facts. We can also say that these facts count in favour of acting in this way, thereby treating reasons as activities, or as what facts do when they count in favour of some act. But we need not ask whether reasons really are properties, or entities, or activities. These are merely different ways of making the same claims. Just as it makes no difference whether we say that certain facts are reasons, or give us reasons, or count in favour of some act, it makes no difference whether we say that it might be true that certain people will exist, or that there are these possible people. These are two ways of stating the same Possibilist view.

Though we should often try to be more precise, and draw new distinctions, we should also try to avoid distinctions which are merely linguistic. We should not think, like the English speaker: ‘The French call it a couteau, and the Germans call it a messer, but we call it a knife, which is, after all, what it really is’. Nor should we think: ‘When others say that there is some possible child whom Sarah might have, we say that there could be such a child, which is what is really true.’ We should not mistake these differences in wording for differences in meaning, and differences in the beliefs that these different words express.

4

I have defended Possibilism for several reasons. First, if we are Actualists, that may lead us to fail to recognize, or to deny, some important truths.

I have claimed that

  1. (A) when we are deciding what to do, we should try to form true beliefs about our different possible acts, and their possible effects.

Though Actualists deny that there are any such merely possible acts and other events, (A) is so obviously true that Actualists are unlikely to be (p.744) led astray. Things are different, however, when we turn from possible events to possible people. Scanlon writes:

… the beings whom it is possible to wrong are all those who do, have, or will actually exist.

Many other writers make such claims. These claims may suggest that

  1. (B) our acts cannot be wrong unless there is or will be, at some time, some actual person whom we have wronged, and to whom we owed it not to act in this way.

And many people have believed that

  1. (C) we cannot be acting wrongly if we know that there will never be any actual person whom our act will affect for the worse, or whose rights our act would violate.

These claims are, I believe, mistaken. As I have argued:

  1. (D) When we are making certain choices that will have effects in the further future, such as choices between two energy policies, we should consider the possible effects of our different choices, not only on actual future people, but also on the many possible people who, if we had acted differently, would have later existed.

And we should believe that

  1. (E) our choice of one of two policies may be wrong, because it will greatly lower the future quality of people’s lives, even though we know that, because our choice will affect who it is who later lives, this choice will never be worse for any actual future person.

If people in every generation chose such policies, the quality of future lives would steadily decline. The world would be slowly wrecked.

To recognize that, in choosing such policies, we have acted wrongly, we must consider the ways in which, if we had acted differently, our acts would have affected some people who never actually exist, but were (p.745) merely possible. It will be easier to ignore such facts if we are Actualists, since we shall then believe that

  1. (F) there is no sense in which there are any such merely possible people.

If there was no sense in which there are such people, we couldn’t think about them, since such thoughts would be about nothing.

Possibilism, I have also claimed, is the thin end of a wider wedge. As the arguments for Possibilism help to show, we should reject the Single Sense View. And if we believe that there are, in the wide sense, some merely possible entities and events, we should believe that there are entities of many other kinds. Some examples are:

words, meanings, philosophical theories, nations, human needs, overdrafts, symphonies, courage, fictional characters, literary styles, problems, explanations, numbers, logical truths, duties, and reasons.

Since such entities are abstract, they do not exist in the narrow actualist sense as concrete parts of the spatio-temporal world. But unlike entities that are merely possible, some of these abstract entities can be claimed to be actual in another, wider sense. There are, for example, many actual words, with actual meanings, and many actual philosophical theories, nations, and symphonies.

When people cease to be Actualists, they might turn to another view. According to what we can call

Alethic Realism: There cannot be anything that is not part of reality. Nor can any claim be true unless there is some part of reality to which this claim corresponds, and which makes this claim true.

Alethic Realists can believe that there are some abstract entities of the kinds just mentioned. These entities are created by us, or depend on the activities of human beings. By using language, we make it true that there are certain actual words, with actual meanings. And there are some (p.746) actual symphonies, theories, and nations because some people have composed these symphonies, developed these theories, or lived together in certain ways. These facts are enough to make these abstract entities part of reality. Alethic Realists can also claim that, since we know how we create these entities, and what their existence involves, these entities are not metaphysically mysterious. There is nothing puzzling in the existence of these words, meanings, symphonies, and nations.

Alethic Realism can also be applied to some entities and events that are merely possible. It may seem that, since these entities and events never actually exist or occur, they cannot be in any sense part of reality. But that is not so. What actually happens depends in part on what might have happened. If I lose some game of chess, for example, by failing to make some move that would have won this game, my mistake was a part of reality. Since my actual move was a mistake only because I could have made a different, winning move, the fact that such a move was possible can also be claimed to be a fact about reality.

Compared with Actualism, Alethic Realism covers more of the truth. But we ought, I believe, to reject this view. In cases of the kind just mentioned, facts about what is possible depend on facts about what is actual. I could have made some winning move only because the rules of chess allowed such a move, given the actual position of the chess pieces on the board. Similar remarks apply when there is some merely possible person who would have existed if some actual ovum and sperm cell had been successfully united. But there are countless other ways in which things could have gone differently, so that different people and other entities would have existed, and these possibilities cannot all depend on facts about what is actual. There are also countless other more remote possibilities. Reality might have contained entities of very different kinds, and the laws of nature might have been very different. These facts about how reality might have been cannot all depend upon, or be made true by, facts about how reality actually is.

Nor can Alethic Realism be defensibly applied to some other kinds of abstract entity, and to some necessary truths, such as certain logical, mathematical and normative truths. One example is

  1. (G) there are prime numbers greater than 100.

(p.747) Though we created the phrase ‘prime number’, and the meaning of this phrase, we did not create prime numbers, nor did we make (G) true. Even if we had never existed, there would have been prime numbers greater than 100. Similar remarks apply to:

  1. (H) No proposition can be both wholly true and wholly false.

  2. (I) If P implies Q, and P is true, Q must be true.

  3. (J) If we know both that P implies Q, and that P is true, we have decisive reasons to believe Q.

  4. (K) We have reasons to prevent or relieve the suffering of any conscious being, if we can.

We did not create these truths, nor does their truth in any way depend on us.

Since we did not create these necessary truths, or make them true, these truths raise some deep and difficult questions. But these truths are not, I believe, metaphysically mysterious. When we claim that there are such truths, we can use the phrase ‘there are’ not only in the wide sense, but also in a narrow non-ontological sense. On the view that I believe we should accept, which I call

Non-Metaphysical Cognitivism, these necessary truths are not made to be true by there being some part of reality to which these truths correspond. Since any truth can be said to be really true, there is a trivial sense in which these truths can be said to be about reality. But these truths are not about metaphysical reality, since they do not imply that certain things exist in some ontological sense.

This form of Cognitivism cannot conflict with what Russell calls our ‘robust sense of reality’, since these claims are not about metaphysical reality. When some view has no metaphysical implications, it cannot be open to metaphysical objections.

Alethic Realists may object that the words ‘there are’ and ‘exist’ cannot be used in any such relevant non-ontological sense. If we say that there are certain truths, but we deny that these truths exist in (p.748) any ontological sense, our claim may seem to be a contradiction. In considering this objection, it may help to compare Alethic Realism with two other similar views. According to

Spatialism: Nothing can exist that is not in space.

On this view, there can’t be any thing that isn’t anywhere. But that is not so. Though the Eroica Symphony was composed in Vienna, and has been performed in many places, this symphony itself, as an abstract entity, couldn’t be anywhere. Nor could many other abstract entities, such as the meanings of our words, philosophical theories, jokes, overdrafts, or the rhyme scheme of a Petrarchan sonnet. These entities could not exist anywhere in space. But there are such entities, since there are, in the wide sense, many actual words, meanings, symphonies, and many abstract entities of many other kinds.

According to

Temporalism: Nothing can exist that is not in time.

As Temporalists could point out, though most abstract entities could not exist anywhere in space, some of these entities do exist only at certain times. Symphonies exist only after they have been composed, and before there were any language users, there were no words, or meanings. Other abstract entities, Temporalists might claim, exist at all times. If we use a language with tensed verbs, such as English, we may assume that Temporalism must be true. Temporalists might say that, when we claim that there are prime numbers greater than 100, we must mean that there are now such prime numbers, though we can add that these numbers always have existed, and always will.

We ought, I believe, to reject this view. We cannot defensibly claim that, though many abstract entities could not exist in space, all such entities must exist in time. When we claim that there are prime numbers greater than 100, we should use ‘there are’ in a tenseless and timeless sense. Mathematical claims are not about what is always true. We should not, for example, claim that we know some facts about the future, because we know that there will always be prime numbers, and that two plus two will always equal four.

(p.749) Return now to Alethic Realism. I have claimed that

  1. (L) there are, in the wide sense, certain necessary truths, such as those stated by (G) to (K). Though such truths are about certain abstract entities and properties, they have no ontological implications. These truths are not about metaphysical reality.

Alethic Realists might object:

  1. (M) When you claim that there are these necessary truths, that must be a claim about reality. To be is to be real.

If they are Nominalists, these Realists might claim:

  1. (N) Since such truths and abstract entities could not exist in space or time, there cannot be any such truths.

Some Platonists would reply:

  • (O) These truths and entities exist in some part of reality that is not in space or time.

When they make such claims, these Platonist Realists are, I believe, too close to their Nominalist opponents. These truths and entities don’t have to exist in any part of reality, not even a special non-spatio-temporal Platonic realm. These truths are real only in the trivial sense that they are really true. And, since these truths are necessary, they do not have to be made true by there being some part of reality to which they correspond. This dependence goes the other way. It is reality that must correspond to these truths.