The main objection to this Presentist account of change involves an appeal to Truthmaker Theory (2.1T/2.1A.1T/2.1A.1A.1T). If we accept some form of Truthmaker Theory, then we have to ask what makes past- and future-tensed truths true. We will consider several arguments of this form in Chapter 21.
20.5 Arguments for Tensism
In the remainder of this chapter, we consider six arguments for Tensism.
20.5.1 The “Thank Goodness” Problem
We'll start with Arthur Prior's “Thank Goodness That's Over” argument (Prior 1959). Prior was a pioneer in the formal study of time and tense. Prior (1959) points out that we have very different attitudes towards events in the near future and the near past. If we know that we must undergo some painful surgery, then, upon learning that the surgery is over, we naturally express relief: “Thank goodness that's over!” Prior thought that this fact could be turned into an argument for Tensism.
In order to reconstruct Prior's argument, we need the notion of a Priorean proposition:
Def D20.3 Priorean Proposition. x is a Priorean proposition if and only if (i) x can be expressed by a sentence employed in some context, (ii) x is a fundamental bearer of truth and falsity, and it can be true or false simpliciter (absolutely, without qualification), and (iii) x can be the object of belief or other intentional attitudes (such as hope, fear, relief, etc.).
A proposition is a tensed proposition if it necessarily has its truth or falsity temporarily, while an untensed proposition necessarily has its truth or falsity permanently. A proposition that predicates a tensed property of something is a tensed proposition. Prior's argument can be taken as a proof that there are tensed propositions. If tensed propositions exist, then Tensism is true, since the truth of a tensed proposition is an A-property.
Let's try to reconstruct Prior's argument for the truth of Tensism:
Mental attitudes, like relief or dread, are relations between persons and Priorean propositions. (Premise)
Priorean propositions are truth-bearers: they are true or false simpliciter. (Def 20.3)
When a person is relieved that an experience is just over, there is no untensed proposition that could be the object of that attitude (and similarly for other temporal attitudes). (Premise)
Hence, mental attitudes, like relief or dread, are relations between persons and tensed propositions. (From 1 and 3)
Tensed propositions exist and are true or false simpliciter. (From 4 and 2)
Tensed propositions change their truth-values over time: they are sometimes true and sometimes false. (Definition of tensed propositions)
Hence, Tensism is true. (From 5 and 6)
There is now almost universal consensus that premise 3 is true. Premise 2 is a stipulative definition, and premise 6 seems to follow from the fact that A-truths have truth only temporarily. Hence, if the argument fails, it must fail in the very first step: mental attitudes don't take Priorean propositions as their objects.
We shouldn't overlook the fact that there is a lot to be said for premise 1. As David Lewis points out in “Attitudes De Dicto and De Se” (Lewis 1979a), there is a great theoretical advantage to assigning the objects of the attitudes to a uniform ontological category. Beliefs, desires, and other attitudes interact in ways that reflect the logical relations between their objects, and it is propositions that most naturally and generally possess the required network of logical relations. In addition, in the case of belief, the pull toward thinking of the object as something true or false simpliciter is, as we shall see, very strong.
OLD ANTI-TENSIST RESPONSES How Anti-Tensers respond to Prior's argument depends on whether they embrace Old or New Anti-Tensism. Old Anti-Tensers have two options. First, they can deny premise 1, claiming that untensed propositions can be adequate objects of relief and dread. Second, they can acknowledge that these attitudes do have tensed propositions as objects but deny that any of these tensed propositions are ever true. This second option involves an error theory about our judgments about the truth-values of tensed propositions.
Nathan Oaklander (1994) has taken the second option and defended an error theory about A-beliefs and A-attitudes. In many ways this is an elegant solution, accepting Prior's claim that only tensed propositions are adequate objects of many attitudes. Oaklander accepts the first three premises of the Priorean argument, disagreeing instead with premise 6: that tensed propositions have varying truth-values. Instead, Oaklander claims that all tensed propositions are necessarily false, since the required A-facts are necessarily absent. Such an approach is not unprecedented. There are error theories of morality (J.L. Mackie 1977), mathematics (Hartry Field 1980), and color judgments (Adam Pautz 2014), among others. It's a proposal that is well worth having on the table, but, in our view, should be considered only as a last resort.
What's wrong with untensed propositions as objects of attitudes? As Prior (1959) pointed out, the two obvious candidates are hopelessly inadequate in the case of the Thank Goodness attitude. (28) and (29) are applications of these two candidates to a particular case:
(28) Thank goodness that the pain ended on February 7, 2007 at noon CST.
(29) Thank goodness that the pain ended just before the occurrence of this token.
As Prior remarks, it is inconceivable that one would thank goodness for either of these facts, nor do we have any explanation of why we would be inclined to feel thankfulness toward these B-facts at some times rather than at others, since we might be aware of these facts long before or long after the indicated time has passed.
In his 2006 dissertation, Max Goss proposed that we each have changing now concepts that pick out for us the present moment via a description that ascribes to that moment the total contents of our immediate experience at that moment. These concepts are, according to Goss, “n-senses” for particular times. Goss's account of a thank goodness attitude would go something like what is represented by (30):
(30) Thank goodness that the pain ended just before I had a total experience state of type N.
Goss argued that, although it is clear that this is a contingent rather than a necessary fact, it is nonetheless quite plausible to suppose that each moment of time during which one is conscious is associated for each person with a unique total experience type, so the various definite descriptions formed in this way will succeed in picking out a unique time. Moreover, it is plausible to suppose that there is a kind of contingent inaccessibility of the experience type for one moment at any other moment. Only when one is in an experience state is one in a position to be able to refer to that experience state in thought. At other times, the state is too complex and precise to be reproduced by the limited resources of one's general concepts or imaginative memory. Thus, Goss claims that his account can explain the de facto inaccessibility of these temporal thoughts at any time but the appropriate one.
However, even if it is a fact that one can think (30) only at a time while one is experiencing a state of type N, the complement of (30) still fails to capture what it is that one is grateful for. The only reason that one cares that the pain ended before one experiences a state of a certain type is that one knows that one is experiencing that type (uniquely) in the present. The problem is that there is only a contingent, accidental association between Goss's n-senses and their referents. Consequently, we cannot suppose that the n-sense delivers the right intentional element to the object of belief. Hence, Goss's analysis hasn't really avoided the need for a tensed proposition as content.
new anti-tensist responses: DE SE attitudes
To deal with the thank goodness problem, many New Anti-Tensers have invoked the idea of de se attitudes. The phrase ‘de se’ is Latin for ‘about or pertaining to oneself’. A de se attitude is a belief or other intentional relation directed towards oneself as oneself. If, for example, one is conceited, then one has a very favorable opinion of oneself. To be conceited, the opinion can't just happen to be about oneself. Suppose RCK thought that the tallest philosopher in the department was very smart,
thinking that Professor Juhl is the tallest member, while in fact it is the case that RCK is the tallest philosopher in the department. Such an opinion doesn't count as making RCK conceited, since he is not attributing intelligence to himself as himself, but rather as the tallest philosopher in the department.
It is also possible to have an attitude that is about oneself de re (‘about the thing’) that happens to be oneself but not about oneself de se. De re attitudes have Millian or Russellian propositions as their objects. A Millian (Russellian) proposition is a proposition that contains some concrete thing as a part or component and not just some concept or identifying property of that thing. The Millian proposition ‘Mt. Blanc is snowy’ is an entity that contains Mt. Blanc itself, “with all its snow fields”, as Gottlob Frege put it when objecting to such propositions to Bertrand Russell. Fregean propositions, in contrast, are wholly abstract. The corresponding Fregean proposition about Mt. Blanc contains some uniquely identifying concept or property of Mt. Blanc, not the actual mountain. Millian philosophers, including Bertrand Russell, have argued that Millian propositions are needed as the object of de re attitudes, including beliefs, desires, and other attitudes that are directly about some thing with which we are acquainted. Suppose, for example, RCK catches sight of a very handsome man in what he take to be a window, and RCK thinks that he is quite handsome (where the ‘he’ is directed de re to the thing RCK is seeing). If it turns out that RCK is really seeing a reflection of himself in a mirror, the opinion expressed is not conceited, since it is about a thing that happens to RCK himself, but it is not about RCK as himself. The reality of de se attitudes towards oneself provides substantial support for Anti-Tensers' claim that our attitudes take something other than propositions as their objects. When one thinks I am tired, one seems to be either in a special de se attitude toward a property (the property of being tired) or toward an incomplete proposition (‘x is tired, or ___ is tired’) or else one is thinking about the proposition that RCK is tired in a special, self-reflexive, de se way.
Anti-Tensers can plausibly argue that we can also have de se beliefs about time. There is a difference between believing that the war is now over and believing that the war is over in April 1865, even if “now” is April 1865. Sentences using tense and temporal adverbs like ‘now’ are de se sentences about the present moment as the present moment, not in the sense that the present moment has some special property, but only in the sense that the person thinking and speaking is wholly located in that time. It's as if the time in question were thinking of itself through the thinker of the now-thought.
One way out for New Anti-Tensers is the one taken by John Perry (1979) in accounting for the essential indexical as the object of a de se attitude. Perry argues that a belief state must be characterized both by its object, which is an untensed proposition, and by the way in which that untensed proposition is grasped. This “way” is an associated “character”, which is a function which determines a unique proposition given a particular context. The notion of the character of a thought or proposition was introduced by David Kaplan in his seminal work “Demonstratives” (Kaplan 1989). To apply this to Prior's example, my thankfulness stands in a relation to one of the untensed propositions (28) and (29) above, but in addition it is related to the character corresponding to the sentence, ‘The pain ended just before now’.
Alternatively, Anti-Tensers could instead suppose that some mental attitudes, like dread and relief, take objects other than complete untensed propositions. Here are some possibilities that have been proposed:
(a) Propositional functions or Kaplanian characters: functions from times to untensed propositions (Kaplan 1989, Theodore Sider 2001: 19–21)
(b) Properties (David K. Lewis 1986a: 28–30, 54–55; Roderick Chisholm 1989, 129–138)
(c) Sets of centered worlds (Lewis1979a)
A final option, briefly mentioned by Perry (1979), is still more radical. It involves giving up any uniformity to the objects of belief and the other attitudes. Instead of a single attitude of belief, there would a large family of belief-like attitudes, each taking a different ensemble of entities as its object. There would be a believing relation that took pairs of particulars and properties as its objects, another that took particulars, properties, and times, and so on. There would be a family of now-believings: if RCK now-believes
We will not pursue this most radical option. It seems to introduce an unacceptable degree of complexity and heterogeneity within our theory of the attitudes. This already runs afoul Ockham's Razor (PMeth 1). Further, in order to get many obvious inferences to come out as correct, we would have to postulate a large number of self-evident axioms relating the various versions of belief and the other attitudes to one another. In fact, it is not at all obvious that, once begun, it would be possible to place any limit at all on the complexity introduced. This seems like a blatant violation of the First Corollary of Ockham's Razor (PMeth 1.1).
Thus, the most plausible Anti-Tenser solution takes belief and other attitudes have something other than propositions as their objects. Philosophers have often assumed that it is one and the same kind of thing—namely, propositions—that play two crucial roles. First, propositions are the fundamental bearers of truth and falsity. Second, they are the object of belief and other attitudes. Anti-Tensers who adopt this solution must give up this assumption. Propositions are the bearers of truth or falsity, but what we believe belongs to a different category. Mental attitudes are relations to something like functions from times and persons (the ‘now’ and the ‘me’ of the thinker's world) to propositions. Hence, we cannot ask whether what a person believes is true or false simpliciter. We can only ask whether what the person believes yields a true proposition when applied to the thinker and the time of the thought.
THANK GOODNESS AS AN INFERENCE TO THE BEST EXPLANATION Prior's Thank Goodness observation could be used to support Tensism in an entirely different way, as part of an inference to the best explanation. Tensers can offer a plausible explanation for the rationality of our relief at the recent passage into the past of something painful and of the rationality of our joy at the imminent arrival of something pleasant from the anticipated future. Anti-Tensers have no comparable explanation.
Of course, Anti-Tensers can always posit that it is simply, as a matter of brute fact, rational to take differing attitudes toward the past, present, and future. However, this additional postulation counts against Anti-Tensism by way of Ockham's Razor (PMeth 1). Tensers have a simpler, less ad hoc explanation. The near future has the potential of being (fully) real, while the near past does not.
Tensers' explanation of the rationality of the thank goodness reaction appeals to the following principle of practical rationality:
Value of Movement. It is a bad thing for a bad thing to be approaching us; it is a good thing for a bad thing to be moving away from us.
The explanation must also appeal to the Vehicle Thesis:
Vehicle Thesis. As the present moves through time, those of us who continue to exist move with the present.
The Vehicle Thesis explains the thank goodness phenomenon, and it clearly entails the truth of Tensism. Does it favor one type of Tensism over the others? Simple Tensism seems inconsistent with the Vehicle Thesis. As the spotlight of actuality moves forward, it would not seem to be carrying anything forward with it. Presentism, however, is the version of Tensism that is most obviously consistent with the Vehicle Thesis. In fact, Presentism might even entail the Vehicle Thesis. If so, then Presentism gains credibility by virtue of its greater explanatory power.
20.5.2 Evidence against indexicality: Parallels between ‘now’ and ‘actually’
As we have seen, the strongest version of Anti-Tensism is one that takes A-thoughts to be indexical in character, as something like functions from the th
inker and the moment of thought to some untensed proposition. Anti-Tensers rely heavily on relatively uncontroversial examples of such indexicality, like sentences involving ‘I’ and ‘here’, to provide grounds for thinking that tense is just another indexical element. Unless one is a Solipsist, one doesn't think that the irreducibility of the first-person pronoun is any evidence for the claim that one is uniquely real or existent. Consider the egocentric, de se proposition expressed by (31) and its ordinary de re counterparts, (32) and (33):
(31) I am now leaking sugar from my grocery cart.
(32) John Perry is now leaking sugar from his grocery cart.
(33) That man is now leaking sugar from his grocery cart.
Clearly, there is a difference in significance between (31) and (32), even when both are being thought by John Perry. Perry may not realize that he is John Perry (or the John Perry who figures in proposition (31)). Maybe he is suffering from amnesia, for example. Similarly, there is a difference in significance between (31) and (33), even when (33) is being thought by John Perry and ‘that man’ picks out John Perry himself. Perry might see himself in a mirror without realizing that it is he himself whom he is seeing.
Anti-Tensers use these facts to argue that we have no reason to believe in special egocentric or ‘I’ facts in order to ground the truth of (31). (31) is made true by the same thing (whatever it is) that makes (32) and (33) true.
Anti-Tensers can then argue that, like egocentric propositions, tensed propositions give us no reason for positing metaphysically real or fundamental tensed truths. Instead, we should take tensed truths as indexical, as made true by whatever makes true the corresponding untensed proposition, as determined by the time of utterance or thought. An indexical sentence like ‘It is now raining’ is simply a function from moments of usage to untensed propositions. The same is true of past- and future-tensed sentences. The tensed sentence ‘It will be raining’ is a function that, when applied to any given moment of time, yields an untensed proposition stating that some time later than that time is one at which rain occurs.
The Atlas of Reality Page 72
