A TENSELESS ACCOUNT OF THE PRESENCE OF EXPERIENCE
By
Joshua M. Mozersky
Department of Philosophy
Queen’s University
Kingston, OntarioK7L 3N6
Canada
Tel.: (613) 533-2182
Fax: (513) 533-6545
E-mail:
Abstract
Tenseless theories of time entail that the only temporal properties exemplified by events are earlier than, simultaneous with, and later than. Such an account seems to conflict with our common experience of time, which suggests that the present moment is ontologically unique and that time flows. Some have argued that only a tensed account of time, one in which past, present and future are objective properties, can do justice to our experience. Any theory that claims that the world is different from how we experience it must nonetheless be consistent with the having of that experience. Accordingly, in this essay I defend the tenseless theory by arguing that it can indeed account for certain key features of our experience of time without recourse to tensed properties.
1
A TENSELESS ACCOUNT OF THE PRESENCE OF EXPERIENCE
1. Introduction
In what follows I examine the prospect of defending a tenseless or ‘B-series’ account of time from a particular challenge. The tenseless theory I wish to defend is austere. It entails that the only temporal properties exemplified by events are the relationsearlier than, later than and simultaneous with.[1] In particular, tensed or ‘A-series’ terms such as ‘past’, ‘present’ and ‘future’[2]either fail to refer or in fact function to ascribe some combination of tenseless relations.[3] There are, in other words, no such monadic properties as being past, present or future. Claims such as ‘1945 is past’ or ‘2045 is future’can be true, but their truth conditions must be spelled outonly in tenseless terms (for how this can be done see Mozersky 2001).
The challenge I wish to consider is that manifested byconscious experience. Two elements of such experience appear to be relevant. First, the presentis experientially privileged in that we are only ever capable of experiencing that which occurs in the present. To put this observation another way, though we may know all week that the movie and Friday, 1:00 p.m. are simultaneous, when we learn that Friday, 1:00 p.m.is present and, therefore, that the movie starts now we seem to learn a new fact. Accordingly, tenseless relations cannot be all there is to time and the best explanation of the presence of experience is that the present is ontologically privileged, more real than other times.
Secondly, as we interact with the world it appears as if time, in some non-metaphorical sense, passes; what was future becomes present and then past. However, since tenseless relations are permanent—if ‘x is earlier than (or later than or simultaneous with)y’ is true, then it is always true—the B-series lacks the conceptual resources to describe either the ontological uniqueness of the present ortemporal passage. For if it is never false that x is earlier than (later than or simultaneous with) y, how can there be time at which either x or y fails to exist?[4] Furthermore, if temporal relations are temporally invariant, how can it be that any one time, t, changes from being future to present to past?
So the question becomes, can the tenseless theory of time account for the obvious facts of human temporal experience? It does no good to deny that we have such experiences. Their occurrence is, as Richard Taylor puts it, a Datum (Taylor1992). Rather, a metaphysical theory must be able to explain why it is we have the experiences we do as theyare the starting pointsfor investigation into the world.[5]
In this paper I focus on the presence of experience. I wish to present a tenseless explanation of why it is that the present is experientially privileged and why we seem to learn a new fact when we learn that an event is present or now,given that all times stand on an ontological par. I shall do so via an analysis of present tense beliefs and utterances. I believe that the account offered below does in fact provide some clues as to why we experience events sequentially and take differential attitudes towards the past and future. However, a B-theoretic account of the experience of passage deserves separate treatment and I shall bypass it here. Nonetheless, I concludethat tenseless theories of time can be defended against one important experiential challenge.
2. The problem
How are we to construe the truth conditions of tensed sentences such as ‘2003 is present’ or ‘it is now 2003’ and the thoughts they express? Simple disquotation will not do:
(0)‘2003 is present’ is true iff 2003 is present.
For one thing, the context of evaluation may differ from the context of utterance rendering (0) false even if the original sentence token is true. But more germanely, the B-theorist must reject this on independent grounds: it ontologically commits one to the property of presentness.[6] If there are no tensed properties, then the specification of the truth conditions of temporal sentences must refer only to tenseless relations. The natural suggestion is to employ ‘simultaneous with’ in place of ‘present’ (or ‘now’), ‘earlier than’ in place of ‘past’ and ‘later than’ in place of ‘future’ (see Smart 1963) and to index utterancesto a time. This would lead to the following:
(0*)‘2003 is present’ said at t is true iff 2003 is simultaneous with t.
This suggestion can lean on weighty authority. Recent work on indexicals inspired by Kaplan (1989) and Perry (1979)[7]has lent significant credence to the idea that temporal indexicals, such as ‘now’ or ‘present’ refer to their time of utterance directly, not via a mediating sense. Consequently, all that ‘present’ contributes to the proposition expressed by ‘2003 is present’ is a referent, namely that time at which the sentence is uttered (or thought).
So far so good, but this story apparently conflicts with Perry’s well-known problem of the essential indexical. For imagine, to change the example slightly, a 2003 utterance of ‘It is now 2003’ or more simply ‘Now is 2003’. We have the following schemata for present tense sentences:
(PT1)For all t: ‘Now is t’ said at t* is true iff t = t*
(PT2)For all t: ‘e is now’ uttered or thought at t is true iff e and t are simultaneous
which entail that the truth conditions of ‘now is 2003’, uttered in 2003 will be the tautological:
(0**)2003 is(simultaneous with) 2003.
If this is correct, then the tensed ‘Now is 2003’ said in 2003 is logically equivalent to the tenseless‘2003 is 2003’.
This seems counterintuitive for suppose I desire to perform a certain action, A,in 2003. Suppose, further, than my actions are caused by my beliefs and desires. My desire to A in 2003 combined with my belief that ‘2003 is 2003’ is true cannot prompt me to act at the right time for this latter belief can be held truthfully at all times; it cannot prompt me to move at one time rather than any other. It seems, however, that my desire to A in 2003 combined with the belief that ‘2003 is now’ is true[8] will move me to act. So, the A-theorist may insist, how can it be that ‘now is 2003’ said in 2003 is equivalent to ‘2003 is 2003’? (PT1/2) must be missing something for tensed and tenseless sentencesobviously differ in cognitive significance.
The complaint is put as follows in a recent defence of tensed time:
My wedding anniversary is March 12. Suppose that I have memorized this date. Maybe I had it inscribed on my wedding ring. So I know the following: My wedding anniversary is March 12.
Now suppose I am in my office late in the afternoon one day next March. I may say to myself: “My fifth anniversary is March 12. I should think about buying my wife an anniversary present.” I might then wonder how much time I have. I take out a calendar and find out today’s date and discover to my horror that it is March 12! I shout “My fifth anniversary is today!”
In this little episode it is clear that I had two distinct utterances:
(1)
My fifth anniversary is March 12.
(2)
My fifth anniversary is today.
It is also clear that when I uttered (2) I had knowledge that I did not have when I uttered (1), and this extra knowledge appears to be reflected in the difference between (1) and (2). Thus, it is arguable that (1) and (2) have different semantic contents. (Ludlow 1999, p. 6)
If the tenseless account of tensed utterances outlined above is correct, then (2) uttered on March 12 has the same semantic content[9]as (1) uttered anytime. But how can standing in relation to one and the same propositional content on two different occasions not only prompt very different responses but do so in a way that enables timely action?
It will help to lay out Ludlow’s reasoning in steps:
- According to the B-series view of time, the truth conditions of ‘e is at t’ believed[10]anytime (Act I) are the same as the truth conditions of ‘e is today/now’ believed at t (Act II).
- So, the two acts of belief are relations to the same content.
- But the two acts of believing lead to very different actions, utterances, subsequent beliefs, emotions, concerns, etc.
- It is therefore implausible to assume the two mental states involve relations to the same content.
- Hence, #1 is false. The best explanation of the noted differences is that in Act I one is related to a tenselesscontent while in Act II one is related to a tensedcontent.
- So, the B-theory is false; there exist tensed truth conditions.
The first thing to notice here is that Ludlow’s argument only goes through on the assumption that the content of uttered sentences and the thoughts they express istransparent, i.e. apprehended directly and not via some sort of mediating entity. In support of thisthe A-theorist would naturally turn to Frege (1918), who identifies propositions with the meanings of that-clauses found in intentional contexts and belief with a with apprehension of propositions. This helps explain why
(3)Nancy believes that Los Angeles is in California.
can be true, while
(4)Nancy believes that America’s largest city is in California.
is false. Since the meanings of ‘Los Angeles’ and ‘America’s largest city’ differ they correspond to different propositions and Nancy need not be related to both on any given occasion.
Similarly,
(5)Peter believes that e is att.
may be false while
(6)Peter believes that e is now.
is true because different propositions are believed in each case. From here the A-theorist seems well positioned to argue that the truth conditions of (6) must quantify overproperties that (5) does not as the relationship between meaning and truth conditions is certainly close.
However, this is far from decisive as thereare clear cases where different Fregean propositions have identical truth conditions. The sentences ‘I am the happiest person in the room’ and ‘You are the happiest person in the room’ differ in meaning, and so on Frege’s accountwould correspond to different propositions. However, if the first is believed by John, and the second by Nancy referring to John, the two beliefs do not differ in semantic content. Similarly, a difference in meaning between ‘e is now’ and ‘e is att’ does not suffice to show that the truth conditions of any two tokens of these utterances differ. But the latter is precisely what the A-theorist needs if the cognitive features of tensed sentences are to have ontological bearing. As Perry (1977) has pointed out, the Fregean view has some difficulty explaining indexical belief in general.[11]
These considerationssuggest two points. First, there is reason to reject or upgrade the Fregean account of the belief. Secondly, the idea that the truth conditions of belief are directly apprehended is rendered dubious precisely by the case of indexicals. So, a satisfactory account of temporal belief seems destined to posit something intermediate between a believer and the content of his or her beliefs.[12] Fortunately, there is a stock of items that seems ideally suited for this role: the sentences of natural language.
3. Indexicals, linguistic meaning and semantics
One of the functions of sentences is to tell ourselves and others how things stand. At least in the case of indexicals, language serves to mediate between believers and truth conditions because how truth conditions arepresented matters. So belief is a relation between a believer and a propositional content, but a two-part relation: one between a believer and a sentence, the other between that sentence and its truth conditions. In other words, to believe that P one accepts as true a sentence, ‘Q’. If done correctly, ‘Q’ is true iff P. But indexical examples teach us that it might also be the case that ‘R’ is true iff P, and this difference can be significant.
Let me pausebefore continuing. I am not asserting that in order to believe that P one must explicitly think (utter, say) ‘Oh, “Q” is true’. On the contrary, one may believe that P without adopting a posture of semantic assent. However, some kind of representational item is necessary in order to have a thought with the content that P. This itemmight be a sentence in the language of thought, or even a picture (mental or otherwise). However, even if one thinks at the object level that, for example, protons are heavier than photons, one must have in mind something, call it ‘Q’, with the grammatical and semantic properties such that it is true iff protons are heavier than photons. This is why non-linguistic creatures can’t have beliefs about atomic physics (or many other things, such as the difference between the A-series and the B-series). However, so long ‘Q’ is true iff protons are heavier than photons, then it is harmless to equate the belief that protons are heavier than photons with the acceptance of ‘Q’ as true.
Let us see how all this applies to Ludlow’s example:
(1)My fifth anniversary is March 12
(2)My fifth anniversary is today [believed on March 12].
(1) is a tenseless sentence so its truth conditions are uncontroversial. Letting e= Ludlow’s Anniversary and t = March 12, (1) is true iffe and t are simultaneous; for simplicity I will shorten this to Sim(e, t). But, according to the B-theory I defend here, (2) has the same truth conditions because‘now’ refersdirectly to its time of utterance/thought,[13]so that(2)thought or uttered on March 12 is true iff Sim(e, t), as illustrated:
Act I:
‘e is at t’
Accepts as trueExpresses (is made true by)
S at any timeSim(e, t)
Act II:
‘e is now’
Accepts as trueExpresses (is made true by)
S at tSim(e, t)
Fig. 1: two ways to relate tothe same content
Kaplan provides some clues as to why the two acts lead to different actions:
It is the manner of presentation, the character and not the thought apprehended, that is tied to human action. When you and I have beliefs under the common character of ‘A bear is about to attack me,’ we behave similarly … Different thoughts apprehended, same character, same behaviour. When you and I both apprehend that I am about to be attacked by a bear, we behave differently … Same thought apprehended, different character, different behaviours. (Kaplan 1989, p. 532).
When you and I both think ‘A bear is about to attack me’ is true, these two acts of belief have different truth conditions yet we are both prompted to run. The reason, in part, is that we have grammatical knowledge concerning the indexical, ‘me’, namely that the referent of ‘me’ is the speaker/believer. So, if you find yourself believing something about ‘me’, then you believe something about yourself. Similarly, when you find yourself believing something about ‘you’, a different set of behaviours is appropriate since in any context of utterance/belief the referent of ‘you’ is not the speaker/believer.
Now recall (PT2):
(PT2)For all t: ‘e is now’ uttered or thought at t is true iff e and t are simultaneous
This plus the transitivity of simultaneity tells one that any act that is simultaneous with e must occur at the same time as the thought that ‘e is now’ is true. However, if we consider ‘e is at t’, the time of utterance or thought is semantically irrelevant:
(PT3)For all t, t*: ‘e is at t’ uttered/thought at t* is true iff e and t are simultaneous.
Therefore, one does not commit oneself to believing that anything simultaneous with e must also be simultaneous with any act of accepting ‘e is at t’ as true and, knowing this, one sees that ‘e is at t’ cannot be cognitively useful in the way that indexical sentences can be, which is just as Perry pointed out (Perry 1979). In other words, if I want to A when e occurs (for example enter a theatre when a movie starts) I must refrain fromA-ing until I accept that ‘e is now’ is true and begin moving the moment I do accept this. However, knowing (PT3), I know not to let any propositional attitudemediated by ‘e is at t’ influence my decision to act. This is why one will typically behave differently depending on which sentence is used to express a given content. So, relating oneself to a tenseless content via a tensed sentence can prompt one to behave and believe in ways that won’t arise with a tenseless sentence. Thus far, there appears to be no needto resort to tensed properties.
4. Indexicals, perception and causation
I have argued thus far that the cognitive significance of present tense beliefsis the result of, first, our knowledge of the linguistic rules—(PT1), (PT2), etc.—that govern temporal indexicals and, secondly, the employment of logico-semantic beliefs, such as the transitivity of simultaneity, that enable us to time our actions appropriately(more on background beliefs in section 6 below). But notice that these beliefs are all temporally stable: if true they can be truthfully believed at all times. Accordingly, it is not clear how thislinguistic cum behavioural knowledge can promptone to form a belief or to act at one time rather than any other. William Lane Craig writes:
On a B-theoretic ontology, all times are equally real. So why does it is 4:30 at 4:30 present itself as ‘now’ to someone and only to someone at 4:30? Why is the person tenselessly existing at 4:30 thus distinguished from his fellows? The obvious answer is that ‘is presented’ is not tenseless, but tensed, and there is a uniquely present moment, other times being . . . unreal. Because 4:30 really is present, the cognizer comes to believe that ‘It is now 4:30’, not just a tenseless proposition (Craig 1996, p. 257).
Present tense beliefs concerning events seem subject to a similar worry: why does e is simultaneous with t, a relation that holds at all times,present itself as ‘now’ at and only at t? What is required is a tenseless account of how one is triggered to form present tense beliefs at just the right times.