Cheryl E. Fitzgerald

 

ÒNowÓ is Not an Indexical

            If we imagine that everyone on the planet all at once utters the word ÒIÓ, we know that each and every single token of ÒIÓ has a unique referent.[1]  And if everyone utters the word ÒHereÓ, we know that each and every single token of ÒhereÓ has a unique referent.  But if everyone on the planet all at once utters the word ÒNowÓ, each and every single token of ÒnowÓ has the very same referent.  What explains the difference?

            If there is a difference, and this difference has the result that ÒnowÓ is not an indexical, then the tenseless theory of time must be false, for it requires that ÒnowÓ be indexical.  This would render the tensed theory of time more plausible, for it does not require that ÒnowÓ is indexical.  Quentin Smith wishes to establish, in his book Language and Time,[2] that ÒnowÓ ascribes a real A-property of presentness to the moment at which it is uttered.  Smith does refer to ÒnowÓ as an A-indexical, but I will argue that is not correct to call ÒnowÓ an indexical of any sort, for it lacks certain features that other established indexicals, such as ÒhereÓ and ÒIÓ, possess.  I will show that this is both an implication of his theory and a benefit to it.  But if any tenser wishes to convince any detenser that ÒnowÓ is not indexical, the discussion must take place on the level of language, and not within the scope of a discussion about the nature of time; otherwise, both sides of the debate run the risk of merely talking passed each other.  It is my purpose in this paper to show at the level of language that it is not possible for ÒnowÓ to be indexical, for it lacks the required features that make a word indexical.  The features in question are only those having to do with the rules of usage in the language. 

            Indexicals are frequently defined as words whose reference is context-sensitive.  This characterization is vague at best; at worst, it is entirely unhelpful in distinguishing true indexicals from non-indexicals.  Thus, an important task for this paper is focusing on the features of words that are taken as unproblematically indexical, and specifically determining which of those features are essential features of indexicality.  I take two words to be mostly unproblematically indexical, ÒIÓ and ÒhereÓ.

I.  The Avoidance of Question-Begging Premises

             Utterances of the indexicals ÒIÓ and ÒhereÓ are dependent for their reference upon the conditions of the speaker; utterances of ÒnowÓ are not.  A detenser would retort, ÒOf course they are!  They are dependent upon his location in time.Ó  If the detenser holds that ÒnowÓ is an indexical, this belief must be a conclusion of his theory and not a premise, for whether or not ÒnowÓ is a true indexical really is the debate between the tensed and tenseless theories of time.  It is not open to the detenser to explain the supposed indexicality of ÒnowÓ by stating that its reference is dependent upon the time it is uttered, for doing so would be question begging in favor of his theory.  The references of utterances of ÒIÓ and ÒhereÓ are dependent upon something very specific about the nature of the speaker, either who or where the speaker is, and thus it is the speaker that plays the role of that to which the reference of the word is relative.  On the other hand, there is nothing about the nature of the speaker at all that relativizes the reference of utterances of ÒnowÓ.  It is not the case that the relativizing factor is the time of utterance; in order for there to be a relativizing factor, something about what occurs must be relative.  Utterances of ÒIÓ and ÒhereÓ are clearly relative to their speakers for their references.  But utterances of ÒnowÓ are clearly not relative to their speakers, as I demonstrated above in the imagined scenario that everyone utters ÒNowÓ all together. 

In order for the tenseless theory of time to hold merit, the difference between how utterances of ÒIÓ and ÒhereÓ refer and how utterances of ÒnowÓ refer requires an explanation that is not question-begging.  But no such explanation can be provided.  In order to show the difference between the nature of ÒnowÓ and that of the true indexical ÒhereÓ, it is necessary to give a full explication of what about ÒhereÓ makes it indexical, and then show that ÒnowÓ lacks those features. 

II.  Establishing the Indexicality of ÒHereÓ

            In each instance[3] that I utter the word ÒhereÓ, the reference is fixed by my location, my own location since I am the speaker, and the word refers to my location whether I intend it to or not.  This is a function of the word ÒhereÓ, a rule of its usage.  The only aspect of the reference that is under the control of my intention is how wide the scope of ÒhereÓ is, literally, how large of a spatial area it covers, so long as that area includes myself within its boundaries.  Thus, for example, you might ask me if I received a package you sent, and I might respond, ÒYes, itÕs here, but I left it in my office.Ó  So long as the intended scope of my use of ÒhereÓ in that sentence is wide enough, if I meant the city I live in, for example, then I would not have said something contradictory. 

            At this point, the only difference between ÒhereÓ and ÒnowÓ is what has already been stated, that the reference of ÒhereÓ depends on where the speaker is, while the reference of ÒnowÓ does not.  This is certainly not enough to convince any detenser that there is a significant difference between the two that makes the latter non-indexical. 

            The difference is revealed by considering counterfactual scenarios in which utterances of ÒhereÓ and ÒnowÓ occur.  Imagine that you are brought to an empty theater and told that you may choose from all possible seats where you would like to sit.  After careful consideration, you choose a seat and announce, ÒI will sit here.Ó  There are at least as many counterfactual situations as there are seats other than the one you actually chose.  In each of those counterfactual situations, your utterance of ÒhereÓ refers to a different chair.  Since each of the other chairs that you did not choose was equally a possible choice, then your utterance of ÒhereÓ could have referred to any of the other chairs in the theater.  The reference of your particular use of ÒhereÓ is relative to where youÕve chosen to place yourself.

When the reference of a word or phrase is relative to something, it is a requirement that it could have referred to another thing had it been relative to something else.  When considering counterfactual situations, we must consider those that are closest to the actual, and those that are closest are those in which utterances are made at precisely the same time as they actually are.  Thus, if we specify that you actually utter the sentence at 5:37 pm on April 29, 2006, then in each of the counterfactual situations under consideration, you utter the sentence at the time in the counterfactual world that is the equivalent of the time referred to by Ò5:37 pm on April 29, 2006Ó in the actual world.  This allows for the possibility that across counterfactual worlds, the same object is referred to by different words or phrases.  It is certainly possible that I might have been baptized with a different name than I in fact have, so that in some possible world, I am referred to by something other than ÒCheryl Elizabeth FitzgeraldÓ.  But in order that we may still talk about me in that counterfactual world, we will have to refer to the counterfactual me by using the name I actually bear along with the qualifier Òin the actual worldÓ.  Essentially, all that matters for counterfactual considerations is the nature of the objects, not the names attached to them.  Therefore, in each counterfactual world, all other elements of the scenario being identical to how they actually are, and the only variable factor is the chair you choose, each ÒhereÓ across counterfactual worlds refers to a different chair.  It is only in this sense that we can truthfully state that your utterance of ÒhereÓ could have referred to a different chair.

III.  ÒNowÓ is Not Similar to ÒHereÓ

Smith specifically assumes the absolutist theory of time, and thus takes ÒnowÓ to be identifying the moment in which the utterance occurs as a member of the set of events that occupy the moment.[4]  The moment is a metaphysically distinct thing from the set of events that occupy it, but might not have occupied it.  Now consider that instead of saying, ÒI will sit here,Ó you utter the sentence, ÒI have now chosen where I want to sit.Ó  Is it possible that your utterance of ÒnowÓ could have referred to a different moment of time other than it actually did?  But a detenser might accuse me of now begging the question in my own favor, so I need to justify why the time of the utterance cannot be a variable factor across counterfactual situations.  First, in regards to the reference of ÒhereÓ, whether a change in time is allowed or not, each chair in the theater is an equally possible choice.  Since no variation in time is a stricter condition for closeness to the actual world, then that is the condition we should apply when considering counterfactual scenarios that are closest to the actual, unless there is no counterfactual scenario that does not include a variation of time, i.e., unless one is considering counterfactuals about variations in time.  Utilizing the stricter condition of no variation in time across counterfactual worlds, I have shown that your utterance of ÒhereÓ is truly relative in the sense that it could have referred to a different chair, all else being equal.  And it is this particular feature of the relativity of its reference that contributes to the indexicality of ÒhereÓ.  Thus, we hold ÒnowÓ to the same strict condition for determining whether it shares this feature with ÒhereÓ that would make it indexical.  And clearly, it cannot: inevitably, the reference of ÒnowÓ in each counterfactual scenario under consideration is identical.  At the time it is uttered, there are no other possible alternative references it could have had.  Thus, the reference of an utterance of ÒnowÓ is not relative in the same way that an utterance of ÒhereÓ is relative that makes ÒhereÓ an indexical.[5] 

IV.  A Further Difference from ÒHereÓ: the Non-Relativity of ÒNowÓ

            1.  Relevance of the Type-Token Distinction

But, the detenser would object again, utterances of ÒnowÓ are indeed relative, namely, to the time they are uttered.  Even if they were, this would not be the same kind of relativity that ÒhereÓ bears that makes it indexical; this difference is significant enough to demonstrate that ÒnowÓ does not share the features that ÒhereÓ possesses that provide its indexicality.  I said, ÒEven ifÉ,Ó for it is not correct to say that utterances of ÒnowÓ are relative to the time they are uttered.  Utterances are tokens of sentences-types, and a sentence-token is an event that necessarily includes the time it is uttered as part of its nature, that is, its time of utterance is something we would necessarily include when determining what it means, what it refers to, and whether or not it is true.  Furthermore, we specifically distinguish between two sentence-tokens as individual and distinct because they are uttered at different times and/or by different individuals.    A token of ÒDescartes is deadÓ is true iff, in the very least, DescartesÕ death has the B-relation of being simultaneous with the sentence-token.  If the sentence is uttered when Descartes is alive, clearly, it is false.  But if the sentence is uttered in 2006, clearly, it is true. 

But the word ÒsentenceÓ in both of the previous statements is ambiguous, for it may mean sentence-type or sentence-token, and the difference is crucial.  If we mean sentence-type, then the truth-value is relative to when it is uttered, since it is a tensed sentence, and in this sense only can we say its truth-value varies over time.  But if we mean sentence-token, then the truth-value is not relative to when it is uttered, for being a token is to be uttered (or thought, or inscribed, etc.); its truth-value when it is uttered is determinate and cannot change, for being a token is being an event at a time, which is a feature that cannot change over time.  To speak of the same sentence as being spoken at different times is to speak of two separate sentence-tokens of the same sentence-type.  It is no mystery that tensed sentences obtain their truth-values, and so their references, from their being tokened; if we desire to talk about the nature of tensed sentences, we must be referring to their sentence-tokens.[6]

The thrust of my argument at this point hinges on my claim that the time of the tokening is an essential feature of a sentence-token.  That this must be the case can be deduced from the very simple way in which we define and treat and distinguish sentence-tokens.  A token is, by definition, an event, whether it is an utterance, an inscription, or a thought.  As an event, it occurs in time and cannot be removed abstractly from that time, just in the way that it does not make sense to remove a soccer game abstractly from time; the nature of a soccer game requires time, and so the event of a tokening requires time.  But most importantly, we distinguish between two separate and individual tokens of the same sentence-type by the fact that they are tokened at different times, or by different individuals even if at the same time.  For the latter, it is irrelevant at the moment, but I will address it and what significance it does have.  Considering only tokens uttered by the same person, of two utterances of exactly the same sentence-type, we specifically claim that so long as they occur at different times, even when separated by only seconds, then they are two, separate, distinct tokens and, thus, not identical.  If we distinguish between different sentence-tokens by this method when considering only tokens in the actual world, then we ought to hold tokens in possible worlds to the same standard.  It does not follow that one could not possibly have uttered the same thing at a different time, it only follows that we are required to recognize that Òto have uttered the same thing at a different timeÓ means Òto have uttered a different sentence-token of the same sentence-type at a different time.Ó 

Because tensed sentences obtain their references from their tokenings, analysis of tensed sentences must take place on the level of sentence-tokens.  It follows that an utterance of a tensed sentence-type at some time in the actual world is not identical to an utterance of that same tensed sentence-type at a different time, however slight, in some counterfactual world.

It is still up to me to discuss the relevance and significance of the fact that two individuals can utter the same sentence-type at the same time, but we distinguish their utterances as two, separate sentence-tokens.  As demonstrated at the very opening paragraph, when more than one person utters ÒnowÓ at precisely the same moment, both of their utterances refer to that moment in the same way.  The conditions of the speakers, who they are, where they are located, the context of the semantic meaning of their utterances, etc., bears no difference on the reference of ÒnowÓ.  All other possible factors may be varied, but so long as their utterances occur in the same moment, the reference of ÒnowÓ in each utterance is identical.  This means that the reference of ÒnowÓ has nothing to do with the speaker in any way, is not relative to the speaker, but rather has something to do with the world.  Since this is the case, then ÒnowÓ does not behave like any other indexical word, for unproblematically indexical words depend on some condition of the speaker for their precise reference.  If it is the case that utterances of ÒnowÓ really depend for their reference upon some state of the world, and have nothing to do with the speaker, then at best, we can say that the reference of ÒnowÓ is relative to the world, which states no more information than the statement, ÒThe laws of nature are relative to the universe.Ó

2.  Relevance of Presentness

There is a second reason why a token of ÒnowÓ cannot possibly refer to any other moment of time than it actually refers, but this reason follows as a consequence of SmithÕs theory of presentism.  This further discussion is therefore to be taken as a consideration and analysis of SmithÕs presentism in light of the current discussion, and not necessarily an argument for the non-indexicality of ÒnowÓ, which I think I have successfully established.  I wish to specifically demonstrate how it is that the non-indexicality of ÒnowÓ is not only compatible with SmithÕs theory, but follows directly from it.  And I do believe that this bears further, more significant consequences beyond the discussions of the nature of time.

In our scenario of the empty theater and your choice of where to sit, at the moment of your utterance, ÒI will sit here,Ó there are a number of possible chairs that ÒhereÓ can refer to, as has been already established.  But there are not more possibilities than there are chairs in the theater—imagining, for the sake of argument, that in this situation you cannot choose any location outside of the theater—for there do not exist any more chairs than there actually are at this time.  We can imagine that one year from now, the theater is expanded and more chairs are installed; but right now, those chairs are not possible choices for you because they do not exist yet, they are not present here before you to choose from. 

With that in mind, if we consider that Smith claims that ÒnowÓ refers to the moment of its utterance and ascribes presentness to it,[7] and does not refer to any other moment, thus, not ascribing presentness to any other time, then there is no other possible time to which your utterance of ÒnowÓ can possibly refer to.  Under SmithÕs theory, presentness is a real property, a property of whatever event or time one is specifically giving attention to, but it would be more accurate to really say that presentness is a property that belongs to a state of the whole world: we can give local attention to any event in our environment, or we can give attention to the moment in time, but we must not forget that during that moment there are several other events that possess the property of presentness, possibly infinitely many.  But the important point here is that there is only one moment in time that is present when you utter the word ÒnowÓ; in other words, there are no other possible alternative moments that your utterance of ÒnowÓ could refer to, for it is logically impossible for more than one moment of time to be present at the same time.  This presents a significant difference from the possibilities of chairs in the theater that are present before you to choose from. 

V.  Implications of the Non-Indexicality of ÒNowÓ

            If ÒnowÓ refers to a state of the whole world, of everything all at once, then ÒnowÓ is the most immediate access we can have to that whole of everything, the most readily available awareness that we can have of the whole of all that exists.  The belief that ÒnowÓ is merely an indexical prevents access to this awareness, and promotes a perspective that localizes oneÕs awareness and focus to the spatially immediate.  In other words, taking ÒnowÓ as indexical links it with ÒIÓ and ÒhereÓ, to form the localized perspective of ÒI, here, nowÓ.  But this would be like swimming in the Atlantic Ocean without being aware that it connects with every other ocean, combining into a unified body of water.  The awareness of the whole that one can access via the feeling of ÒnowÓ would provide an opening to the global awarenesses and global feelings that Smith discusses and describes in his book The Felt Meanings of the World.[8]  An awareness of the ÒnowÓ, the pure presentness of the whole of all that exists, is equivalent to the felt meaning of global fulfillment, via the feeling of global joy, for the existence of the world-whole that Smith discusses in Chapter IV of Felt Meanings. 

One concern that Smith very briefly addresses in this book is that most people are unable to find access to these global feelings.  An implication of the realization of the non-indexicality of ÒnowÓ might be the opening of a vein of access to those feelings.  However, this is possible only under the perspective of the reality of tense, and the privileged present.  Under the tenseless theory of time, there is no privileged access to the whole of all that exists through the presentness of ÒnowÓ, and each and every moment is but one small piece of the whole timeline, which is inaccessible to us because we, too, only exist as small pieces in that timeline.  But under the tensed theory of time, the existence of the whole is always, fully accessible by us through the feeling and awareness of ÒnowÓ. 



[1] There are uses of ÒIÓ in which the referent is not the speaker.  These uses are specifically intended by the speaker to refer to someone else in particular, and the reference is understood by others only because certain qualifications have been set up for that particular occasion.  Clearly, if I am repeating what someone else said, and that person used the word ÒIÓ in a sentence, then my use of it refers to that person, not myself, for I am quoting him or her.  But this is not an ordinary use of ÒIÓ, for if you walked into the room in the middle of a conversation and heard me utter a sentence that included the word ÒIÓ, you would undoubtedly immediately believe that I was referring to myself.  You might exclaim with a surprised, ÒYou whatÉ?!Ó  And it would make perfect sense for me to respond with, ÒNo no!  I was just repeating what X said earlier.Ó  When ÒIÓ is uttered unqualified by any other phrase, its default reference is the speaker.

            Another common but unordinary use of ÒIÓ might be a situation in which I am dressed up in a costume and I ask you to guess, ÒWho am I?Ó  Conversationally, this use of ÒIÓ is intended to refer to the person whom I am dressed up as: ÒI am Queen Elizabeth I.Ó  But it is more accurate to say that this use of ÒIÓ is an abbreviation for ÒI am dressed up asÉÓ or ÒI am pretending to beÉÓ, etc.  These expansions of this particular unordinary use of ÒIÓ need not be stated, for they are normally understood through the circumstances of familiar conversational implicature. 

[2] Smith, Quentin.  (1993).  Language and Time.  New York: Oxford University Press.

[3] I specifically mean ordinary instances, and not instances in which I am quoting another person, acting in a play, etc.  For all such unordinary uses of indexicals, it is clear that the referents of those indexicals will be abnormal.  Such abnormalities are easily explained by revealing the unstated qualifying expressions that tag alongside those utterances, but that is not the topic of this paper.

[4] Smith, 115.

[5] A similar demonstration to show that the indexicality of ÒIÓ is like the indexicality of ÒhereÓ should be conceivable without any difficulty. 

[6] All of this amounts to something very similar that Smith himself discusses in section 4.3, pp. 98-105.  Essentially, detensers have confused two different notions of the truth-value of tensed sentence-tokens, their tensed truth-values and their tenseless truth-values.  The tenseless truth-value of any sentence-token is merely the truth-value it has when it is tokened.  But the tensed truth-value of a sentence-token is whether or not it is true at some time, for example, whether it is now true, which amounts to whether it would be true if it were tokened now.  If someone now uttered the sentence, ÒDescartes is dead,Ó this token, call it D, is now true.  But had someone uttered it when Descartes was still alive, that token, call it D*, would have then been false.  Since D was uttered simultaneously with DescartesÕ death, then it is also tenselessly true.  And since D* was uttered earlier than DescartesÕ death, then it is also tenselessly false.  The mistake that detensers have made was in not understanding this distinction.  To state that D (is) true iff it occurs simultaneously with DescartesÕ death does not tell us whether D is now true or false, for even when Descartes is alive it is true that ÒD (is) true iff it occurs simultaneously with DescartesÕ death.Ó  It is clear that source of the confusion is not distinguishing properly between sentence-types and sentence-tokens, and what consequences follow when we are dealing specifically with sentence-tokens. 

[7] Smith, 119-120.

[8] Smith, Quentin.  (1986).  The Felt Meanings of the World.  West Lafayette, Indiana: Purdue University Press.