Time, Tense, and Eternal Recurrence

Sam Cowling

October 22, 2010

Abstract

Nietzsche’s discussion of eternal recurrence (hereafter, ER) raise arange of interpretive questions. In this discussion, I have two aims.The first is to consider the tenability of ER as a metaphysical thesis inlight of recent commentators’ objections to it. The second is to exam-ine how, if at all, we might develop a metaphysics of time and tensethat accommodates the central commitments of ER. After cataloguingthese central commitments, I survey a dilemma that arises in recon-ciling them with a plausible metaphysical picture. I then consider anumber of views about the metaphysics of time and tense, and defend aversion of dynamic eternalism that captures the central commitmentsof ER and avoids the dilemma posed earlier. I conclude by noting theinterpretive challenges for this metaphysics of ER.

1 Introduction

Nietzsche’s thesis of eternal recurrence (hereafter, ER) finds its most vividpresentation in the form of a thought-experiment:

What, if some day or night a demon were to steal after you intoyour loneliest loneliness and say to you: “This life as you now liveand have lived it, you will have to live once more and innumer-able times more; and there will be nothing new in it, but everypain and every joy and every thought and sight an everythingunutterably small or great in your life will have to return to you,

2

all in the same succession and sequence—even this spider andthis moonlight between the trees, and even this moment and Imyself. The eternal hourglass of existence is turned upside downagain and again, and you with it, speck of dust!”

Would you not throw yourself down and gnash your teeth andcurse the demon who spoke thus? Or have you once experience atremendous moment when you would have answered him: “Youare a god and never have I heard anything more divine.”1

Nietzsche accords ER enormous importance. In Ecce Homo, he describesthe thesis as “the hardest, most terrible insight into reality” and the “mostabysmal thought.”2 In The Gay Science, he accords it “the greatest weight”3

and urges us to “stamp the form of eternity upon our lives.” He concludesThe Twilight of the Idols by marking himself as the “teacher of the eternalrecurrence.” And, in The Will to Power describes the doctrine as “powerfulenough to work as a breeding agent: strengthening the strong, paralyzingand destructive for the world-weary.”4

In light of the privileged place Nietzsche gives ER, the many difficultissues it raises cannot be ignored. Here, we might distinguish four sorts ofissues. The first family of issues are interpretive. They concern the natureand role of ER in Nietzsche’s thought. Among these is perhaps the mostcommonly addressed question about ER: Does Nietzsche conceive of it asgenuine cosmological hypothesis or something like a heuristic for normativeguidance?5 The second family of issues is normative. They concern what,if any, normative implications ER has. Among these is the question ofwhat it would mean to live one’s life in light of the “abysmal thought”.6

The third family of issues is epistemological. They concern the peculiar1Gay Science [hereafter, GS] §341.2Ecce Homo [hereafter, EH] §3: TSZ-6.3GS §341.4Will to Power [hereafter, WP] §862.5The prevailing consensus would seem to be the latter. See Magnus (1978), Nehamas

(1985), Long (1987), and Clark (1990) for defenses of a non-cosmological interpretation.6Extended inquiries into these normative issues include Magnus (1978), Nehamas (1985),

Clark (1990), Reginster (2006), and Hatab (2005).

3

epistemic situation that ER gives rise to and the question of whether weare forced into skepticism about its truth.7 The final family of issues aremetaphysical. These are questions about what ER—understood as a thesisabout the nature of the world—would require the world to be like. Amongthem are the following: What is the relation between individuals in differentstages of recurrence? What must the passage of time be like in order toaccommodate recurrence?8

As should be evident, these issues cannot be neatly pried apart. Ques-tions about the possibility of free action in the face of eternal recurrence are,for example, partly normative and partly metaphysical. Furthermore, whensurveying the literature on ER, it is common to see arguments proceed fromnormative, epistemological, or metaphysical issues to interpretive conclu-sions.9 Despite these intersections, my discussion will be primarily focusedupon the metaphysical issues raised by ER. In particular, I have two aims:The first is to consider the tenability of ER as a metaphysical thesis in lightof recent commentators’ objections to it. The second is to examine how, if atall, we might develop a metaphysics of time and persistence that naturallyaccommodates the central commitments of Nietzsche’s ER. Here, I will pro-ceed by focusing on his remarks directly regarding ER. Only in concludingwill I address the much more difficult issue of how a metaphysics of ER canbe made consonant with Nietzsche’s broader philosophical commitments.

The implications of these projects remain considerable: If I am successfulin showing that arguments against the metaphysical coherence of ER fail, Iwill undermine a common interpretative strategy: concluding that ER mustbe a mere heuristic since it is an untenable metaphysical hypothesis. Inaddition, I will argue that the most natural interpretation of Nietzsche’s7For discussion of skepticism and ER, see Kain (1983) and Loeb (2006).8Arguably the most popular metaphysical issue regarding ER is the importance and

efficacy of Nietzsche unpublished statistical arguments for ER. I discuss these in Section

Four. For discussion, the locus classicus on this issue is Simmel (1907), who argues that

Nietzsche’s arguments are unsuccessful. See also Danto (1965), Soll (1973), Zuboff (1973),

and Magnus (1978).9For example, Magnus (1978) and Clark (1990) produce metaphysical arguments against

ER in defense of their preferred non-cosmological interpretations.

4

remarks regarding ER casts Nietzsche as a proponent of provocative andinteresting view regarding the “reality of tense”.

The discussion will run as follows: In Section One, I outline the centralcommitments of ER and survey a dilemma that arises in understanding themetaphysics of ER. I then consider a response to this dilemma and arguethat it is unsatisfactory. In Section Three, I distinguish a number of viewsabout the metaphysics of time and tense, and sketch a metaphysics of timethat both captures the central commitments of ER and avoids the dilemmaposed earlier. I then address some objections to this proposed metaphysics.In Section Four, I conclude by briefly noting the interpretive challengesthat this face-value account of ER faces once we consider his more generalcommitments.

2 The Case against Recurrence

In this section, I consider some features of ER as well as some metaphysicalobjections to it. These features will be introduced roughly at first and thenclarified as objections to ER are examined.

The first feature of ER is what I will call the perpetuity requirement.This requirement, alluded to in Nietzsche’s remark above, holds that recur-rence, however it is to be understood, extends infinitely “backwards” and“forwards”. As he remarks in The Will to Power, your life “has alreadyrepeated itself infinitely often and plays its game ad infinitum.”10 For ourpurposes, we can call each instance of recurrence an “epoch” and take theperpetuity requirement to be the following: any satisfactory metaphysics ofER must hold that, in addition to our own epoch, countless epochs precedeit and countless epochs follow it.11 So understood, neither our epoch nor anyother is the first or last of the series of recurrences. Although this perpetuityrequirement, characterized roughly, is an uncontroversial feature of ER, wewill soon see that is not obvious how the demands of infinite “backward”10WP §1066.11This terminology owes to Lewis (1986) who employs worlds of “eternal recurrence” to

examine a number of issues in the metaphysics of modality.

5

and “forward” repetition are best understood.The second feature of ER concerns the relation between the plurality

of epochs. It holds that there is some relation of “uniformity” betweenthem. According to the uniformity requirement, every epoch is “the samesuccession and sequence”. Intuitively, the uniformity requirement holds thatthere is no deviation in the events that make up each epoch. There are,therefore, no alterations to the course of history. Everything—down to themost minute details—repeats itself. And, while this uniformity is easy tograsp intuitively, many of principal metaphysical problems raised by ERconcern precisely how this uniformity is understood.

Perhaps the most natural way to understand the uniformity require-ment is as follows: every epoch is qualitatively indiscernible from everyother epoch. According to this qualitative conception of uniformity, epochsare related by perfect resemblance; each epoch is an exact duplicate of ev-ery other epoch. So, while epochs are alike in terms of their qualitativeproperties, they differ numerically from one another.

If the qualitative conception of uniformity is accepted, a familiar argu-ment against ER can be offered. This argument turns on the Principle ofthe Identity of Indiscernibles (hereafter, PII), according to which any ob-jects that share all their qualitative properties are identical.12 The argumentruns as follows: Given the qualitative conception of uniformity, epochs arequalitatively indiscernible, and, since PII requires qualitatively indiscernibleobjects to be identical, epochs are identical. But, if there is only a singleepoch, ER is false, since the perpetuity requirement cannot be satisfied.13

As an attempt to show ER to be metaphysically impossible, this argu-ment is not convincing. While some have argued for PII way of the Principleof Sufficient Reason or, more recently, the implausibility of a multitude of co-located qualitatively indiscernible objects, these arguments are by no means12The qualification that the properties be qualitative prevents the principle from being

trivial: if some qualitatively indiscernible objects, a and b, also shared their non-qualitative

properties, the logic of identity alone entails their identity, since both instantiate being

identical to a and being identical to b.13As I consider in Section Five, this last inference also turns on certain heavy metaphysical

assumptions.

6

decisive.14 In addition, arguments owing to Black (1953) and Adams (1979)suggest that there is very good reason to reject PII in order to accommodatecertain metaphysical possibilities. So, given the highly controversial statusof PII, those who take it to refute ER overestimate its dialectical force.15

Although the qualitative conception of uniformity is not refuted by mereappeal to PII, a number of commentators have argued that the qualitativeconception of uniformity is, for independent reasons, unsatisfactory. Thesource of the problem, as we’ll soon see, is that qualitative indiscernibilityis alleged to fall short of a further demand on ER: the concern requirement.

According to the concern requirement, you must be appropriately con-cerned for the welfare of “yourself” in other epochs. Put differently: ERmust account for how and why you care for “yourself” in other epochs, sincefailure to do so would leave ER without the apparent normative significanceNietzsche affords it. Admittedly, this requirement is far from precise. Inlarge measure, this is because it is an open question of how exactly ER is tohave its alleged normative implications. But, in any case, if ER is to havethe implications Nietzsche envisions, something like the concern requirementmust be satisfied.

There are a number of ways we might attempt to clarify the concernrequirement, but, in each case, complications proliferate. We might, forexample, point out that while we care about our own persistence into thefuture, we would not care in a comparable fashion about an individual that isdistinct yet qualitatively indiscernible from us. As should be clear, however,this presupposes that identity is “what matters” in a way that Parfit (1984)and others have notably challenged.

Perhaps a better analogy arises in considering a debate regarding modalmetaphysics. According to counterpart theory, individuals have their de remodal properties in virtue of standing in relations of qualitative resemblanceto various individuals.16 So, for example, Kit is possibly a goldfish if there14Leibniz is most famously associated with arguments for PII based on the Principle of

Sufficient Reason. See Della Rocca (2005) for discussion of arguments for PII based on

co-location.15For example, Magnus (1978) suggests that PII suffices to show that ER is untenable.16See Lewis (1986) for discussion.

7

is some individual that qualitatively resembles Kit to a suitable degree andinstantiates the property being a goldfish. Against a counterpart-theoretictreatment of de re modality, Kripke (1972) and others have argued that, inevaluating modal claims, we are not concerned with how someone very simi-lar to us is, but, instead, with how we ourselves could have been. Again, theworry is that the relevant measure of concern is undermined by the meta-physical theory in question, but, again, this is to presuppose that identityis “what matters”.

Regardless of the merits of the present example, it should be clear thatthe concern requirement is primarily intended to show that, in some way,individuals must be tied together across epochs by some connection of nor-mative significance. I leave it open, however, whether this must be identityor mere resemblance.

According to the following objections, the qualitative conception of uni-formity is incompatible with the concern requirement. The first objectionto the qualitative conception of uniformity owes to Soll (1973):

I want to suggest that only if the recurrences of one’s life werelinked to one’s present life as different states of consciousness areunited in one consciousness should one have any personal concernabout the prospect of one’s present pleasures or pains recurring.If I am concerned about the possibility of suffering pain laterin this life, it is because those future states of consciousness inwhich I might suffer pain are united with my present state of con-sciousness. And this unity depends, at least in part and in somecomplex manner, upon there being linkages of memory amongthese states of consciousness. Since I am not connected to myrecurrence in another cycle by being united with him in one con-sciousness through links of memory, I am not identical to him inthe sense necessary for me to view his suffering pain tantamountto my suffering further pain. Only by inappropriately construingthe suffering of some future recurrence on the model of sufferinglater in this life does the question of the eternal recurrence of

8

one’s pain weigh upon one with “the greatest stress”.17

This objection holds, in keeping with the concern requirement, that thenormative significance of ER is a function of our concern for “ourselves” inother epochs. But, if we are merely qualitatively indiscernible from “our-selves” in other epochs, this relation of indiscernibility does not provideadequate grounds for concern. As a result, something stronger than thequalitative conception of uniformity is needed for ER to satisfy the concernrequirement. To establish this, Soll argues that something like continuityof consciousness or perhaps numerical identity is required for the concernrequirement to be satisfied. But, given the numerical distinctness of indi-viduals across epochs as well as the lack of causal connections between theseindividuals, Soll holds that there is no way to ground the needed concernfor individuals across epochs.18

Viewed solely as argument to show that individuals cannot be “appropri-ately” concerned for individuals in other epochs, Soll’s argument is unsatis-factory. It simply assumes that either identity or continuity of consciousnessis what grounds the “appropriate” level of concern for individuals, but theseassumptions are highly controversial. One could, for example, accept theSimple View of personal identity and hold that, although there is no psy-chological or physical criterion for personal identity, individuals are, as amatter of brute fact, identical across epochs.19 Alternatively, if personalidentity were to be understood in terms of psychological dispositions likedesires, personality, or behaviour rather than continuity of consciousness,17Soll (1967: 340).18For an endorsement of Soll’s argument, see Clark (1990: 267): “[Soll’s] argument may

require slight modification because some will find it as reasonable to have a special concern

for the destiny and suffering of the person qualitatively identical to myself in the next cycle

as to have such concern for the person I would be after an operation or an accident that

removed all memory connection to my present self. But if there is not continuity of

consciousness in the latter case, there is a kind of continuity. It still makes sense to think

of my present life as in some (perhaps limited) kind of continuity with my life after the

memory loss. In the case of eternal recurrence, on the other hand, no continuity exists.

There is neither continuity of consciousness nor any kind of continuity or traceability

through time and space.”19See Partfit (1984: 209).

9

the common dispositions of individuals in different epochs could be taken toground their concern for one another.20

Viewed as problem for understanding Nietzsche’s conception of ER, Soll’spoint is more serious. Nietzsche’s remarks suggest that the connectionbetween individuals within various epochs is of unambiguous importancewhereas qualitatively indiscernibility seems to carry no obvious or immediatenormative importance. Furthermore, Nietzsche’s remarks seem to suggestthat it is not mere qualitative indiscernibility, but, instead, literal iden-tity that explains why the concern requirement is satisfied. The animals ofZarathustra stress this apparent identity of individuals across epochs in thefollowing passage:

Behold, we know what you teach: that all things recur eternally,and we ourselves, too; and that we have already existed an eter-nal number of times, and all things with us I myself belong tothe causes of the eternal recurrence. I come again with this sun,with this earth, with this eagle, with this serpentnot to a newlife or a better life or a similar life: I come back to this same,identical life.21

Given this passage and its focus upon one and the same individual com-ing back to “this same, identical life”, there is reason to believe that Ni-etzsche holds numerical identity rather than qualitative indiscernibility toground the concern for individuals across epochs.

A second objection to the qualitative conception of uniformity is putforward by Magnus (1978) and also suggested by Simmel (1920). This argu-ment proceeds as a thought-experiment: Suppose that, instead of eternallyrecurring epochs, the demon told you that there were instead an infinite ar-20Among the prevailing views in contemporary metaphysics, the denial of numerical iden-

tity as a prerequisite for the relevant kind of concern is commonplace. For example,

both stage theory and perdurantism—two of the leading contenders—reject the view that

our persistence across times requires the numerical identity of a single object located at

distinct times. See Sider (2001) for discussion.21Thus Spoke Zarathustra [hereafter, TSZ] III: 13

10

ray of simultaneously occurring epochs that are qualitatively indiscernible.Magnus (1978) puts the objection as follows:

One can imagine, let us suppose, a universe precisely like ours,a counterpart universe in which all objects, events, persons, andrelations are indiscernibly like our own. What are we to say of“the space” occupied by this duplicate universe, this counterpartworld? It must be the same as ours, must it not? Or considerthe “n” number of right hands which may be said to occupy“the same” space as the right hand with which I now write, butwhich my perceptive and apperceptive faculties are too gross todiscern. Is it necessarily contradictory to say in both cases—duplicate world or hands—that two or more identical objectsoccupy precisely the same space at the same time? If it is con-tradictory, and I believe it is, it is probaly also contradictory tosay that two or more “recurring” identical entities occupy thesame time (and space).22

This objection is well understood as reductio: If the qualitative conceptionof uniformity is correct, our concern for “ourselves” in other epochs is afunction of our qualitative indiscernibility. But, if there were myriad ver-sions of “ourselves” that exist simultaneously but were also qualitativelyindiscernible from us, we would have no concern for them, so the qualitativeconception of uniformity cannot satisfactorily explain how the concern re-quirement gets met. Without this concern, ER is not significant in the wayin which Nietzsche envisions, so we must reject the qualitative conceptionof uniformity.22Magnus (1978: 108-109). The same argument is considered in Clark (1990): ”Suppose

the demon, instead of announcing eternal recurrence, proclaims that there now exist an

infinite number of duplicates of our solar system and therefore an infinite number of in-

dividuals qualitatively identical to myself. Although this might arouse much amazement

and even interest, few people would perceive the existence of such duplicates as adding

infinitely to the suffering or joy of their own lives. But how is there any stronger basis for

such perception in the case of eternal recurrence? .... If this is correct, a clear concep-

tion of the lack of connection involved in eternal recurrence would remove the practical

significance it would seem to have on a personal level.”

11

I take this argument to succeed where Soll’s argument fails. Intuitively,we need not have a profound concern for individuals that are simultaneousand qualitatively indiscernible with us. But, if qualitatively indiscernibilitydoes not succeed in grounding concern here, then qualitative conception ofuniformity cannot explain how the concern requirement is met within theframework of ER. Furthermore, qualitative indiscernibility seems a ratherpeculiar requirement anyways. Suppose, for example, that each epoch dif-fers from the last by virtue of a single electron moving an infinitesimallysmall distance further from the center of the universe.23 Since individualsin different epochs would discernible with respect to their distance from thiselectron, they would fail to be qualitatively indiscernible. But this differencein properties is unlikely to strike us as having any significance for whetherthey are appropriate objects of concern. In this way, it seems quite oddthat qualitatively indiscernibility rather than some interdeterminate mea-sure of qualitative similarity would ground our concern for individuals inother epochs. For this reason, and given Nietzsche’s emphasis on the recur-rence of we ourselves—understood as requiring numerical identity—there isgood reason to reject the qualitative conception of uniformity.

A dilemma now arises in our efforts to accommodate the uniformity andconcern requirements. In light of Nietzsche’s remarks and the objectionsjust considered, the qualitative conception of uniformity ought to be re-jected. We must therefore hold that something over and above qualitativeindiscernibility must unite individuals across epochs. This relation mustthen be a non-qualitative relation, but the only plausible non-qualitativecandidate is that of identity. And, if we adopt this identity conception ofuniformity, epochs are uniform by virtue of being one and the same. But, asthe argument from PII seems to show, it will not do to claim that epochs areidentical to one another: if there is only one epoch, then there is no eternalrecurrence. There is merely a single occurrence. As a result, the identityconception of uniformity forces us to abandon the perpetuity requirement.23This proposal would, of course, butt heads with the uniformity requirement as well as the

proposal in Nehamas (1985) that the essences of individuals are so wildly interconnected

that each individual’s essence necessitates every other individual’s essence.

12

This, then, is the dilemma: neither qualitative relations nor non-qualitativerelations seem suitable for explaining uniformity and, at the same time,satisfying the requirements of perpetuity and concern.24

The dilemma just surveyed presumes that there is only one way to ex-plain the uniformity of epochs. It holds that epochs are either merely qual-itatively indiscernible or numerically identical. There is, however, a thirdview about uniformity that requires us to distinguish between the relationthat unites material objects like persons, serpents, and specks of sand acrossacross epochs and the relation that holds between the times at which differ-ent epochs take place.25 On the mixed view, individuals or material objectslike you and I are numerically identical across epochs.26 We are one and thesame throughout the myriad cycles of eternal recurrence. In contrast, thetemporal regions at which you and I recur are distinct from one another.On such a view, there is an objective ordering of an infinitely extendingtemporal region, and, at regular intervals within this ordering, you and Ipass into and fade out of existence.

To better make sense of this proposal, it will be helpful to introduce adistinction between certain views about the ontology of space and time.27

According to substantivalists, spatial and temporal regions are not depen-24The acceptance of literal identity across epochs does a nice job of alleviating another

problem regarding ER. According to what we might call the Doppelganger Problem, if

you are distinct from every individual in the epochs that precede you and your actions

must perfectly comport with the actions of these agents, then your acts are, in some

sense, “determined” by the actions of someone qualitatively indiscernible yet distinct from

yourself (i.e., your doppelgangers). But, since ER requires something like an affirmation of

one’s own actions, the notion—endorsed by the qualitative conception—that individuals

in other epochs are mere doppelgangers is highly problematic.25Given Nietzsche’s opposition to an ontology of matter and the acceptance of a meta-

physics of forces, my talk of “material objects” should be taken loosely. All such talk of

material objects would, on the present view, be recast in terms of the complex of forces.26Here, we must set aside Nietzsche’s claim that “there is nothing identical as such”. (GS

§111)27Othodoxy now holds space and time are not distinct entities, but rather a single unified

entity, spacetime. Here, I treat space and time in keeping with Netizsche’s pre-Einsteinean

physics.

13

dent upon material objects.28 Their existence is sui generis and independentof the existence of the objects that occupy them. According to relationalists,spatial and temporal regions are dependent upon material objects. Spatialand temporal regions are therefore merely complexes of relations that holdbetween objects. For this reason, relationalists hold that space and timedepend for their existence upon the existence of objects.29

On the mixed view, substantivalism is an extremely natural assump-tion.30 Not only does it allow for the desired objective distinction betweentemporal regions—a distinction that is not reducible to facts about objects—it also allows for a fairly intuitive conception of ER: Imagine the sum ofspatial and temporal regions as an enormous “container” for objects. Theseobjects come into being and pass away over the course of an epoch, but,in subsequent epochs—differentiated by occurring at numerically differenttimes—they merely reappear within the container and behave just as theydid previously. In this way, the distinctness of temporal regions allows forthe existence of distinct epochs, but the identity of objects also provides anaccount of how the concern requirement is met. You care about “yourself”in future epochs because you yourself will come into being and pass awayover and over again.

This mixed conception of uniformity is a plausible solution to the dilemmaabove. For example, Clark (1990) remarks that, in explaining how epochsare to be distinguished, “Only a difference in temporal position fits the bill—28One might worry here that substantivalism is in immediate conflict with Nietzsche’s

claim that “there are no eternally enduring substances”. (GS §109) It is worth noting,

however, that such a claim stands in opposition to a metaphysics of substantial matter

rather than space or time.29One might defend the interpretation of Nietzsche as a relationalist on the strength of

his remark that “the world... does not exist as a world ‘in-itself’; it is essentially a world

of relations.” (WP §625). See also Richardson (1996: 105).) Here, I incline to take

Nietzsche’s talk of “the world” to denote the totality of putative objects, which, given his

power ontology, emerges as a vast complex of relations embedded with the container of

space and time.30The question of whether relationalism can accommodate ER is, I believe, an open one.

For our purposes, substantivalism proves useful, since the distinctness of temporal regions

is “built into” the metaphysics.

14

the recurrence of my life must be part of an earlier or later cycle of cosmichistory than my present life (which is, of course, also a recurrence).”31 Animportant question about this metaphysics for ER is whether it might rea-sonably be viewed as the one Nietzsche intends. On this score—and almostany other concerning Nietzsche’s metaphysics—there are layers upon layersof complications regarding both his perspectivism and the extent to whichhis power ontology can be reconciled with our superficial talk of material“objects.”32 Here, I will set these broader issues aside and note that a cru-cial commitment for the present proposal is that Nietzsche’s view of ERcomports roughly with substantivalism. And, when drawing on Nietzsche’sremarks surrounding ER, there is reason for optimism. For example, hisdiscussion of statistical arguments for eternal recurrence suggest a realismabout the reality of space and time:

The total amount of energy is limited, not “infinite”: let us be-ware of such excesses in concepts! Consequently, the number ofstates, combinations, changes, and transformations of this en-ergy is tremendously great and practically immeasurable, but inany case finite and not infinite. But the time through which thistotal energy works is infinite. That means the energy is foreverthe same and forever active. An infinity has already passed awaybefore this present moment. That means that all possible trans-formations must already have taken place. Consequently, thepresent transformation is a repetition, and thus also that whichgave rise to it, and that which arises from it, and so backwardand forward again! Insofar as the totality of states of energyalways recurs, everything has happened innumerable times.33

31Clark (1990: 267).32For example, “We are operating only with things that do not exist – with lines, surfaces,

bodies, atoms, divisible times, and divisible spaces.” (GS §112). Here, it is interesting to

note that, if Nietzsche’s denial concerns the divisibility of time and space, then he would

seem to be affirming the opposing thesis that time and space are indivisible and, therefore,

real.33Nietzsches werke in Drei Brande XII: 51.

15

For Nietzsche, time is said to be infinite and plays the role of a container“through which this total energy works”. This suggests a background con-ception of time along substantivalist lines, where we can deduce featuresabout the recurrence of energy-states from metaphysically independent factsabout the objective temporal order. Given these remarks, I take it that thewe are not out of place in developing a metaphysics of ER along substantival-ist lines even while Nietzsche’s ultimate metaphysical commitments remaincontentious.

Although the mixed conception of uniformity provides a prima faciesatisfactory treatment of ER, I contend that it is an inadequate interpreta-tion. The problem faced by the mixed view arises when we recall the initialthought-experiment:

This life as you now live and have lived it, you will have to liveonce more and innumerable times more; and there will be noth-ing new in it, but every pain and every joy and every thought andsight an everything unutterably small or great in your life willhave to return to you, all in the same succession and sequence—even this spider and this moonlight between the trees, and eventhis moment and I myself.34

We saw earlier that the qualitative conception of uniformity is inadequate,since Nietzsche’s remarks demand that one and the very same individuallike you or I recurs. This requirement, along with the objections surveyedabove, pushed us towards the mixed view. But, here, Nietzsche claims thatnot only individuals like ourselves but even “this moment” and other tempo-ral regions must recur. If we are to take this passage seriously, we are forcedto treat talk of individuals and times with parity: their recurrence must beexplained in the same fashion. And, since the recurrence of times, under-stood in the same fashion as the recurrence of objects—straightforwardlyrules out the individuation of epochs in terms of an objective temporal or-dering, the mixed view is unsatisfactory. Nietzsche’s conception of ER drawsno ontological distinction between the recurrence of times and individuals,34GS §341.

16

so another response to the dilemma above is needed. In the next section,I outline a metaphysics of time and tense that accepts the same treatmentof the recurrence of times and individuals and that also allows to overcomethe dilemma poster earlier.35

3 Why So Tense About Eternal Recurrence?

My defense of an alternative metaphysics for ER requires focus upon twoissues in the metaphysics of time. The first concern the “reality of tense”;the second concerns the reality of non-present times.

Debates about the “reality of tense” can be conducted in many ways.Here, we can understand the debate to concern whether tensed propertieslike being past, being present, and being future are reducible to tenselessproperties like being five minutes prior to and being thirty-three years after.So understood, question about the reality of tense turn on whether there isa metaphysically primitive distinction between the present and non-presenttimes. For example, is there some fundamental property had only by presentthings that separates them—albeit briefly—from the past and future things?Those who believe in a fundamental distinction that privileges the presentaccept what we can call a privileged view of time. In contrast, those whothink that there is no fundamental distinction opt for a unprivileged view oftime.36

A primary virtue of the privileged theory is that it captures the intu-ition that the present enjoys special ontological significance. This intuitioncan, however, be captured in other ways. This brings us to the second is-35An interesting interpretive consequence of the present proposal is that it furnishes us

with a natural explanation for why Nietzsche’s statistical arguments remained unpub-

lished: even if Nietzsche believed them to be sound, the conclusion they purport to es-

tablish is not consonant with a view of ER that accommodates the numerical identity of

times across epochs.36The debate over the fundamentality of tensed properties is connected yet importantly

distinct from the debate over whether a semantic theory can be given that does away with

any primitively tensed notions. Unfortunately, both issues sometimes travel under the

name of the “A-Theory/B-Theory Debate”. For more on these issues, see Sider (2001)

and Zimmerman (2006).

17

sue that we will focus upon: the reality of non-present times. Accordingto presentists, the present is metaphysically significant since the only timethat exists is the present time. Understood atemporally, presentism holdsthere are no past or future times and, therefore, no past or future entitieslike dinosaurs or the 2050 World Series Champions. On such a view, thepresent is privileged by being coextensive with the domain of all entities. Onthe contrasting view, eternalism, past, present, and future times all exist.Presently existing entities are, therefore, a mere subset of all the entities thatexist, since, in some atemporal sense, both dinosaurs and the 2050 WorldSeries Champions exist. They simply fail to exist now.

Of the above described views, two natural package deals emerge: privi-leged presentism and unprivileged eternalism. For the privileged presentist,the metaphysical importance of tense is explained by noting that only invirtue of being present can something exist. For the unprivileged eternalist,the temporal features of reality can all be explained in terms of the variousrelations that hold between past, present, and future individuals.

In addition to these views, two other heterodox views remain. The firstview, unprivileged presentism, borders on triviality or incoherence. It holdsthat there is no fundamental distinction that the present enjoys, but deniesthat non-present entities exist. The second view, privileged eternalism, ismore interesting.37 On this view, past, present, and future entities exist,but presently existing entities are held to instantiate a fundamental property,being present, that distinguishes them from past or future entities.

What I want to now suggest is that the metaphysics of ER is best un-derstood as one according to which privileged eternalism is true. I take thisproposal to square with the central features of ER Nietzsche outlines in thepassages surveyed above. In particular, it succeeds in capturing the require-ment that temporal regions and individuals recur in the same fashion. Sucha view is also forced upon us by the dilemma considered above. Earlier,we saw the qualitative conception of uniformity to fail. Here, I will argue37There is some issue over which philosophers actually endorse such a view. Perhaps

the only uncontroversial proponent is Smith (1993). See Zimmerman (2006) for further

discussion.

18

that the non-qualitative conception can be salvaged by appeal to privilegedeternalism.

To see how privileged eternalist view sustains ER, it will be helpful tocontrast it with the mixed view considered above. On that view, extraontology was proposed as a solution to problems with ER. Specifically, ob-jectively distinct temporal regions were posited to “house” the multitudeof epochs. We saw, however, that this proposal was unsatisfactory, sinceNietzsche’s remarks regarding ER suggest that both objects and tempo-ral regions must themselves recur. Given this requirement, ER cannot beaccommodated by expanding our ontological commitments (i.e., increasingthe domain of objects or temporal regions). Instead, primitive ideology—metaphysically basic distinctions—are needed to explain ER, and, in theform of privileged eternalism, the passage of time involves exactly this: aprimitive distinction that demarcates the present temporal region from othertemporal regions. This distinction is constantly “moving” and, in this way,an account of “becoming” and the passage of time can be accounted for. Inlight of the dilemma above, it is natural to help ourselves to the ideologi-cal resources of privileged eternalism—specifically, the primitive property ofbeing present—to make sense of ER.

The first step in developing a privileged eternalist view of ER is to graspone horn of the dilemma surveyed earlier: we should accept the identityconception of uniformity. Rather than taking ER to require a recurringworld to be a stretched out like a infinite chain of paper dolls, the worldis identified with what commonsense would take to be “the universe”—thesum of the spatial and temporal regions between, say, the Big Bang and alooming Big Crunch. Compared to the mixed view, this delivers a rathermodest ontology, since the mixed view would be required to duplicate theuniverse infinitely many times over.

The second step is to notice an errant assumption in the argument fromPII considered earlier. This assumption should be clear from our discussionof the reality of tense above. That earlier argument purported to showthat ER was impossible if the world was not “stretched out” by havingnumerically distinct temporal regions for each epoch. Here, however, we

19

need to distinguish between ontology and ideology. Most notably, if thereis but one “universe”, there might still be myriad recurrences provided thatthe distribution of the primitive property, being present, follows a certainpattern. Specifically, ER can take place within a single universe, if beingpresent is instantiated infinitely many times by each and every temporalregion of that universe. Adopting the familiar metaphor of the “movingspotlight”, ER can be understood as follows: after having shone on theentire universe in sequence, the spotlight returns to the beginning of theunivesre and begins to shine on the entire univesre once more.38 By positinga primitive property of being present, the privileged eternalist can hold thatevery moment and every individual has been and will be illuminated adinfinitum.

On this view, the perpetuity requirement is met, not by positing distincttemporal regions in the form of infinitely many universes, but by holding aprimitive temporal property to cycle through the temporal regions of a singleuniverse over and over again. As Nietzsche puts it, “the eternal hourglass ofexistence is turned upside down again and again.” The concern requirementis met by virtue of individuals being identical to themselves “across” epochs,since there is but a single universe that encompasses all of us and no meredoppelgangers. The uniformity requirement is met in a similar vein: we areboth identical with and qualitatively indiscernible from individuals in otherepochs, since there are no other “epochs” (i.e., entities or temporal regionsdistinct from ourselves and our fellow universe-mates).

On the present view of ER, the spotlight of the present “moves” through-out temporal regions. Taken at face value, this talk of a “moving present”would seem to require a commitment to a kind of higher-order temporaldimension. Such a view is, however, likely to strike many as unattractive.It will be helpful, then, to note that recent commentators on the movingspotlight view have defended privileged views from this charge by takingon a commitment to conceptually primitive tense operators, which, onceaccepted, obviate the need to provide an analysis of the “moving” of the38For more on the ”moving spotlight”, see Sider (2001: 17-21).

20

present.39 Here, our present view of ER ought to follow suit and grantthat the distinction between present and non-present times is conceptuallyprimitive and that no appeal to higher-order temporal dimensions is needed.

A notable virtue of the present view of ER is that it makes good onNietzsche’s repeated remarks that time is, in some respect, “cyclical”. Noticethat, on the competing mixed view, time is not itself cyclical, but, instead,material objects exhibit a bizarre kind of regularity. On the present view,however, the cyclical character of time is a function of the spotlight of thepresent shining over the entire temporal region of the universe ad infinitum.In this respect, better sense can be made of a salient feature of Nietzsche’stalk regarding ER.

I have suggested that we can understand ER in terms of a single “uni-verse” that exhibits a peculiar pattern of instantiation with respect to afundamental property, being present. I have indicated that such a viewaccommodates the perpetuity, concern, and uniformity requirements. Inthe next section, I consider the interpretive consequences for the view justsketched and the broader question of whether it might comport with Niet-zsche’s broader metaphysical commitments.

4 Conclusion

In considering the metaphysics and interpretation of ER, Long (1987) states:

The vision of eternal recurrence is precisely the experience ofBeing as Becoming without end. Quite simply, there is no lan-guage which can capture this as an idea. Not only is the truthof the eternal recurrence outside the domain of the public, as wehave seen, but the very meaning of eternal recurrence cannot beexpressed in language.40

Long’s passage typifies a natural strategy for reconciling ER with other Ni-etzschean commitments: A broader commitment that is borne out in the39For discussion, see Sider (2001).40Long (1987: 440).

21

Nietzschean corpus—here, a certain interpretation of his perspectivism—is granted interpretive primacy over Nietzsche’s apparently intelligible re-marks on ER. This outcome is, in some respects, highly understandable.Nietzsche’s remarks regarding perspectivism and his rejection of “Being”in favour of “Becoming”, upon quite natural interpretations, threaten toreduce ER and any other putative metaphysical hypothesis to “mere” inter-pretations. On such interpretations, there is little interest to paying detailedattention to the metaphysical peculiarities of ER, since it cannot reasonablybe taken as thesis about something like an “objective world”.

For those who incline to give Nietzsche’s perspectivism a central rolein interpreting his metaphysics—one that deflates the possibility of any ro-bust metaphysical hypotheses—the preceding is liable to be of only limitedinterest. In contrast, for those who incline to take his talk of “the hard-est, most terrible insight into reality” at face-value, the picture of ER thatemerges is likely to be very different. Here, my aim has only been to considerwhat Nietzsche’s immediate remarks suggest as the default metaphysics ofER. And, as we have seen, the resulting view of ER—a brand of privilegedeternalism—is a provocative thesis when set against the backdrop of presentorthodoxies regarding time and tense.

References

[1] Adams, R.M. 1979. “Primitive Thisness and Primitive Identity.” Jour-nal of Philosophy 76: 5-26.

[2] Black, Max. 1952. “The Identity of Indiscernibles.” Mind 61: 153-164.

[3] Clark, Maudemarie. 1990. Nietzsche on Truth and Philosophy. NewYork: Cambridge University Press.

[4] Danto, Arthur. 1965. Nietzsche as Philosopher. New York: MacMillan.

[5] Della Rocca, Michael. 2006. “Two Spheres, Twenty Spheres, and theIdentity of Indiscernibles.” Pacific Philosophical Quarterly 86.4: 480-492.

22

[6] Hatab, Lawrence. 2005. Nietzsches Life Sentence. New York: Rout-ledge.

[7] Kain, Philip. 1983. “Nietzsche, Skepticism, and the Eternal Recur-rence.” Canadian Journal of Philosophy 13: 365-387.

[8] Kripke, Saul. 1972. Naming and Necessity. Cambridge: Harvard Uni-versity Press.

[9] Lewis, David. 1986. On the Plurality of Worlds. Oxford: Blackwell.

[10] Loeb, Paul. 2006. “Identity and Eternal Recurrence.” In A Companionto Nietzsche (ed. Keith Ansell Pearson) Oxford: Blackwell.

[11] Long, Thomas. 1987. “Nietzsches Eternal Recurrence—Yet Again.” Ni-etzsche Studien 16: 437-443.

[12] Magnus, Bernd. 1978. Nietzsches Existential Imperative. Bloomington:Indiana University Press.

[13] Nehamas, Alexander. 1985. Nietzsche: Life as Literature. Cambridge:Harvard University Press.

[14] Parfit, Derek. 1984. Reasons and Persons. Oxford: Oxford UniversityPress.

[15] Reginster, Bernd. 2006. The Affirmation of Life. Cambridge: HarvardUniversity Press.

[16] Sider, Theodore. 2001. Four-Dimensionalism. Oxford: Oxford Univer-sity Press.

[17] Smith, Quentin. 1993. Language and Time. New York: Oxford Univer-sity Press.

[18] Lewis, David. 1986. On the Plurality of Worlds. Oxford: Blackwell.

[19] Simmel, Georg. 1907. Nietzsche und Schopenhauer. Leipzig: Dunckerund Humboldt.

23

[20] Soll, Ivan. 1973. “Reflections on Recurrence.” in Nietzsche: A Collec-tion of Critical Essays. (ed. Robert Solomon) New York: Anchor Books.

[21] Zimmerman, Dean. 2008. “The Privileged Present: Defending an A-Theory of Time.” in Contemporary Debates in Metaphysics (eds. Sider,T. et al.. Oxford: Blackwell.

[22] Zuboff, Ivan. 1973. Nietzsche and Eternal Recurrencein Nietzsche: ACollection of Critical Essays. (ed. Robert Solomon) New York: AnchorBooks.