Given the strangeness of there being ways of being that are made up of distinct parts, it is not surprising that some commentators who focus on the parts of time doctrine (perhaps, inadvertently) seem to credit Descartes with offering a substance-like account of time. This is a problematic tendency insofar as scholarship related to Descartes’ account of time has primarily focused on the nature of these temporal parts. This has been the focus because analysis has generally related to determining whether the parts in Descartes’ account of time are continuously or discontinuously linked. In Descartes’ Metaphysical Physics, Daniel Garber provides a brief overview of this classic debate between proponents of the discontinuity and continuity theses. Garber explains that the former, “received” view proposes that Descartes presents time as being composed of discontinuous “temporal atoms.” [11] As such, the received view clearly follows the language of “independent parts” sketched above. Accordingly, this type of interpretation is particularly amenable to regarding Descartes as positing time to be an independent entity that exists apart from those things that are in time.
One can see how a proponent of discontinuous temporal atomism might credit Descartes with offering a substance-like account of time in Ken Levy’s recent (2005) article, “Is Descartes a Temporal Atomist?”[12] In this article, Levy argues that Descartes offered a strongly discontinuous[13] variety of temporal atomism. He explains his interpretation by using the analogy of a cut string. He writes:
[Some philosophers] become susceptible to one or another version of temporal atomism when they treat time as though it were a piece of string that can be cut into separated parts and then mistakenly take this model to be an accurate representation of time. Descartes, I shall argue, is just one such philosopher.[14]
In other words, Levy argues that Descartes held time to be something that exists as a composite entity composed of disconnected temporal atoms (which are conceived of as really distinct parts). As a participant in the atomism debates, Levy’s argument (like the arguments of many others) does not begin by inquiring about the ontological status of the string, but rather by focusing on its internal structure.
In conceiving of Descartes’ account of time as a composite whole (i.e., as a string with parts), Levy shows a tendency to treat time as if it were only dependent on its “parts” and not dependent on some substance—as would be the case if time were an attribute. Levy argues, for example, that the question of whether one moment of time will be followed by another can be treated independently from the question of whether a substance “in” time will continue to endure. He writes:
If it is logically possible that time end at any given moment t, then it is certainly logically possible that I—who exist in time—cease to exist at t. But it is not clear that [my not being guaranteed to exist at a next moment] depends on [the fact that the parts of time are independent of each other]. . . Even if every earlier part of time logically guaranteed that there be a later part of time, if it were (therefore) logically impossible that time end, it would still be the case that I could cease to exist at any moment. Time’s immortality would hardly guarantee my immortality. So the fact that my ceasing to exist is logically possible does not necessarily depend on whether or not it is logically possible for time to end.[15]
In this argument, if one were to conceive of time as an attribute of a thing in time, then Levy would be arguing that the existence of the substance is not dependent on the existence of the attribute. Though it is of course correct that attributes depend on substances and not vice versa for Descartes, by divorcing the question of time’s continuation from the question of the continuation of a thing in time, Levy shows his tendency to presume that the existence of something in time is irrelevant to the question of time’s existence. This tendency treats Cartesian time as if it were a Newtonian container. But, if time might be a type of attribute, then such a divorce is a mistaken way to conceive of Cartesian time. Though it would still be fair for Levy to claim that an infinite time would not guarantee the immortality of any particular substance, such a determination would guarantee the eternal existence of some substance or other. Since there can be no free-floating attributes, if logic demanded time’s eternal existence, then an eternal time would also demand the eternal existence of some substance of which time is an attribute.
Levy elsewhere makes his presupposition that time is a substance more explicit. Indeed, Levy goes so far as to argue against regarding time as a type of attribute. He does this by arguing that time is something other than duration, and that only duration is an attribute that inheres in a substance. He claims, “it does not really make sense to say that time itself is ‘in the very things of which [it is] said to be [an] attribute or [mode].’. . . we tend to think not that time is in (no less an attribute of) all things but rather just the opposite – that all things are in time.”[16] Accordingly, Levy concludes that “time is not equivalent to ‘duration taken in the general sense’ and therefore is not necessarily ‘[an] attribute or mode.’”[17] These passages clearly indicate Levy’s belief that Descartes offers some account of time that is not merely an attribute of substance. Levy argues that time in Descartes is a thing in which temporal things endure (like the Newtonian container view), and not that time is something dependent on its enduring within temporal things.
Though Levy’s account only shows one example of someone attributing a sort of independent existence to Cartesian time, reflecting on the “parts of time” language has shown why such a tendency might be reasonable. Such a tendency carries significant metaphysical weight, however, since it further assumes that time is a type of Cartesian substance and thus that Descartes is presupposing an absolutist account of time. Having looked at the textual evidence that seems best to support an absolutist reading of Descartes, I can now consider the significant evidence against such an interpretation. The contrary evidence proves that any tendency to presume a sort of absolutism in Descartes’ account of time is not only mistaken, but one that establishes a significant road block to understanding the nature of Cartesian time.
Four Arguments against Time being a Cartesian Substance
Though Descartes’ “parts of time doctrine” may seem to invite the interpretation that time is a sort of composite substance, Descartes provides other discussions that exclude this interpretation. By considering these discussions, I will provide four arguments against time’s being a Cartesian substance. I will refer to these arguments as:
The Argument from Attributes
The Argument from Principles I.57
The Argument from Radical Contingency
The Argument from Numerically Distinct Parts.
Whereas arguments (1) and (2) will focus on various metaphysical commitments that Descartes offers in the Principles, arguments (3) and (4) will offer related arguments that focus on Descartes’ proof for God’s existence based on the (troublesome) “parts” of time. I will conclude that contrary to first appearances, Descartes’ “parts of time doctrine” does not entail the substantiality of time. Rather, this doctrine itself supports an attribute theory of time.
The Argument from Attributes
The first textual argument against time’s being a Cartesian substance can be extracted from Principles I.52 and I.53. This ‘argument from attributes’ follows Descartes’ claim at I.52 that “corporeal substance and mind” are the only two types of created substances. This claim, when conjoined with I.53’s claim that there is just one primary attribute characteristic of each of these types of substances (i.e., extension for body and thought for mind) provides a test for determining if something is a created substance; namely, if X is a created substance, then X is either a corporeal substance or a mind. If X is a corporeal substance, then it is extended in “length, breadth and depth” and if X is a mind, then thought constitutes its nature. Thus, for time to be a Cartesian substance, it would need to be either spatially extended or thinking. Yet, both of these descriptions seem inapplicable to an account of time. Indeed
, it is difficult to conceive how any account of time could credit time itself with being either spatially extended or thinking. Thus, unless a proponent of the substance interpretation could show how the substantial time it defines could be either spatially extended or thinking, then this account is inconsistent with the metaphysical landscape that Descartes defined.[18]
The Argument from Principles I.57
A second textual argument against time’s being a substance is what I refer to as ‘the argument from Principles I.57.’ In I.57 Descartes explicitly states his view that time is a type of attribute. This passage clearly offers an argument against time’s being a substance so long as one grants that Descartes offers a consistent account of time between this principle and the “parts” language of the Meditations and elsewhere. In Principle I.57, Descartes addresses the question of what types of beings time and duration are, given that, “some attributes are in things and others in thought.”[19] Since I.57’s stated agenda is merely to determine the type of attributes that duration and time are, this passage shows that Descartes is not even considering the possibility that time might be a sort of substance. That time is some sort of attribute is a fact that he simply takes for granted. Given that I.57 simply assumes that time is a type of attribute, and since Descartes could not consistently maintain that time is both a type of attribute and a type of substance, this passage provides strong evidence against reading the “parts” language of Med III as positing time to be a type of aggregate substance.
Though it may seem as though I.57 is alone sufficient for refuting the hypothesis that time is a Cartesian substance, one might doubt the conclusiveness of this passage by arguing that I.57 is so unlike the “parts” account of time that one ought to simply assume that Descartes offered inconsistent views on time. Indeed, when Ken Levy briefly discusses Principles I.57,[20] he suggests that the apparent inconsistency between the parts account of time and the account offered in I.57 is significant enough to warrant treating this passage as a mere anomaly. Indeed, he claims that one can infer very little from “minor passages tucked away in the Principles.”[21] Even if one disagrees with how Levy’s evaluates the significance of I.57, one might still agree that this passage is so unlike the discussions of time given in the Med III proof that it is appropriate to treat it as anomalous.
It is certainly reasonable to suppose that Descartes’ views on time might have undergone changes, or even to think that his views might not always have been consistent. In the following two arguments, however, I not only provide additional grounds for denying that time is a Cartesian substance, I also show why I.57 should not be read as anomalous. That is to say, I show how a close examination of Descartes’ claims about the parts of time indicates why this “parts” account is itself inconsistent with time’s being a substance.
The Argument from Radical Contingency
For my third argument against time’s being a Cartesian substance, I focus on the proof for God’s existence based on the parts of time. Remember that this argument seems to presuppose the independence (and thereby the substantiality) of time. Contrary to first impressions, this proof actually presupposes the impossibility of the substantiality of time. In the following ‘argument from radical contingency,’ I consider what is required for the possibility of individual moments. Recall that in the Med III proof Descartes explains that a “lifespan” can be divided into “countless parts, each completely independent of the others.”[22] From this, Descartes concludes, “it does not follow from the fact that I existed a little while ago that I must exist now, unless there is some cause which as it were creates me afresh at this moment—that is, which preserves me.”[23] In other words, the Med III proof establishes the radical contingency of substance. Descartes establishes that the existence of every substance is contingent on God’s concurrence (i.e., creation/preservation) at each and every moment as the immediate cause of its continued existence.
By focusing on the radical contingency thesis, one learns something about the nature of created substances and how they are dependent on God. In Principles I.51, when Descartes defines substance he explains that created substances are not substances in the same way that God is a substance because “they can exist only with the help of God’s concurrence.”[24] The analysis of the parts of a lifespan given in Med III clarifies the nature of this concurrence. Descartes argued that created substances are substances insofar as they depend on nothing other than the concurrence of God. As the radical contingency thesis shows the way in which a substance is dependent on God, it provides a standard for determining if something within the Cartesian metaphysic might be a substance; namely, does the thing in question possess the sort of independence whereby it could exist at successive moments given divine concurrence alone? In the following argument, I will show why time cannot meet this necessary condition for substantiality:
The Argument from Radical Contingency
Time has parts (namely, moments). (Med. III)
According to radical contingency, substances can exist at one moment (that is, could exist even if there were no prior or later moments of them).
Suppose time is a substance. (for reductio)
If time is a substance, then time can exist at one moment. (2, 3)
Thus, each moment of time could exist even if there were no prior or later moments of time. (The most charitable interpretation of the consequent of 4)
Moments of time are essentially relational.
Anything that is essentially relational requires the existence of at least two relata for its possibility (namely, itself and that to which it is related). (Definition of ‘essentially related’)[25]
It is not possible for a moment of time to exist if there are no prior or later moments of it. (6, 7)
Lines 5 and 8 contradict.
Therefore assumption 3 is false. Time is not a substance.
There are certain oddities in this argument that cannot be avoided, in addition to some controversial premises. It is undoubtedly odd, for example, to try to conceive of time existing at a moment (i.e., line 4). Though it may indeed seem bizarre to conceive of time at a time, the oddness in this maneuver is compelled by the analysis of radical contingency just considered. If my analysis is correct, then Descartes’ radical contingency doctrine provides a test for all substances; namely, if X is a substance, then X can exist at a moment. Thus, to analyze the hypothesis that time itself might be a substance, I plug time in for X, and end up with line 4 of the argument. As I indicate in line 5, it seems that the most charitable way of applying this general rule to time is to take it as suggesting that if time were a substance then it would be possible for one moment of time to exist in isolation of other moments. Accordingly, as the argument proceeds, I consider whether it would be possible for a substantial time to exist at a single moment, and determine that the answer is no.
The most controversial premise of this argument is surely found on line 6, where I claim that “moments of time are essentially relational.” This argument attempts to construct the most reasonable position that someone positing time to be a Cartesian substance might offer. If one is arguing that time is a substance in Descartes, then one is explicitly arguing that time is capable of independent existence. To argue that time is capable of independent existence, however, one must be able to offer some account of what it would be for time to exist at all. Given that Descartes posits time to be a collection of parts (i.e., moments), one must be able to give some account of what it is for a moment to exist as such. In my argument’s assertion that moments are essentially relational, I merely suggest that what it is for a moment t2 to exist is for it to be a moment subsequent to moment t1 and prior to moment t3. In other words, I argue that t2’s nature is “essentially relational.” As I explain in line 7, “anything that is essentially relational requires the existence of at least two relata for its possibility (namely, itself and that to which it is related).” With this seemingly plausible account of moments in hand, I then simply consider t
he criteria for substance that are provided by Descartes’ account of radical contingency in conjunction with this account of how moments might be individuated. The Argument from Radical Contingency shows that time could not both meet the criteria for substance suggested by his account of radical contingency and maintain a relationalist account of moments. Thus, time could not be a Cartesian substance, because it would not possess the sort of independence whereby it could exist at a single moment given divine concurrence alone. Time could not exist “at a moment” by divine concurrence alone because each moment of time seems minimally to require the existence of an additional moment to which that moment could be related.
The Argument from Numerically Distinct Parts
Though I believe that the Argument from Radical Contingency assumes a reasonable account of moments and then proves that time could not be a type of Cartesian substance, it seems that one might object to the important premise that moments are “essentially relational” on textual grounds. One might claim that this particular account of moments (despite its plausibility) could not have been held by Descartes, since in Principles I.21, Descartes claims that the “parts [of time] are not mutually dependent, and never coexist.”[26] Since Principles I.21 denies that moments are mutually dependent, it seems that I.21 might show that Descartes cannot appeal to the external relations between moments as establishing the nature of an individual moment. He could not appeal to these external relations since this would make the nature of individual moments mutually dependent. In other words, the very idea behind the Argument from Radical Contingency (namely, that time could not be a substance because no moment can exist in isolation) appears to be denied by Principles I.21. This principle seems to suggest that the nature of a moment cannot be determined in virtue of its location in the timeline, since this hypothesis makes moments mutually dependent and Descartes explicitly states that moments are not mutually dependent.
Descartes' Temporal Dualism Page 6