If we take any baby born today (let’s call the individual we pick ‘Baby’), the history of its life has been born at exactly the same time. That history (let’s call it ‘H’) will grow relentlessly from nothing, adding a year to its girth every time Baby ages a year. If Baby eventually dies, at that moment H will reach its greatest size, and it will subsequently remain frozen in that state for the rest of time.
Such histories are clearly entities that inhabit time, because they can change (grow). The interesting thing is this: whatever our theory of how Baby persists through time may be, H has to persist by enduring. It can’t have temporal parts. (See here and here for the jargon.)
That last claim sounds silly at first. Surely, for instance, the history of Baby’s first day at school (let’s call this ‘H*’) is (or will be) a temporal part of H, with an extent of one day?
Well, H* is certainly a part of H, but it looks like it can’t be a temporal part. The reason is that it doesn’t exist only for one day – once H* exists, it will exist for ever. In fact, the day that forms the content of H* is the last day when H* doesn’t (fully) exist. A second candidate for a temporal part of H is the history of Baby’s life up until its first day at school, which we can call ‘H#’. We could say, whereas H* is the history of that day, H# is the history as ofthat day. It’s just possible to think that, unlike H*, H# doesn’t exist forever, because it is a temporary version of H and is in some sense replaced or consumed by later versions of H. But sub-histories like H# also can’t be temporal parts of H. For one thing, if Baby dies, all those parts from times after Baby’s death will be identical with H, and it’s absurd that a thing could be a proper temporal part of itself. For another thing, the proposal has trouble assigning parts to intervals of time (as opposed to mere moments) except by making the history as of an interval identical with the history as of its first moment. But that solution couldn’t deal with histories of beings that have existed for an infinite period, because the interval of time up to any point in the life of such a being couldn’t be assigned a corresponding part of its history.
It’s difficult to think of any alternative to things like H* and H# as temporal parts, though, because it is the role of such parts to account for changes in H by having different of its properties at the different times they each correspond to. Since the most significant change in H is its growth, its temporal parts will have to be the sort of thing that can have lengths in the sense that histories have lengths, so they’re going to be histories themselves. It looks like one and the same history is present in any interval or moment of time that H exists, though that one history is changing all the while.
*Another example is what I’ll call Baby’s ‘destiny’, which existed unchanged from the beginning of time to the moment of Baby’s birth, since when it has been shrinking inexorably, and will continue to shrink; if Baby ever dies, its destiny will expire at precisely the same moment.
So, some things endure.* For an example of what doesn’t, we can just look to the (hopefully) decades-long event (let’s call it ‘E’) that is Baby’s entire life. As an event, E persists by having temporal parts. It forms a kind of perduring trace in time of Baby’s enduring history, H. It’s pretty clear that, for any history, there will be a similar corresponding event, and vice versa. (Non-temporal) parts of H, like H*, will correspond exactly with temporal parts of E. It is probably this duality that makes it initially attractive to think that H* is a temporal part of H. Maybe the existence of an E for each H, as it were, will lead people to suggest that there are no histories like H except as unnecessary re-descriptions of events like E. I’m not really sure what to say about that potential view, except that I don’t like it and I wish it wouldn’t keep getting in the way of perfectly good metaphysical whimsy.
On this point, it might also be tempting to suggest that histories as I’ve described them don’t exist, and I just have a confused picture of Baby’s history as a set of sets of propositions that is parametrised by times (so that any time T corresponds to the set of propositions that have been true of Baby by T, or whatever). I want to insist, though, on the existence of histories that genuinely come into existence and experience change.
What use all the above is, I’m not sure. I suppose it’s an obstacle for the perdurantist attitude – which I find very appealing – that if a persisting thing has one property at one time and another at another, it’s because it has temporal parts for those times with those different properties. Since a history can have different lengths at different times, but can’t have parts with different lengths at those times (because to have lengths in the right way they’d have to be sub-histories and, as above, that doesn’t work out), that attitude would lead us to give the wrong answer for histories.
On a meta-disciplinary level, perhaps this issue could be used by a perdurantist about things like Baby to explain why others are tempted to wrongly believe that Baby endures – maybe they are confusing Baby with something like its history. (If you are committed to a strong narrative conception of the self then maybe that will sound like a proper association rather than a confusion.) At the same time, an endurantist about things like Baby could point to the difference between the sub-histories of a history, like H*, and temporal parts, like those of events, and suggest that perdurantists are seeing the latter where they should be seeing the former.
It’s also quite possible that something like the above has already been widely recognised and it turns out that nothing interesting comes of it. Or that it hasn’t, because it’s already obvious that nothing interesting can come of it. If either is the case, someone please tell me.