The thing about time is that it elapses — obvious, really, but nevertheless a fact that has been ignored or denied by time theorists from Boethius to Einstein to my own past self. Any adequate model of time must take account of the fundamental fact that time passes. It just does. No static model can be considered a model of time.
One person who made a valiant effort to deal with the fact that time elapses, and to work out the ramifications of that idea, was the philosopher and airplane designer J. W. Dunne. He first published his ideas in An Experiment with Time, where his theory of time, expressed in confusing mathematical jargon and illustrated with even more confusing diagrams, was overshadowed by that book’s more sensational claims about dreams and precognition. He followed up Experiment with the even more abstract and confusing The Serial Universe (tying Dunne’s theory of time in with then-new developments in physics), and then with the relatively lucid The New Immortality, Nothing Dies, and Intrusions? — but by then no one was paying attention. Of his works, only Experiment was ever widely read, and by now even it has faded from public consciousness.
Nevertheless, Dunne was right — or at least partly right — or at least he approached the problem from the right angle, and it is on his foundation that we must build in trying to solve the riddle of time and its passage. This post represents my stab (perhaps the first of many) at reformulating Dunne’s key insight in my own words.
It was sometime in September or October 1963, a month or two before the assassination of John F. Kennedy, that Bob Dylan wrote the lines that will serve as our theme: “The present now will later be past . . . for the times, they are a-changin’.” In the discussion that follows, it will be convenient to adopt the fiction that that point in time is the present — since the real present is a moving target, and the real future is unknown. Let us suppose, then, that it is currently the autumn of 1963, that Kennedy is alive and kicking, that Ed Sullivan has discovered the Beatles but not yet booked them for his show, and that Bob Dylan is at this very moment jotting down some lyrics for a song which will later give its name to his third studio album.
Now let’s consider some states of affairs that (from the standpoint of our designated “present”) are the case, some that were, and some that will be. Ideally, we should be looking at entire states of the universe, but it will be simpler to limit ourselves to one constituent fact of each such state: the identity of the sitting U. S. President. We will represent these states of affairs graphically, as below.
As should be fairly obvious, each of the head shots above represents the fact of that person’s being the sitting President. The red frame around Kennedy’s photo represents the fact that he is the President, while the other fellows (from left to right, Eisenhower, Johnson, Nixon, the bluegrass fiddler Clarence “Tater” Tate, and Truman) are not.
While Kennedy’s presidency is uniquely real, the other five are not equal in their unreality. Truman and Eisenhower were Presidents in the past, which is something anyway, and Johnson and Nixon will be, but nothing of the kind can be said of Tater Tate. We can add this information to our diagram by means of additional colored frames to indicate the ontological status of each state of affairs: black for past states, green for future ones, and no frame at all for the one that is wholly counterfactual.
Notice that we still don’t have a timeline, per se; the photos are arranged alphabetically by last name, and all we can say about the fact represented by each is that it is true, was true, will be true, or none of the above. We do not as yet have a way of expressing the idea that, say, Truman was President “before” Eisenhower. That will come later.
But first, let’s tackle Dylan’s obvious-yet-cryptic statement: The present now will later be past. How might we represent that? Kennedy’s presidency is present now but will later be past. Our diagram must represent it as both present (red) and past (black), and must further indicate that the former is its status now (red), while the latter will be its status later (green). The only way to represent this is with two different levels of colored frames.
As before, red, black, and green frames represent present, past, and future states of affairs, respectively, but every state now has two different levels of temporal status. Kennedy’s photo appears in both a red frame (for “present”) and a black one (for “past”), but the red-framed Kennedy is inside a larger red frame (indicating that this state is the present now), whereas the black-framed Kennedy is inside a larger green frame (indicating that it will later be past).
This introduction of a second level of time is necessary in order to model Dylan’s statement; otherwise “will later be past” would be a flat contradiction in terms. At best, it could be interpreted as a counterfactual — as meaning “If what is in fact the future were instead the present, then what is in fact the present would instead be the past.” If we want to ascribe any sort reality at all to the past-ness of Kennedy’s presidency, to represent it as anything other than a purely counterfactual hypothesis like that of the Tater Tate administration, we need this second level of time.
This new diagram also allows us to express the idea that Nixon’s presidency comes “after” Johnson’s: In the larger frame where Johnson’s frame is red, Nixon’s is green; where Nixon’s is red, Johnson’s is black. We are therefore now justified in arranging our five successive states in a sequential timeline — or, rather, in five timelines forming a two-dimensional “time-plane.” (At this point, we can also dismiss Tater Tate from further consideration; he has served only to emphasize that past and future states are not non-existent in the same sense as purely counterfactual states.)
This two-dimensional arrangement of states gives us a convenient way of referring to the two different levels of time: the “horizontal” time in which Kennedy comes after Eisenhower and before Johnson, and the “vertical” time in which red-Kennedy comes after green-Kennedy and before black-Kennedy. (This is, I repeat, simply a matter of convenient terminology. There is obviously nothing actually “vertical” or “horizontal” — or “red” or “green” — about either type of time, nor am I necessarily trying to say that they should be considered two actual “dimensions” in a physical sense.)
We can therefore annotate Dylan as follows: “The present (in horizontal time) now (in vertical time) will later (in vertical time) be past (in horizontal time).”
One problem raised by the above diagram is that it seems to mean that the future “already exists” in some sense and is therefore predetermined. Kennedy is President now, but it is already true to say that he will be succeeded by Johnson and then by Nixon — which seems to negate the free will of the assassins and voters responsible for that series of events.
One possible solution is to deny that future states have any sort of existence at all — to say that now, while Kennedy is President, Johnson’s presidency has no more reality than Tater Tate’s. But this seems inconsistent both with Einsteinian relativity (specifically, with the relativity of simultaneity) and with the data of precognition (since no one could see the future unless there were something to see). Fortunately, our model allows for another solution: that the future already exists but can be changed.
The idea of “changing the future,” while common in our everyday thought and speech, is strictly nonsensical in a one-dimensional model of time, just as much so as that of “changing the past.” To say that the future “was” A but “later” changed to B is meaningless and self-contradictory without a second dimension of time. Our two-dimensional model allows us to express the idea that the future was literally changed by the lone gunman on the grassy knoll or whatever they call him.
The diagram above expresses the idea that Kennedy was going to serve a full term or two, after which some of his charm would wear off, the other party would get its turn, and he would be succeeded by Nixon and then by that other guy. You know the one I mean. (Notice how neatly our English expressions “was going to” and “would” — past-tense forms of future constructions — denote the green-within-black states of our two-dimensional model.) But at this very moment Mr. Grassyknoll has just made his fateful decision to take out Kennedy prematurely, so as of now the future has changed and Johnson is going to be the next President (though, with the grim inevitability of Greek tragedy, Nixon’s turn in the White House will still come round in the end).
Now whether people actually can change the future (or the past, for that matter) is an open question, but the point is that this model of time at least allows us to ask it, whereas the standard one-dimensional model would rule it meaningless and not-even-false.
The ideas laid out in this post represent a first step only in creating a satisfactory model of time, but I think it is an important one. I hope to get deeper into Dunne’s theory, and my interpretation of it, in future posts. (I might throw in a bit of Ouspensky for good measure.) The question of agency is inseparable from that of time, so finding or creating a workable model of time — one which, whether ultimately “correct” or not, will at least give me a starting point from which to think — is a high priority.