The Conservation of Five Weeks

I have discovered a temporal anomaly. For some considerable time now, my house move has been approximately five weeks away.

This ought not to be possible. Five weeks is a measurement of duration, and duration is generally expected to diminish as time passes. If something is five weeks away today, then barring interference with the space-time continuum, next Sunday it should be four weeks away. Ours doesn't work like that. Time has continued to pass in the usual direction, there has been measurable progress towards moving house, and yet the distance between me and actually doing it has remained stubbornly unchanged.

There are several established phenomena that appear capable of explaining this, but irritatingly none of them quite does.

The planning fallacy certainly has a hand in it. Humans are notoriously optimistic about how long future tasks will take, even when experience provides ample evidence that optimism is unwarranted. Hofstadter's Law is more knowing: everything takes longer than you expect, even when you take Hofstadter's Law into account. Both explain why an anticipated date might prove hopelessly optimistic, but neither explains the peculiar preservation of the interval itself. I am not merely discovering that five weeks was actually eight. I am using up the weeks and finding the original five apparently untouched.

Zeno's paradox at least has the decency to shrink. Travel half the remaining distance, then half of what's left, and you approach the destination without ever quite arriving. It is a paradox of diminishing returns. Mine has stopped diminishing. Mine has acquired tenure.

There is also the receding horizon, which is closer still. Move towards a horizon and it moves with you, since the destination is partly defined by where you are standing. I recognise this. I do not recognise the part where a horizon eventually hands you the keys.

What I seem to be dealing with is a hybrid: the optimism of the planning fallacy, the inevitability of Hofstadter, the unattainability of Zeno and the mobility of a receding horizon. Each accounts for one property of the thing, but no existing law seems to encompass the whole peculiar arrangement.

The part I particularly like is the mathematics. If the predicted moving date is always the present date plus five weeks, then the predicted date is not fixed at all. The interval is. Time can pass perfectly normally on both sides of the equation while the answer continually relocates itself into the future.

Five weeks, of course, is merely the value it has taken in my case. There is no reason to suppose it is universal. Another uncertain process might settle at four weeks, six weeks or three months. The important thing is not the number but its persistence. Once the estimate has stabilised, the interval remains constant while the predicted date moves.

I have therefore provisionally named this the Fixed Interval Effect: during any sufficiently prolonged process with an uncertain end date, the estimated time remaining may stabilise at a particular interval. Thereafter, the predicted end date advances with the present, preserving that interval despite the passage of time.

There is an obvious difficulty with this. Things do eventually happen.

But that doesn't disprove the effect. It depends upon uncertainty. Eventually the end date stops being an estimate and becomes an actual date. There is no reason to suppose the interval will shrink politely when this happens. I may remain five weeks away until some entirely ordinary Tuesday afternoon when certainty finally arrives and I discover that I am, in fact, moving in nine days. The intervening four weeks have not passed; they have simply ceased to exist. The Fixed Interval Effect doesn't necessarily decay. It collapses, leaving you abruptly deposited back into conventional time rather closer to the event than conventional time would consider reasonable.

So the Fixed Interval Effect is not ended by time passing, nor apparently by progress towards the thing itself. It ends when uncertainty does.

In my case, the conserved interval happens to be five weeks. I have already spent considerably longer than five weeks being five weeks away.

Five weeks is not five fucking weeks.