Stein asked:

I was wondering if someone has read Michael Dummett’s ‘Bringing about the Past’ and actually understood his arguments. I am not a newbie to philosophy, but I thought this article was quite hard to grasp. I was first presented to this article 5 years ago, when attending a course in epistemology/philosophy of science at the University of Oslo (a bachelor’s degree). Since then I’ve been attempting to understand it without success. And I don’t know where to look for a nontechnical explanation of his article either a (long enough) summary, or a stepwise explanation. Every response to the article ‘Bringing about the Past’ either is way too technical or picks out some specific feature. And that, I feel, is no help.

I ‘understand’ the beginning of the article what he means by our prejudice when it comes to the future and the past, but it’s only because the asymmetry idea is too deeply ingrained in us, that we have to (or better for argument’s sake, try to) free ourselves of our prejudices about asymmetry. But from here and on I get COMPLETELY lost I cannot reproduce his argument(s). Don’T know if my lack of understanding hinges on my interpretation of the phrase ‘bring about’ which I take to mean ’cause to take place’ (implying replacing whatever event/circumstance was already present in the past). I can see no other good interpretation of the phrase ‘bring about’. You, of course, have the special case where no event/circumstance already was in the past, in which case ‘bring about’ would imply replacing nothing, i.e. replacing/filling up some empty time/space stretch with some event/circumstance.

Taking ‘bring about’ to mean just what I did above implies quite strange results (but actually not more strange than ridding oneself of the conception of asymmetry between future/past). One could ask (a natural question for any rational person): how can one generalise this? If this is not a special case of prayer and this particular tribe that Michael uses, then it must apply to everything? We could pick literally any example and play with it. For instance I could (in principle) cause myself not to be born in the past, which would yield the strange (but no more strange than annulling the concept of asymmetry between past/present) result of me disappearing right now in the present (or perhaps like in ‘Back to the Future’, where the body disappears in a stepwise fashion right before the protagonist’s (Michael J Fox’s) eyes.

Or I could kill Michael Dummett when he was alive before he produced this article, which in turn would remove the CAUSE for me sitting here and wondering what the hell Michael (Dummett) was saying.

Since this conclusion seems TOO absurd to me, I am inclined to think that I made the wrong assumption somewhere along the line, or that my reasoning just went astray (or it seems that this is how philosophers argue if a conclusion disagrees with intuition, then the conclusion goes overboard, not the intuition). But I still maintain the importance of understanding this article as it must have caused a lot of havoc when it was published. And if it caused havoc (I know there exists many responses to this article), it must be because the article made an impression by bringing up something profound.

Anyway, I have attempted to explain to the best of my ability what I wonder about and hope that someone could throw some light on the issue.

Answer by Craig Skinner

Where you go wrong is in thinking that the past could be CHANGED, that there could be different versions of the past. There can only be one version, what actually happened, and that’s it, done and dusted. Any actions by time travellers from the future are already built into past events. Thus if Jack Flash, a time traveller from the 28th century went back to the 14th century and signed the Magna Carta, his signature is on the document, for all to see, now and at all times since he signed it all those years ago. And you can’t go back and kill Dummett: he wasn’t killed, he lived to a ripe old age. Nor could you go back and prevent your birth: your birth wasn’t prevented. You might go back and try to do these things , but you wouldn’t succeed because they happened despite your efforts.

Now, to Dummett. He is notoriously opaque. He writes for other philosophers, not for beginners, and doesn’t spell out his underlying assumptions or views (his informed readers know what they are).

His view of the past is as I have spelled out, but he doesn’t make this clear.

His article makes four points:

1. We experience an asymmetry between past and future. You have got this point. We are entropy-producing creatures, and live our lives along the direction of increasing entropy (into the future, since the universe started out with very low entropy which can only increase). We experience causation in this direction, we experience time in this direction, remembering the past and planning for the future. In short, the entropic, temporal, causal and psychological arrows all point in the same direction (past to future).

2. Although we experience causation as antegrade (past to future), retrograde or backward causation (future to past) is logically possible, and may sometimes occur (some interpretations of quantum mechanics imply this). I think you can see this point, provided you cut out the idea of changing the past, and stick to the notion of causing or ‘bringing about’ the past (by future action).

3. Experiment can’t prove or disprove retrograde causation.

You might think it would be simple to disprove the matter by what is called ‘bilking’. This means: find out whether the relevant past event occurred or not. If it did, take steps to prevent the alleged future cause. If, contrariwise, the past event didn’t happen, take steps to ensure that the future alleged causal action does occur. In this way the disconnect between alleged cause and effect would be shown.

But Dummett, using his example of the chief dancing (the cause) to ensure that the tribesmen were brave some days earlier (the effect), shows that matters are less straightforward.

Take the case where it has been reported to us that the tribesmen didn’t act bravely. We try to ensure that the chief dances. But:

* sometimes the chief can’t dance (he has broken his ankle, he is in bed with a fever), so he can hold on to his view.

* sometimes the chief does dance, accepts that the men were not brave, but just says well, it doesn’t always work, it’s probabilistic rather than deterministic causation.

* sometimes the chief does dance, but it turns out that the report of the men’s cowardice is mistaken (communication error, or somebody trying to impugn the tribe’s good name, or somebody trying to undermine the chief’s credibility). Now, of course, it is open to the chief to say, whenever his dancing seems to have failed, that the report of the men’s cowardice is in error.

A combination of these three ensures that the chief can plausibly hold on to his view about retrograde causation.

4. Even if a reliable association between dancing and past bravery were shown, we needn’t accept backward causation. We can say that the bravery caused the dancing. Of course we would struggle to think of a mechanism by which this happens. But here we are no worse off than on the retrograde assumption where we struggle to explain the mechanism by which dancing causes past bravery. As Hume famously said, we only ever observe a conjunction, never the connection. Hence modern ideas of causation which strive to establish the connection by defining the cause as what we need to manipulate to produce a change in something else (the effect) e.g. altering smoking habits in a population changes the later incidence of lung cancer, hence smoking is a cause of lung cancer.

I hope things are clearer to you.