“How Many Lives has the Higgs Boson?”

The New York Times is running an article about one quirky explanation for why we haven’t found a Higgs boson yet.

A pair of otherwise distinguished physicists [Holger Bech Nielsen and Masao Ninomiya] have suggested that the hypothesized Higgs boson, which physicists hope to produce with the collider, might be so abhorrent to nature that its creation would ripple backward through time and stop the collider before it could make one, like a time traveler who goes back in time to kill his grandfather.

The picture should be familiar enough from Lewis’s discussion of time travel. What would happen in a time travel world if someone tried to go back in time to kill their grandfather? They’d fail. They might fail in surprising ways. At this point of the story banana peels start to play a prominent role, leading to one of the better paper titles of our time.

But as Nielsen acknowledges in the article, it isn’t quite so clear why creating the Higgs boson would be akin to killing one’s grandfather. And at this point the analogy with time travel starts to look strained.

If we’re doing speculative physics on the basis of Lewisian philosophy, I think we should be looking not at Lewis on time travel, but Lewis on QM. In How Many Lives has Schrödinger’s Cat?, Lewis discusses what it feels like to be the inhabitant of a world where the Everett Hypothesis is true. And his conclusion is that inhabitants of such a world will appear to have, from their perspective, a surprising number of near-death experiences.

The picture is that whenever we get near to death, there’s a possible evolution of the world in which we don’t die. (That tumour quantum tunnels its way out of the patient’s body, etc.) Now from an external perspective, such bizarre possibilities are generally ignorable quirks. But from the perspective of the agent nearing death, things look quite different. They don’t experience anything after they die. So all of the possibilities in which they keep experiencing are ones in which they survive. So if the only way to survive a time period t is through a series of events that are incredibly improbable, all of the agent’s future experiences will be in worlds where those events happen. So despite the improbability, the agent should expect those events to happen, in the sense that those events will probably be part of the only future the agent knows.

Now suppose that creating a Higgs boson (or more precisely isolating one) will lead to the instantaneous destruction of the universe. (Actually we just need that it leads to the death of everyone on earth.) And suppose the Everett hypothesis is true. Then unless the creation of a Higgs boson has probability 1, there are some really existing futures in which the Higgs is not created. And we should expect (with probability 1) that we’ll find ourselves in one of them. So we should expect (with probability 1) that the attempt to create a Higgs boson will appear to fail.

If future attempts to create a Higgs boson fail in more and more unlikely ways, I think we’ll have to start taking this hypothesis seriously. So far we’ve had two attempts to create the Higgs boson fail in mildly surprising ways. I think we need to get to more like 8-10 failures before we start worrying about the hypotheses sketched here, but if the improbable failures start accumulating, the lessons of How Many Lives has Schrödinger’s Cat? start to look more and more pressing.

Hopefully they’ll get the LHC back working soon and we can put crazy ideas like these to bed!

11 Replies to ““How Many Lives has the Higgs Boson?””

  1. For people, like me, who are under-informed as to what a Higgs boson is, it’s worth mentioning that In Our Time (the BBC4 radio program, which is in general highly recommendable) had an episode on these topics not long ago. The Higgs boson gets explained towards the end, but the rest of the discussion is also quite enlightening. Hear it here:


    - the webpage also has useful links to further reading etc.

  2. Don’t forget the (very surprising) failure of the SSC. That counts for at least another failure or two. My credence in many-worlds plus collider-kills has been boosted by at least a factor of 1000 or so from its prior. If this goes on, I’m not sure whether we should be mourning all our dead counterparts or worrying about all the deaths that face us in the near future. But I suppose that we should at least be grateful to Everett for the survival mechanism.

  3. The moral psychology of the Everett hypothesis (should we be mourning, worrying etc) is kind of fascinating actually. If I know measure 0.99 of my ‘branches’ are going to be trimmed, I don’t think I feel particularly worried about it.

    Indeed, the ethics appropriate to agents in an Everett world are tricky.

    If A sets up a device that has a 0.8 chance of killing B, has A done anything wrong? You might think no – after all, B will see that the device doesn’t work. But that’s too simplistic. B’s friends will see him killed in 80% of the branches, and what A does is terrible to them.

    But what then if A sets up the device to kill all of B’s friends, and all of their friends, and so on as well? Then has A harmed anyone? It’s kind of hard to say.

  4. It gets better. Suppose A’s device produces Higgs bosons when the temperature of B’s beer (wine cellar, planet) rises above comfortable levels. It is tempting to think that A has done B a big favour, by guaranteeing him a comfortable future, with no effort on his part. (This is Euan Squires’ quantum Russian roulette.)

    The nice thing about the Higgs boson, on your hypothesis, is that it seems to be solving Lewis’ tails problem, by disfavouring branches in which survival is unpleasant.

  5. If Higgs bosons are so nefarious, surely they wouldn’t confine their conspiratorial powers of temporal sabotage to just physical events—clearly we can’t form consistent ideas about them and can’t see that inconsistency since that would guarantee that no apparatus could be thus physically constructed that would produce them. Not just many-worlds then, but many-minds…

  6. Lewis’s argument rests on (1), no?

    1. I should spread my expectations for what will be the case at t exclusively over possibilities in which I have experiences at t.

    That seems false to me, even if the Everett Hypothesis is true. Of course, (2) is true:

    2. I should spread my expectations for what I will experience at t exclusively over possibilities in which I have experiences at t.

    But (2) is trivial. (2) is equivalent to

    3. I should spread my conditional expectations Cr(P given that I have experiences) for what will be the case at time t exclusively over possibilities P in which I have experiences at t.

    And (3) is not equivalent to (1).

    Anyway, that’s what I’ve always thought about Lewis’s paper, but I’ve never spoken to anyone about it. Am I missing something?

  7. I quote from an email about this from my mate John Price, who says

    My point is just that Lewises point is that more probably we find ourself living on the higher intensity branches – but not on branches where we are dead. So for any crunch, after crunch time we may find ourself alive on a branch that from a statistical point of view, looks very low-intensity. Barring coincidences, that’s when on the high intensity branches we are dead, and (for us) they are culled and we are here —->.

    (John continues): I see no way that this process will retrofit any culling of branches to the past? Branches on which CERN persists with LHC and branches on which it quits are for the time being both still branches on most of which you are alive, I hope? Nu?

  8. What would happen in a time travel world if someone tried to go back in time to kill their grandfather? They’d fail.

    Hmm. I have that you (i.e. Tim) can and cannot kill his grandfather, depending on the facts that you take to be relevant. Some of those are compossible with Tim’s success, some aren’t; you can reasonably choose either one.

Leave a Reply