Read The Best Australian Science Writing 2012 for Free Online

time but at different locations, say flashing a torch, then depending on how you are travelling relative to each of those events, you may see them occurring at different times. For instance, if you are accelerating relative to one then you will see it occur later, as if time is slowed.
    Now let’s have a machine that turns on two torches at the same time. If the first of those is a normal torch and the second emits faster-than-light luminous particles, then from a distance you might think that the second torch was turned on first because you saw it first, even if you aren’t moving relatively – it’s as if it was turned on in the past. So the particles can appear to travel back in time (hence the backwards ordering of the joke about the neutrino entering a bar) … but there’s still no method of accelerating a cyborg killing machine to super-luminal speeds.
    Is the latest potential finding an isolated incident? Apparently not. In 2007 the MINOS experiment, located in the US and with a similar source-detector distance to the CERN-OPERA path, observed the same thing, albeit with a smaller significance(1.8 standard deviations – not enough to get excited about) and a larger error (allowing for neutrinos travelling exactly at c ). Measurements of arrival times of photons and neutrinos from supernova SN1987a in 1987 provided a much better agreement with the speed of light, but those neutrinos were of a much lower energy. The possibility remains that the velocity of a neutrino depends on its energy. Somewhat less rigid explanations include neutrinos taking ‘shortcuts’ through extra dimensions. Undoubtedly many more possible explanations will arise if all conventional sources of error are excluded.
    The much more likely scenario is that the analysis has overlooked some seemingly insignificant but critical aspect, and that re-analysis will lead to a very good agreement with the speed of light.
    Should that be the case, the follow-up press release will no doubt refer to the ‘Phantom of the OPERA’.
    Postscript: 1 May 2012
    For every complex phenomenon there is an explanation which is simple, concise, and wrong. The OPERA faster-than-light neutrino article generated a flurry of explanations, both online and within the scientific community, claiming to resolve the problem via strange new mechanisms. In the end, a loose coupling in the GPS wires mentioned in the above article was found to be responsible for the apparent velocity increase.
    Even though the OPERA team was agnostic to sources of error, and announced their findings expecting that they would be proven wrong, the fallout has been dramatic. The two OPERA scientists in charge of the collaborative experiment, spokesperson Antonio Ereditato and physics coordinator Dario Autiero, resigned from their posts following a vote of ‘no confidence’. The timing of the signal travelling between the surface GPS and the underground detector was measured originally by Autiero in2008, but was not rechecked recently. Doing so would likely have implicated the loose connection and forced a re-evaluation of the spurious data before making it public.
    Since then, the ICARUS experiment (also based at Gran Sasso) has confirmed the neutrino speed as matching the speed of light to very high precision, a confirmation that would have also prevented the premature publication of the OPERA results.
    The widely varying mechanisms invoked to explain how neutrinos could travel faster than light will no doubt be shelved until the next anomaly. In an ideal world the vanishing of such a discrepancy would serve as a constraint on wacky theories, but in reality the goal posts are too often moved to suit specific purposes. Experimental errors are an ever-present ‘unknown’ in any facility, and events like this should leave us even more impressed with the enormous precision and sensitivity of experimental efforts like the LHC.
    More importantly, we