When Chesley Sullenberger III saved the lives of all 155 people aboard his airliner with a breathtaking landing on the Hudson River last month, he won worldwide acclaim. Within hours, the pilot's achievement had been recognised by a dedicated website, and he found himself the focus of international media attention. But the media weren't alone in picking up on what has been dubbed the "Miracle on the Hudson". Scientists monitoring a network of electronic devices spanning the world have found that the devices also picked up some kind of signal just as Sullenberger's jet ran into trouble after take-off.
Normally the devices churn out a steady stream of random electronic noise. But just after 3pm New York time on Jan 15, their output started to move away from the usual meaningless jittering. Such a shift could have happened completely by chance, but analysis suggests it was statistically significant. So what could be responsible? One possibility is some kind of freak electromagnetic event - but this seems unlikely to have affected a network of more than 60 devices shielded from such interference and located from Europe to Japan and Australia.
In any case, it can't explain the disturbing fact that this is far from the first time that the devices have picked up some kind of global disturbance. Over the last 10 years, many similar shifts have appeared at times of great drama, including the Twin Towers attack of September 11 and the Bali bombings of 2005. Despite the low collective odds, most scientists would probably still dismiss all these anomalies as nothing but flukes. But according to the team operating the Global Consciousness Project (GCP) from Princeton, New Jersey, the network of random signal generators may be giving us the best evidence yet for psychokinesis (PK): the ability of the mind to influence matter.
The GCP is the most sophisticated attempt yet to prove the existence of PK. Until now, much of the evidence has been anecdotal, ranging from claims of furniture moving during séances to the famous "spoon-bending" demonstrations of Uri Geller. Some of the anecdotes have impressed even hard-nosed scientists. In 1970, a Russian woman named Nina Kulagina sparked a brief thaw in the Cold War when Soviet scientists invited their western counterparts to Russia to watch her demonstrate PK abilities, such as moving small objects simply by staring at them. Sceptics insisted Kulagina was using standard conjuring methods like ultra-fine threads or magnets, but the allegations were never substantiated.
The aim of the GCP is to avoid such sterile sniping by steadily accumulating hard evidence for PK that no one can challenge. Events likely to produce an anomalous signal are identified, and the output of the GCP network examined for signs of the event being picked up. This long, slow approach towards confirmation has an impressive pedigree: many of today's most successful drugs were only recognised after years of studies involving thousands of people.
The GCP is not the first to try to do the same for PK. Laboratory studies began in the mid-1930s, and investigated whether people could compel dice to land on pre-specified faces more often than expected by chance. Over the next half century, around 150 studies were published, with most finding positive results. But as with drug studies, many of the studies were so small that their success could easily be dismissed as flukes. Sceptics also suggested other explanations, including the "file-drawer effect", where researchers only publish positive results, leaving the negative ones languishing in filing cabinets.
Even so, a 1989 review of the evidence pointed to a small but highly statistically significant effect consistent with PK - and one hard to dismiss as a file-drawer effect. The overall size of the effect was hardly spectacular, however, amounting to barely one per cent higher than the chance hit-rate. One obvious explanation is that the PK effect is simply too feeble to produce dramatic effects with dice. That has prompted interest in using more sensitive detectors - like devices that generate random signals by subatomic processes.
As with the dice experiment, reviews of the evidence suggest there is something going on - but again the effect is small, at barely one per cent above random chance. The idea behind the GCP is to boost the size of the effect by focusing on events which might trigger especially large PK signals. Since 1998, when its first devices were set running, the GCP network has detected more than 270 cases of anomalous signals appearing at times of emotionally charged events. Taken individually, the results aren't especially impressive, but when combined the result is spectacular, with the odds against chance producing such striking results exceeding 10 million to 1.
Yet on closer inspection, the results raise a host of questions. Why did the terrible Indonesian earthquake and tsunami of December 2004 fail to trigger a signal on the GCP network, while a smaller event that hit the same area six months later produced a highly significant signal? Come to that, why did Capt Sullenberger's heroic act, in which not a single life was lost, generate a clear signal while the July 2000 Concorde crash in Paris, which killed more than 100, produced nothing?
One possibility is that only certain individuals involved in such events are responsible for producing the whole signal. The trouble is, no one has any idea how PK might work. Yet without some mechanism, it's impossible to know if an event that fails to register should count against the reality of PK - or if a signal that does appear is just a false positive. The only conclusion to emerge from the Global Consciousness Project so far is that data without a theory is as meaningless as words without a narrative.
Robert Matthews is Visiting Reader in Science at Aston University, Birmingham, England.