In a parallel universe, this theory would make sense

Yet another of my pens has just disappeared from where I swear I just left it and is probably already with my smug doppelganger in a parallel universe. We all have our favourite take on the existence of parallel worlds; it's a subject that has been fodder for science-fiction writers for quite a while now. The question is whether the idea has a place in serious scientific discourse.Well, it seems the only known recordings of a physicist who predicted the existence of parallel universes have been found in his rockstar son's basement. The tapes document how quantum physicist, Hugh Everett III, developed a remarkable idea at the age of 24 while a graduate student at Princeton in 1957. The recordings are believed to have been made at a conference 20 years later, and were discovered during the making of a documentary in which Mark Everett, the physicist's son and singer of the band Eels, attempts to understand the work that consumed his father.

Those more astute followers of science news may have also become understandably excited at the recent news that, finally, there might be a way of proving the existence of Everett's universes and reclaiming all those lost pens. One online headline read: "Parallel universes really do exist, according to a mathematical discovery by Oxford scientists." Another quoted a respected US physicist as saying: "This work will go down as one of the most important in the history of science." Good grief.

So is this for real? Or is it just sensationalising journalism? What we should first ask is what place the theory of parallel universes has in real physics? So allow me to lead you through gently.

Many cosmologists are quite keen on an idea that requires the existence of other realities as an anthropic argument. That is, in order to explain why our universe seems to be just right for humans we need simply appeal to the obvious: that our universe is not unique. There must be an infinite number of universes in a much larger multiverse and we happen to be in the one that by definition had to be just right. While this is a very attractive idea, I find it rather lazy. In fact, I would argue that it is no more satisfying than creationism. But the notion of parallel universes has a more serious raison d'etre; one that was introduced to science exactly half a century ago.

Quantum mechanics - the theory that describes the workings of the subatomic world - is based on many counterintuitive ideas. This is not news. Most physicists learn to live with the notion of a subatomic particle being in two places at once. We put up with this because quantum mechanics, as a mathematical framework for describing how atoms behave, is so successful. Indeed, I would argue that it is the most important theory in the whole of science - beating Darwin with one hand behind its back, since it explains why our universe is the way it is. Indeed, this year's Nobel prize was awarded for the discovery of something called giant magneto-resistance, a neat quantum effect that has led directly to the technology that allows for more than one song to be stored on your iPod. But there is this frustrating weirdness at the heart of quantum mechanics that won't go away.

It is difficult to translate what is essentially advanced mathematics into words that make sense. Another way of putting it is that while the mathematical formalism of quantum mechanics is not in doubt, nobody has yet found a satisfactory explanation of what it means that is agreeable to everyone. Should this not be a cause for concern? Many physicists feel strongly that there is no problem. They would argue that drawing attention to the conflicts between different interpretations is unnecessary.

But there are ways of explaining quantum weirdness, albeit with yet more weirdness. One such interpretation as to how subatomic particles can do more than one thing at once is for there to be more than one universe. After first being introduced by Everett, this idea has had a steadily growing minority of supporters. Everett proposed what became known as the "many worlds" interpretation of quantum mechanics, which has since come to be regarded as both the most extravagant explanation of reality, and yet the most simple, depending on which side you are on. The basic idea is as follows: whenever a quantum particle is faced with a choice of alternatives (as happens everywhere all the time), it doesn't choose one but rather all. It is only when we look that we force the little blighter to make up its mind. This is not even the controversial bit; we know this happens. But in the many-worlds version the particle never makes up its mind. Instead, it, and the universe along with it, splits into multiple versions of itself equal to the number of options available. When we look at the particle, the universes separate into non-interacting independent realities. We see one version, but our identical counterparts see another.

There was little initial support for Everett's idea. He became disillusioned and left physics to work for the US defence industry - where he made a fortune working out how to maximise kill rates during a nuclear war. I should point out that this is not the natural career progression of quantum physicists.

Oxford physicist David Deutsch, one of the founding fathers of the exciting field of quantum computing, later proposed a variation on Everett's idea in which all possible universes already exist within a quantum multiverse. What we perceive as our reality is just a weaving through this vast shadowy multiple reality, creating our own version of events. While supporters of the multiverse interpretation argue that it is the most sensible explanation, the majority of physicists are sceptical, mainly since it is essentially unprovable.

This brings me finally to the recent headlines. It turns out that there is no proof that the multiverse exists, but rather that one of the main objections to it has been removed by an argument in logic and algebra that has yet to appear in any peer-reviewed scientific journal.

Of course I understand scientists being keen to publicise their research, and that sometimes journalists see a story too juicy to ignore. But for now, let's take parallel universes with a pinch of salt. And if we are going to explain science, let us at least ensure we use the scientific method, with its reliance on peer review and reproducibility. And give up hope of ever getting those pens back.

Jim Al-Khalili, professor of physics and of the public engagement in science at the University of Surrey. He is this year's recipient of the Royal Society Michael Faraday prize for science communication.