Steorn's free energy seems curiously expensive

In spring, a young man's fancy lightly turns to thoughts of love. In late summer, though, it's more a case around here of what on earth to write about. It is the journalistic silly season, when everyone's on holiday and nothing's really happening until September.

Canny people know how to fill that space. One lot who got the timing just right is Steorn, a company apparently composed of three people working from a business park in Dublin. Flicking through the seasonally adjusted pages of an anaemic edition of The Economist last week, I saw their full page advert claiming a "blasphemous" breakthrough in energy generation. Thence to their Web site, which is a creditable production saying the company has a small bundle of aluminium, motors, disks and wires that effectively produces power out of nowhere. Interested scientists are invited to apply to become part of a panel of 12, which will then be asked to test the device.

Coo. And Steorn is putting its money where its mouth is. A full-page advert in the Economist costs many tens of thousands of pounds. The Web site is very professional, and the London PR company involved is one that also handles ITV, Halifax, John Lewis and others of that stature. This is a substantial investment — and, since it doesn't seem aimed at selling anything, inviting investment or producing anything measurable, it's a huge chunk of their own money in what even the company will cheerfully admit is a PR stunt.

It is also pseudoscience of the highest order. The general idea has been around for a while and has spawned many impassioned claims: you spin magnets around in a clever way and get more energy out from a system than you put in. This is generally agreed as impossible: it's perpetual motion, it breaks the laws of thermodynamics, and in the long and gaudy history of pseudoscience it ain't never worked yet. Which is not to say it never will: science is full of astounding discoveries that turn the accepted truths on their head. History is also full of total balderdash masquerading as science.

Fortunately, there are easy ways to tell pseudoscience: grand claims with no way to verify them, important facts that are alluded to and not presented, claims of conspiracy or closed-mindedness by the scientific community, production of claims by press release rather than scientific papers. Steorn more than fulfils all of these: it is, by any objective test, pseudoscience.

So what on earth are they playing at? In a long and very impassioned phone call with Steorn's chief executive Sean McCarthy, I had some theories flatly denied and others half-confirmed. It is not a teaser for an Xbox game It is nothing to do with a TV programme It has nothing to do with promoting anti-fraud systems (Steorn's corporate history is in detecting and preventing high-tech fraud), which was my personal favourite.

The official story — and one they are at pains to emphasise — is that the idea of convening a panel of 12 top scientists to do secret tests is the best way they can think of to get their ideas accepted by the scientific community. Time after time, McCarthy said, they'd tried to get people to look at what they were doing, but nobody was prepared do so. Those who did refused to go on the record.

None of this makes sense. Here's why.

There are two sorts of scientific discovery: the predicted and the unpredicted. Predicted is great: you have a theory, you come up with some physical ramification of that theory...

...you go and look and, if you find what you're looking for, your theory is all the stronger. Unpredicted science is even better, when you see something happen, go "Hold on a sec..." and try and work out why. If you can't work that out using the very best science of the time, then you could be onto something.

Chances are, you're not. You've missed experimental error, an aspect of existing theory, even a straightforward misinterpretation of results. To catch this, you put out a paper saying: "I've found this, and I think it's important". Others — not as attached to your discovery as yourself — then check what you've done. They repeat the experiment, but not to reproduce the results: instead, the idea is to pick holes in your logic, find problems, show why it's not important after all. Get past that stage and then and only then is it deemed good enough to build further theories.

McCarthy says Steorn won't do that bit, and won't really explain why. Or rather, he says it will — but in private with hand-picked scientists, because it wants to have a really good explanation of how it works before it's prepared to demonstrate that it does. Despite returning to this point repeatedly with him, I could never quite fathom that. Put the plans up on the site, I said. What would it hurt? Multiple independent demonstrations are difficult to refute.

The conversation ended up with him saying in effect: "This is how we've chosen to do it". There was talk of intellectual property and patents, yet if you want to patent something you keep quiet about it, not take out full-page ads in the Economist. And basic physics has no IP anyway — although I'm sure somebody in America is currently trying to patent the photon.

And thus, despite their claims to the contrary on their website, it's pseudoscience. They don't have to have a theory for it to be science — in fact, it's clear that they don't — but that doesn't matter. Build a box. Show it working. That's enough. Hans Christian Ørsted showed electromagnetism in 1820 and it wasn't fully explained until James Clerk Maxwell's masterwork nearly 50 years later — but it was all science.

A working journalist gets a nose for this sort of thing. I've had a similar experience with a company called SilkRoad, which had an amazing change-the-laws-of-physics optical data transmission method. I've had one with a British data compression outfit set up by a university professor. I've been assured of unbelievable wireless transmission methods that will change the world. Grand claims with no evidence sink. By comparison, look at Intel, HP, IBM and Hitachi: amazing results through the scientific method that flow through to real products.

Whatever Steorn is doing — and in the utter absence of any testable data, the chances of it being a significant scientific achievement are closer to absolute zero than the contents of Lord Kelvin's freezer compartment — it's an expensive experiment. For the price of that Economist advert and whatever they're paying their PR company, they could have built 10 apparatuses that actually demonstrated their effect and Fedex'd them to the major centres of scientific excellence on the planet.

It wouldn't even cost them that much. If they'll send me the plans, I'll build one. Having built it, I'll convince myself that it does produce more energy than it takes in — which will take a glass of water, a resistor, a thermometer, a couple of test meters and some basic mathematics, all of which I already have. I shall then get on the train to Cambridge and refuse to leave until the nice people at the Cavendish take a look at it.

I shall do all this at no charge to Steorn, because it will make me very famous if it turns out to be true and I'll get a great article out of it if it isn't. Furthermore, I don't think it will happen.

Neither will Steorn's amazing machine. Whether it's being driven by madness, genuine misapprehension or some ulterior motive yet to be revealed, it's not being driven by science. Producing rotational energy out of nothing is a great trick and one that its PR company is clearly very good at, but once the silly season's over the spin will die down.