Physics bans cloning

PHILIP BALL

Attack of the Clones is classically implausible. © Lucasfilm

Perfect clones can't exist, say physicists. They're not doing down the hottest topic in biology, merely pointing out that the laws of classical physics forbid making an exact copy of an object, living or inanimate, just as the laws of quantum physics have been known to do for 20 years.

Scientists have created replicas of individual quantum particles such as atoms and photons with properties almost identical to those of the original. Last month, a team at the University of Oxford, UK, even showed that they could clone a photon - a particle of light¹.

But quantum cloning is only feasible if one accepts a little bit of copying error - as two US physicists proved in 1982 with their 'no-cloning' theorem.

This theorem means that quantum teleportation - recreating a quantum particle in a different location from the original, but in exactly the same state - is only possible if the original is destroyed, so that it does not coexist with its copy. Quantum teleportation of photons was demonstrated experimentally in 1997².

But the laws that govern quantum particles are different from those that apply to everyday systems, which are made up of countless billions of such particles. No one had given much thought to ideas such as exact cloning and teleportation on macroscopic scales, where Newton's laws of classical physics apply.

Until now. Daniel Collins and Sandu Popescu at the University of Bristol, UK, have shown that the quantum property of entanglement has a classical analogue³. Entangled quantum particles have mutually dependent states; they seem to interact by way of an instantaneous action at a distance.

Entanglement is central to quantum teleportation and cloning. Collins and Popescu show that systems that are governed by classical physics can exhibit secret correlations in their behaviour that can give rise to effects rather like teleportation.

And Andreas Daffertshofer of the Free University in Amsterdam and colleagues have proved that there is a kind of no-cloning theorem for classical systems⁴.

The key is that classical, many-particle systems are characterized by statistical distributions. The particles are distributed between a huge number of possible states, like grains of wind-blown sand distributed over a hillside.

Daffertshofer's team shows that it is impossible to copy any arbitrary statistical distribution of one many-particle system onto another to make an exact clone. Attempts to do so will disturb the original distribution. There is a sole possible exception: classical systems in which all particles occupy the same state, like sand grains stacked in a single tiny well.

Biological clones are nothing like these idealized versions. The statistical distributions of particles in an organism and its clone are inevitably very different - so physicists would not accept them as true clones at all.

1. Lamas-Linares, A., Simon, C., Howell, J. C. & Bouwmeester, D. Experimental quantum cloning of single photons. Science 296, 712 - 714 (2002).|Homepage|
2. Bouwmeester et al. Experimental quantum teleportation. Nature 390, 575, (1997).|Article|
3. Collins, D. & Popescu, S. A classical analogue of entanglement. Preprint, arXiv:quant-ph/0107082 (2001).|Article|
4. Daffertshofer, A., Plastino, A. R. & Plastino, A. Classical no-cloning theorem. Physical Review Letters 88, 210601, (2002).|Article|

© Nature News Service / Macmillan Magazines Ltd 2003