Changing your mind in a hurry.

Changing your mind in a hurry

In the microcosmic world governed by quantum mechanics, the observer doing experiments has an important, but not yet precisely defined, effect on the reality of the object being observed. As Carroll O. Alley of the University of Maryland at College Park puts it, "We have a role in creating." Specifically, if an experimenter sets up to look for the wave nature of light, that side of light's dual nature will show itself, but not the presumably equally existent particle nature. Conversely, an experimenter looking for photons -- the particulate aspect of light's being -- will see them but not waves. One question that arises is wether it makes any difference when the experimenter chooses what to look for: long before the apparatus is built, or while the light is on the way through it.

At least three "delayed choice" experiments, which test what happens if the experimenter does not choose until the light is moving through the apparatus, have been done. Alley reported on one conducted with his student Oleg G. Jakubowicz. A group from the University of Munich and the Max Planck Institute for Quantum Optics in Garching, West Germany -- T. Hellmuth, Arthur G. Zajonc and Herbert Walter--did the other two.

The Maryland experiment and one of the Garching experiments involve Mach-Zehnder interferometers, devices that take an incoming light pulse from a laser, split it in two with a half-silvered mirror, send the two halves over different paths and recombine them. This is a standard test of the wave nature of light: A wave splits in two, and when it is recombined it interferes -- that is, its brightness reinforces or cancels depending on whether the two half-beams are still in phase or not. To test the particle nature of light the experimenter leaves out a mirror that recombines the beams and puts particle detons, at the ends of the two paths. These record light as photons, and also tell over which path a given photon came through the apparatus. The delayed-choice part is to switch this second mirror in and out while a single laser pulse is traversing the apparatus. The switch flips in 1 nanosecond.

One way of interpreting the wave-interference situation is to say that a single photon took both paths at once (a very mysterious thing, as a particle shouldn't be able to do that). In the words of John A. Wheeler of the University of Texas at Austin, who inspired these experiments, the switching of the mirror then "decides whether the photon 'shall have come by one route or by both routes' after it has 'already done its travel.'"

The second Garching experiment is what the experimenters call a "quantum beat experiment." Laser pulses of 553 nanometers wavelength and 1.5 picoseconds duration excite a barium atom. They send it into a superposition of states in which it exists in two excited energy states at once. These states are linked together in a way analogous to the way wave and particle natures are linked.

The excited atom immediately reemits some light, and because of the superposition or linkage, the wavelengths that each of the two states might emit separately are combined in an interference or beat signal like two sound waves beating together. With a polarizer, an experimenter can analyze this signal so as to separate the two wavelengths, but then the result is all one or all the other.

The whole process amounts to a photon leaving the laser and being absorbed and reemitted by the atom through either one of the linked excited states (no beat signal) or both of them at the same time (the beat signal). Switching the polarizer in and out after the reemission determines which mode appears.

So far, all three of these experiments support the conventional quantum wisdom that whether you make the choice before or after the event occurs, the effect of the choice is the same.