Why won’t physicists leave these poor cats alone?! Is it dead? Alive? Both?

As if the ideas about where quantum mechanics melds into classical physics weren’t already confusing enough, new experimental work seems to confirm the theory that the transition from multiple possibilities to a single observed outcome is far from instantaneous and is, in some cases, reversible. In my previous post I described the old standard quantum mechanics view of the instantaneous collapse of possibilities at the moment of observation—but the modern view is that this is far from the whole story. It turns out that these observations, or measurements in physics parlance, can vary in “strength” depending on how much information they give us about the state of the system.

To illustrate this point, I’ll appeal to a famous thought experiment put forth by a skeptical Erwin Schrödinger, now known as *Schrödinger’s Cat*: imagine an experiment where an ordinary housecat is placed in a large steel box, along with what Schrödinger called a “diabolical” device consisting of a single atom of a radioactive substance, a detector, and a phial of poison gas. When the atom decays radioactively the detector goes off and smashes the phial, killing the cat. The box is sealed, and the experimenters/cat detractors wait until the probability that the atom has decayed is one half.

N.B. depending on the species of atom, they could be waiting anywhere from nanoseconds to several billion years.

They then ask the question, how do we describe the state of the cat? Classically, we’d say that the cat is either alive or dead, with 50% chance each. Quantum mechanics says that since we haven’t opened the box and observed the cat yet, it is in an equal “superposition” of alive and dead, which is to say, some weird quantum state of both at the same time. Since when opening the box, the state “collapses” to either alive or dead, you might ask what possible difference could it make how you *describe* the cat?, half the time it will be alive, half the time dead.

This is where things get tricksey. Imagine that, instead of opening the box, I put my ear to the side and listen. Cats meow occasionally, when they’re alive, and don’t when they’re dead. So if I hear a meow, then I can safely assume that the cat is alive. But, what if I hear *nothing*? I don’t know for sure that the cat is dead, since it might have simply chosen not to meow while I was listening. However, the fact that I didn’t hear anything still gives me some information about what’s happening in the box. In quantum mechanics, this is called a *weak measurement*. I now expect to be *more likely* to find a dead cat in the box by virtue of my measurement of hearing nothing. If I had a well characterized cat, which meows randomly at a certain average rate, then I could calculate new probabilities of the cat being alive or dead.

This is, in effect, what experimenters at UC Santa Barbara have done. Their “cat” is a loop of superconductor whose electrical properties can be in one of two different states, labeled 0 and 1 for this discussion. They prepare the loop in a superposition of 0 and 1 using a pulse of radio waves. They then perform a weak measurement of the system, where if the state is 1, they have some probability to detect it, like the cat meowing if it is alive. If the state is 0, then they will detect nothing. They then perform a full “strong” measurement, which detects either 0 or 1 (equivalent to opening the box). For the runs where they detected nothing during the weak measurement, they were more likely to get a 0 during the final measurement.

So far, so good, but this is where it gets interesting. They then added a second radio pulse to their procedure after the weak measurement. The effect of this second pulse is to swap states 0 and 1, so if the cat was dead, it would get reincarnated, but if alive, it would be killed (and here the prospects for doing this with actual cats probably end). They then performed *the same weak measurement* on the loop after the swap. For the runs where both weak measurements detected nothing, the final state was exactly the original, equal superposition of 0 and 1!

The weak measurements only *partially* collapsed the quantum state, leaving it in a state which was not an equal superposition, but which could be reversibly “uncollapsed” by performing the same weak measurement on the other state. We used to think, according to the “standard” interpretation of quantum mechanics that the measurement induced collapse was both instantaneous and irreversible, but new experiments in this realm are forcing us to reconsider!

– – – –

The original paper, published to the arXiv. Read this if you’re a physicist, or are hardcore.

Nature News article about this research. This is behind a paywall. Oxford people should read it fine if you’re on campus or the VPN. Others, ask your respective institutions/employers. Read this if you’re a scientist.

Science Daily article about this research. Read this if you’re interested, but don’t have a science background.