First, let me state that I'm a lot less experienced with physics than most people here. Quantum mechanics was as far as I got and that was about 9 years ago, with no use in the meantime.

A lot of people seem to think that the act of observing the universe we are changing the universe. This is the standard woo hand waving produced by people who don't understand the physics -- thank you Fritjof Capra.

In reading about quantum physics out there on the Internet, I find this idea is propagated far too often. Often enough that I question whether my understanding is accurate.

My questions:

  1. Does observing a particle always mean hitting it with something to collapse the wave function?

  2. Is there any other type of observation?

  3. What do physicists mean when they say "observe"?

Related, but doesn't get to the question of what observation is: What is the difference between a measurement and any other interaction in quantum mechanics?


6 Answers 6


Assuming that the incoming "first" particle is prepared in a pure state, interaction with another particle does seem necessary. Such an interaction might simply be the spontaneous emission of a photon or other particle by the original incoming particle, however.

Most importantly, such an interaction is not itself sufficient. For a measurement event to occur (wave function collapse in the Von Neumann formalism) we must also "physically lose track" of the some of the information of the interacting particle after the interaction has taken place, so that we must replace the entangled state description of the second particle after the interaction with a probabilistic mixture of such states, forcing a description of the first particle after the interaction in terms of a real valued probability density matrix rather than as the complex valued pure state amplitude we started with. This change of description automatically includes an increase in entropy, which also occurs physically.

Unless the second, interacting, particle either escapes the apparatus or interacts with a third particle which so escapes, i.e. "interacts with the environment", no measurement has yet occurred, the entire interaction is in principle reversible, and the complex amplitude description remains appropriate. Measurement requires "loss" (via decoherence) of the entangling information by further entangling with the environment and dissipation.

The escaping third particle is often an emitted photon or phonon. See the reference in the linked answer What is the difference between a measurement and any other interaction in quantum mechanics?, particularly the 1939 article by London and Bauer (but avoid their metaphysics) for details. More recently, see this book on quantum measurement theory, particularly page 102 referring to the view of Zeh.

You may have noticed that some ambiguity remains in this descriptipn. This has been analyzed in great detail and resolved by Zurek, but it gets a little tricky. See e.g. http://arxiv.org/abs/1001.3419 and references therein.


"Collapse the waveform" is a loaded term, that would not be agreed to by all physicists. There are a great many "no-collapse" interpretations out there in which there is no special role for measurement that directly alters the wavefunction. There are also collapse-type interpretations in which the collapse happens more or less spontaneously, as in Roger Penrose's theory whereby gravitational effects cause any superposition above a certain mass threshold to collapse incredibly quickly.

As a practical matter, it's hard to think of a measurement technique that couldn't be described as hitting one particle with another. Most quantum optical experiments rely on scattering light off an atom in order to detect the state of the atom, many charged-particle experiments involve running the particles into a surface or a wire in order to detect it, and so on. I think that the solid-state qubit experiment done by people like Rob Schoelkopf at Yale would probably count as an exception, because I believe they use a SQUID to detect the state of their artificial atoms via magnetic fields. If you want to get really picky, though, you could probably consider that a particle interaction as well, though, in some QED sense.

Even there, though, the act of measurement does not leave the initial system unchanged. While there would not be general agreement with the specific phrasing "observation changes the universe," the idea that quantum systems behave differently after a measurement is central to the theory, and can't be avoided.


The thing is there are a variety of different opinions that, since they can not be distinguished by experiment, are around and used by different people to interpret experiments.

The conventional view of quantum mechanics, although it has eroded over time, is that a sharp disctinction has to be made between the classical and the quantum. The apparatus has to be described classicaly, while the quantum describes the measurement results of the experiment. Von Neumann has then tried to show that the distinction need not be sharp and that you can include the apparatus in the quantum description, but it then has to be observed itself by another apparatus which has to be described classically. Wigner argued that this regression of the quantum/classical divide can be carried up to the mind, hence why there is a lot of woo latching onto these ideas, because they seem to justify the importance of the human mind over anything else in the world.

Other approaches have argued that there is no distinction between classical and quantum, at least. One is the Many Worlds Interpretation, which states that a superposition of states is representing actually realized states but in different universes. Another is the Bohmian or pilotwave interpretation, which states that the wave equation is describing a wave that guides particles. An extra equation is then supplemented to the Schrödinger equation to show how this guiding happens. In both theories, there is no need to speak about measurement, at least, not in any deeper sense than what would be done in classical physics.

Here's a non-exhaustive list of interpretations of quantum mechanics

So, within the context of the Copenhagen interpretation and the von Neumann/Wigner paradigm, the answer to the title question would be yes, there is a difference between measurement and hitting with another particle. Within the context of Bohmian or MWI interpretations, the answer would be no.


How do you know the output of an experiment? You "observe" it.

How do you actually do that? Well, with your eyes.

How does that work? Photons emitted or reflected by a surface hit special molecules in your eyes. This leads to a signal which is transmitted into the rest of your brain.

So the actual observation eventually happens by exchanging photons. Now you want to observe really small particles. Particles that are so small that interaction with a photon actually changes them.

Which begs the question whether you can examine something like an electron without hitting it with a photon or anything that might change it's state. This is really hard. So in most cases, observing something does influence it.

There are a couple of tricks like observing something that is by itself influenced by the particle you want to measure (say the electric field of a moving electron could influence another molecule and you could do your measurement on the molecule).

But in the end, all these processes are based on resonance. If you want to measure anything, you must create a resonance of some kind and that always means two-way interaction at some level.


This question reminds me of the Zen koan:

What is the sound of one hand?

According to this site (which also elaborates on the koan) if:

while walking, standing, sitting, and reclining, you proceed straightforwardly without interruption in the study of this koan, you will suddenly pluck out the karmic root of birth and death and break down the cave of ignorance.

... and/or understand quantum mechanics :-) Ok ... maybe not quite well enough for a degree, but you get my point ;)


all particles are the sums or products of other particle interactions, however higher energy collisions are required for the threshold of observations to be made.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.