This question already has an answer here:

A single photon travelling within a single wavelength contains a dual nature, in that it can behave as a particle or a wave depending upon the chosen experiment or measure.

When the duality behaves as a particle, is it true to say that at the point of measurement or detection that the wave-form of the duality has collapsed, that the energy within the wave is now confined to a particle?

Or is it the case that the wave and particle form of the duality remain in tact and that the measurement is simply confined to either form?

Two examples I am thinking of are the photoelectric effect and the detection of photons within the double slit experiment.


marked as duplicate by John Rennie, Jon Custer, stafusa, Kyle Kanos, glS Dec 26 '17 at 23:50

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • 2
    $\begingroup$ You may be relying too heavily on the word "duality". In the current understanding, there is only one object, the field, which obeys a wave equation, and whose excitations are quantized. So it fills space like a wave, but interactions occur in quanta that transfer energy and momentum. Nothing ever changes from one form to another, and nothing ever collapses. $\endgroup$ – garyp Dec 26 '17 at 0:30

A photon does not "collapse" into a particle or a wave depending on which measurement you make. That is specifically a property of wavefunctions, which probabilistically describe the location of the photon. When people talk about wave-particle duality, they are instead talking about the fact that the photon acts like an EM wave, which deals with the distribution of the photon's energy.

As such, I have found it best to think about the wave and particle properties of a photon as two separate, but ever-present attributes of the object. (As an analogy, perhaps think of cars. All cars have a color and a speed. When you measure the speed, you don't change the color, and vise versa. Sometimes one property is more useful to measure than the other). So, in the photoelectric effect, we are most interested in investigating the particle-like properties of the photons generated, so we sometimes ignore the wave-like properties, but they are still there!


Quantum mechanics describes reality as dynamics of abstract vectors in complex vector space with defined norm, which is called by mathematicians, Hilbert space. Every thing you can measure, called observables, are functions acting on elements of this space - vectors. Such functions acting on vectors in linear way, so we call them operators. Given vector, describing some physical state of the system, operator acting on such vector gives you another vector, transforming oryginal one. In some way it looks like rotation. If you want to know what is the result of the measure, you have to choose observable, the craft operator which gives you measurement results, and act operation on vector of the physical state of the system. Then you will get result - a vector describing system after measurement. But what about values of measured things? There's any value. There's only another vector. Why? Because typical ordinary quantum system does not have any certain value of any observable!

But you may say, some of the vectors are special. If you are able to prepare quantum system in a state of that particular vector, some observable, let's say velocity, may have certain value. Other observables usually not, but this one, and any related to it, like momentum, would have.

So for typical quantum system you cannot get any certain value, but for some artificially crafted, you will.

As vector space usually has property it have basis, you can use such crafted vectors as your basis. Typical space of quantum system has infinite dimensionality, because there's a lot of possible velocities. But it is just a technical problem.

In that case you may do the following: you start with your system, described by some, abstract vector in this vast space. But now you are able to decompose this vector into sum of elements of your basis. Remember, every basis vector has defined velocity ( and undefined rest of possible observables, but you do not care for a moment, because you want to know velocity). So sum you would get is large. In typical case it contains many basis vectors with different velocities, every of it with some coefficient. Then you act observable operator on it, one by one, and after that, some coefficients will change. Then you take a look and see, some velocities have bigger coefficients than others. You can check it just performing normal projection in direction of this or that basics vector.

You still don't know what velocity your system have! But is responsible to expect that velocities with bigger coefficients will dominate, and in fact those coefficients measure probability of your measurement gain.

But then you changed your mind, and want to know position instead. Oh, it is just simple: you have to change your vector space basis. Instead using vectors with defined velocity, you would like to use another ones with defined position. Well, as everything comes at price, single vector with defined velocity, usually have to be defined as infinite sum of vectors with definite position. And vice versa.

In fact everything you do in quantum mechanics is just exchange one vector space basis with another and checking of coefficients of such decompositions.

Duality? This is just fancy name for such vector space basis equivalence.

  • $\begingroup$ Hilbert space contains an infinite number of possible dimensions. Within such a space anything is possible and any particular view of reality can be justified or imagined. "Vector space" is an oxymoron and is not helpful. $\endgroup$ – Marcus de Brun Dec 26 '17 at 21:39
  • $\begingroup$ It is precisely defined mathematical object. It is functional analysis term, so infinite dimensionality means only it consists of infinite number of functions, which are vectors in this case. They may be represented in different ways, as ordinary function on coordinate system, describing position of the system, or as Fourier transform in frequency domain in which case they contain information about energy and momentum. This is exactly basics change I described. $\endgroup$ – kakaz Dec 26 '17 at 21:46
  • $\begingroup$ Anything is possible here has precisely defined meaning: there is large number of possible positions, or momentums. This is exactly what QM is about... $\endgroup$ – kakaz Dec 26 '17 at 21:50

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