In the context of the double-slit interference experiment, what exactly does “observation” mean? Specifically, does the act of visual observation by intelligent creatures, such as humans, play a role in determining whether light behaves as a particle or a wave?

I’m trying to understand how the concept of observation impacts the behavior of light in this experiment. How is observation defined in the quantum mechanical sense, and what role does it play in the interference pattern observed on the screen? Does merely looking at the experiment change the outcome, or is there a more technical aspect to what “observation” entails in this scenario?

Any detailed explanations or references to relevant quantum mechanics principles would be greatly appreciated!

Intelligence has nothing to do with it. “Classical” has a lot to do with it. Let me explain.

A “classical” entity: a human, a cat, a brick, a video camera, has well-defined states. A brick is either on the table or on the floor, never both. A human is either moving or standing still, never both. A cat is either alive or dead, never both. (Putting aside pointless philosophical musings about cats being, after all, “quantum” since they consist of a very large but finite number of uncorrelated quantum degrees of freedom… true, but for all practical intents and purposes, bricks, cats, etc., are classical. Any quantum behavior is averaged out to the extent that it vanishes.)

A “quantum” entity is different. A quantum entity does not have a classically defined position or momentum. Rather, it has a state. That state, which may be described (depending on your mathematical preference) by a vector in an abstract space, or by a fancy wavefunction, can be used to calculate the probability of finding that quantum entity with a certain value for its position or momentum when it interacts with something classical. Let me explain how.

What the equations actually tell us is that the state of the quantum entity evolves such that it is constrained by its past, present, and future interactions, including interactions with classical things such as bricks, cats or humans. Quantum physics is manifestly nonlocal (this is the essence of Bell’s famous theorem) so it is subject to constraints that can be arbitrarily far in space or time, yet curiously, somewhat counterintuitively, it remains causal (it is not possible to arrange the future to induce an observable influence on the past).

A quantum entity is neither a particle nor a wave, it just displays behaviors of both under the right circumstances. The state of that quantum entity obeys a wave equation including the ability to exhibit interference patterns. But when a quantum entity interacts with a classical thing, the interaction may be localized, creating the appearance of something resembling a miniature cannonball, i.e., a particle.

Concerning the double-slit experiment, the quantum entity may be a photon or an electron (so let me just call it an electron from now on), which is confined by three things: a) the location and time of emission; b) the slits; and c) the location and time of detection. These are the constraints that determine the wavefunction that characterizes this electron’s state.

One possible interpretation, obviously inspired by the uncomfortable consequence of quantum nonlocality, namely that the location and time of detection is one of the constraints characterizing the system, is to take that constraint out of the picture initially. We then describe the state of an electron that is characterized by the location and time of emission, and the presence of the slits. The resulting wavefunction can be used as a probability amplitude, characterizing the probabilities of possible locations and times for the electron’s detection. When we actually do detect the electron, the picture changes: We now have a new wavefunction that is also constrained by the location and time of detection. This rather abrupt replacement of one wavefunction with another we call “wavefunction collapse”.

Now it is true that this interpretation requires an intelligent creature, such as a human who prefers this probabilistic interpretation (the Copenhagen interpretation). But apart from some (not generally accepted) extensions of the quantum theory that treat wavefunction collapse as a physical process, the collapse is only a clumsy mental aid. It is a philosophical musing, not physical reality. In reality, that electron was always governed by a wavefunction subject to nonlocal constraints (again I stress, this is the fundamental essence of Bell’s theorem.)

So to sum up, a quantum entity (photon, electron, etc.) is neither a particle nor a wave; its state is governed by a wave equation so it sometimes exhibits the consequences of wave-like behavior; its interactions may be highly localized, giving the impression of a particle; and intelligent observers have nothing to do with this unless we buy into the controversial concept of objective collapse (of the wavefunction), and perhaps not even then.