31
$\begingroup$

In inflationary cosmology, primordial quantum fluctuations in the process of inflation are considered responsible for the asymmetry and lumpiness of the universe that was shaped. However, according to the Copenhagen interpretation, any random quantum phenomenon only occurs when the system is observed; before observation, the quantum state is symmetric. So the question is, who has observed the universe while it was inflating? Obviously, there was no conscious creature that time.

Actually, this problem is discussed in the paper The Bohmian Approach to the Problems of Cosmological Quantum Fluctuations (Goldstein, Struyve and Tumulka; arXiv:1508.01017), and the proposed solution to the problem in said to be an observer-independent interpretation (the pilot-wave theory).

$\endgroup$
  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ – rob Jan 13 at 6:01

11 Answers 11

68
$\begingroup$

“Observe” oftentimes causes a lot of confusion for this exact reason. It doesn’t actually refer to some conscious entity making an observation.

Rather, think about how we actually make an observation about something. You have to interact with the system in some way. This can be through the exchange of photons, for example. This interaction is what constitutes an observation having taken place.

Obviously, particles can undergo their fundamental interactions without a nearby sentient entity.

For the sake of analogy, consider measuring air pressure in a tire. In the process of doing so, you let out some air — changing the tire pressure in the process.

$\endgroup$
  • 8
    $\begingroup$ According to the Copenhagen interpretation, the wavefunction collapse is "subjective", it DOES depend on the act of observation, and doesn't have anything to do with physical and experimental imperfections like the exchange of photon. If you say that the wavefunction can collapse automatically, you are advocating "objective-collapse" interpretations, which are proposed by some people like Roger Penrose, but are not mainstream. $\endgroup$ – Ali Lavasani Jan 10 at 22:07
  • 59
    $\begingroup$ You have a misunderstanding. Collapse of the wave function in the Copenhagen interpretation is caused by a thermodynamically irreversible interaction with a classical environment. I agree — it does depend on the act of observation. Observation just doesn’t mean what you think it does. $\endgroup$ – Riley Scott Jacob Jan 10 at 22:10
  • 9
    $\begingroup$ @AliLavasani The point is that exchange of photon is not an experimental imperfection, but that observation without it is impossible, so that it is a fundamental part of the observation. I believe I have been told that the realization of 'you can't observe something without interacting with it and hence changing it' is what led Bohr to the Copenhagen interpretation but maybe I misremember. Others here are more knowledgeable on this history. $\endgroup$ – Vincent Jan 11 at 9:53
  • 7
    $\begingroup$ Aren't observations interactions only between a "classical system" and a quantum system? If we don't divide the Universe in different system, but describe it by a single state, there wouldn't be any observations. $\endgroup$ – jinawee Jan 11 at 15:17
  • 8
    $\begingroup$ The von Neumann interpretation requires consciousness, the Copenhagen interpretation only requires a measurement. But what constitutes such a measurement remains unclear and opinions on it are divided. en.wikipedia.org/wiki/… $\endgroup$ – Michal Paszkiewicz Jan 11 at 16:27
28
$\begingroup$

The Copenhagen interpretation isn't an essential part of quantum mechanics. It isn't required in order to make physical processes happen. It's just a way of describing what seems to happen when an observer makes a measurement. It's not even the only way of describing what it seems like to the observer.

However, according to the Copenhagen interpretation, any random quantum phenomenon only occurs when the system is observed; [...]

If you don't use the Copenhagen interpretation, quantum mechanics still works fine. In your example of the early universe, all the quantum-mechanical processes work in the same way. E.g., a hydrogen atom in an $n=3$ state will radiate light, and at a later time it will be in a superposition of $n=2$ and $n=1$. No randomness, just a superposition.

[...] before observation, the quantum state is symmetric.

I'm not sure what you mean by symmetric here. This seems like a nonstandard description.

$\endgroup$
  • 7
    $\begingroup$ You say there has been a "superposition" of all possible outcomes in the inflation, so what has destroyed the superposition? In Copenhagen, ONLY observation can collapse the superposition. If you believe it has automatically collapsed, you are defending "objective collapse" interpretations, and another option is that the universe is still in superposition (the many worlds interpretation). Either way you are implying one of these two kinds of interpretations, aren't you? $\endgroup$ – Ali Lavasani Jan 11 at 0:20
  • 5
    $\begingroup$ @Wolphram Yes, any interpretation other that Copenhagen has no problem. Copenhagen shouldn't also fail, so my question is how the observation can have been done at the beginning of the universe. I don't know, maybe observation is done NOW when we look at the universe!! $\endgroup$ – Ali Lavasani Jan 11 at 1:34
  • 7
    $\begingroup$ @Wolphram Notice that Copenhagen perfectly works. Interpretations like Bohmian mechanics have more serious problems (nonlocality, or retrocausality in transactional interpretation, etc). $\endgroup$ – Ali Lavasani Jan 11 at 1:46
  • 4
    $\begingroup$ @AliLavasani how about the good old many-worlds interpretation? No wavefunction collapse, no issue choosing when it happens. The only thing that you lose is the notion that only the universe that you see is what exists. It's not that far-fetched either. An electron created by the collision of two gamma photons can only "see" a positron that flies away in the exact opposite direction, which is just a layman's way of saying that a superposition of states evolves the same as if you evolve each state separately and only then sum up the states. $\endgroup$ – John Dvorak Jan 11 at 15:14
  • 4
    $\begingroup$ @John In many worlds, you have the problem that you cannot interpret probability for your "worlds". For example, suppose the probability of quantum event A is 0.7, and the probability of quantum event B is 0.3. What does this mean? Does it mean you have 7 universes in which A happens and 3 ones in which B happens, or what? $\endgroup$ – Ali Lavasani Jan 11 at 15:33
16
$\begingroup$

"Observation" is not about a human actually viewing and consciously perceiving a system. If one state is capable of affecting another state, then the latter is said to be measuring, or observing, the former. The reason conscious observation also constitutes measurement is simply because interaction with the environment is fundamentally necessary for our eyes to be able to perceive an event.

$\endgroup$
10
$\begingroup$

The Copenhagen interpretation is nothing but an impediment to understanding quantum mechanics. There is no such thing as "wave function collapse" within the system described by QM, nor in any falsifiable physical sense outside of the theory. At best it's an artificial glue for sticking quantum and classical models together; less flatteringly it's a mental crutch for people who don't want to accept that the best model of physical reality we can hope for describes not the evolution of a single deterministic state, but rather the deterministic evolution of a probability model of possible observed states.

Ultimately what's attributed to "wave function collapse" from an act of observation is just conditional probabilities, or if you want to go even more basic, correlations between random variables. I like to explain this via analogies with other applications of conditional probability, and usually end up picking something morbid like cause of death. As a random member of a general population, you have some $X$ percent chance of dying of a particular disease. If you get DNA tests done, you might find out that you instead have a $Y$ percent chance of dying from it, where $Y$ is greater or less than $X$. No physical change took place when you had the test done to change the likelihood of dying from that particular disease. Rather, you're just able to make better predictions based on correlations.

Now, neither QM nor any other physical theory is going to tell us much about what fine-grained observations could have been made in the very early universe, because the correlations to anything we can observe are going to be too small. But that doesn't mean the probability model didn't evolve the same way then as it does now, with all the consequences that entails.

$\endgroup$
  • 7
    $\begingroup$ It is true that observing a particle's position can be thought of as just acquiring information about its position. However, unlike the case with the disease, you can't take this to mean that the particle really had a definite, though unknown position the entire time -- such a hidden variable theory just doesn't work. That's the real subtlety of quantum mechanics, the puzzle that led people to adopt the Copenhagen interpretation in the first place. If you simply ignore it, you're not really talking about quantum mechanics at all. $\endgroup$ – knzhou Jan 11 at 17:58
  • 4
    $\begingroup$ Instead, you're just repeating what you already know about classical probability and hoping it all transfers effortlessly to quantum mechanics. The problem is, experiment tells us it doesn't. Quantum mechanics and classical mechanics are different things. $\endgroup$ – knzhou Jan 11 at 18:00
  • 1
    $\begingroup$ @R.. In most QM classes you're told that after a measurement associated to an operator "A" which leads to result "a", the state "collapses" to $|a\rangle$. I'd like to see same formulation with classical correlations. $\endgroup$ – jinawee Jan 11 at 18:44
  • 3
    $\begingroup$ @knzhou: I didn't assert that there is any possible hidden variable theory; quite the opposite, that QM only describes the evolution of a probability model, not of some "real state" among the elements of the probability space. Everything I've said above is roughly equivalent to how you view QM through the MWI, but without its ontological baggage which is problematic just like the CI, but for different reasons. $\endgroup$ – R.. Jan 11 at 21:30
  • 1
    $\begingroup$ @Menno: Nothing requires "probabilities cancelling each other out". The fact that QM describes the evolution of a probability model (determined by the wave function) is not at all controversial. Interference is one consequence of the rules for how the probability model evolves. I'm not clear what you're actually objecting to. $\endgroup$ – R.. Jan 13 at 2:04
7
$\begingroup$

If only an act of observation by a conscious (whatever it means) creature could cause a wavefunction to collapse, then it would be impossible in the first place for conscious creatures to develop in the course of history because the entire Universe would be in a continuously developing superposition of states without any collapse taking place (collapse is a necessary condition for conscious creatures to develop). Which means that conscious creatures making an observation aren't the cause for the collapse (and nor can conscious creatures now cause the collapse at the beginning of the Universe retroactively because conscious creatures couldn't have developed if the collapse is caused by them). So when inflation took place, no conscious creatures were needed to make a wavefunction collapse, and as you stated in your question, obviously there were no conscious creatures (if the collapse is caused by "a thermodynamically irreversible interaction with a classical environment" then by the same token, neither a classical environment will be able to develop).

This means, for example, that the pattern of lines (resulting from the collapse of a whole lot of wavefunctions corresponding to photons) appearing on the screen in the double slit experiment will develop independently of some conscious creature observing the setup.

This doesn't necessarily mean though that an observer(creature)-independent interpretation is one that postulates a pilot wave (or hidden variables). The "inherently probabilistic" interpretation will do as well. Both can make a wavefunction collapse without an observer. I think which interpretation corresponds to reality will remain unknown (unless someone comes up with an experiment to make a decision which I find hard to imagine) and be a question of "taste". Einstein was an advocate for a theory that underlies the apparent probabilistic behavior of matter ("Gott würfelt nicht", that is, "God doesn't play dice"), as a theory of hidden variables does (somewhat like the molecules surrounding a Brownian particle make the particle move in an apparent random way). But many others (like Bohr in the "famous" Bohr-Einstein debate) take an opposite stand.

$\endgroup$
6
$\begingroup$

For an interpretation of quantum mechanics that requires "conscious observers", you can assign our present-day astronomers that role. Certainly their observations are not done at the time of the early universe itself. That's just fine. No problem if you observe 15 billion years after the fact.

The problem only exists if you insist that observations must be done simultaneous with the observed phenomenon. But simultaneity has no place in physics, such a requirement would be at variance with basic physics (relativity). Quantum mechanics does not use simultaneity, and does not prescribe when observations must be made.

$\endgroup$
  • 1
    $\begingroup$ This is I think the correct answer ! Whether observer dependant or not, these interpretations can all account for the observed facts as long as they don't rely on some "time of observation" or "time of collapse". It is the case of both the Copenhagen and the Bohmian interpretation, so same results. $\endgroup$ – Undead Jan 14 at 5:18
6
$\begingroup$

If the Copenhagen interpretation is correct(unknown), and if it requires conscious observers(unknown), our observations of the universe could retroactively collapse the superpositions. https://en.wikipedia.org/wiki/Delayed-choice_quantum_eraser .

$\endgroup$
  • $\begingroup$ This seems to mesh with some work by Hawking I was reading a while back that suggested that the present influences the early parameters of the universe. $\endgroup$ – Michael Jan 11 at 19:34
4
$\begingroup$

Observation does not mean "by a human". Observation is any action on the system by outside of the system. Photons interacting, the confines of the system being changed, etc.

Your comment above about superposition "automatically collapsing in the early universe" is wrong. A hydrogen atom with superposition of it's energy level will collapse when the value of it's energy level is needed (e.g in a physical collision) which counts as an observation. The main takeaway is that when we say observation we mean interaction with a clearly defined outcome.

$\endgroup$
4
$\begingroup$

The problem with this question is that it assumes there is some metaphysical interpretation that we can be sure is true. While we have excellent equations that work incredibly precisely, we are not sure which qualitative interpretation of these equations is real.

There are now countless interpretations, each with their own sub-interpretations. Alexander R. Pruss splits these interpretations into two main groups - No collapse theories with a deterministically evolving wavefunctions and wavefunction collapse theories.

Out of the collapse theories, we have the Copenhagen Interpretation, where the wavefunction collapse is triggered by a measurement. Definitions of what constitutes a measurement can differ a lot depending on the physicist/philosopher. The Ghirardi-Rimini-Weber theory is another collapse theory where the collapse is triggered at some particular rate over time. The trouble with this theory is that no spontaneous collapse has been observed in any way, and an additional parameter - that of the rate of collapse - has to be introduced and explained in some way.

There are also many no collapse theories such as Bohmian Mechanics, the Many Worlds Interpretation, Many Minds Interpretation and Traveling Forms interpretation. In these, the universe continues to develop deterministically, but each have their own reasons as to why we can only get stochastic results from the deterministic systems upon measurement. Each of these interpretations also have their own problems. Bohmian Mechanics has the problem of nonlocality. The Many Worlds Interpretation is unclear as to how splits occur and is a bit bizarre to try to reconcile with, for example, the conservation of energy. The Many Minds interpretation leads to bizarre absurdities such as Boltzmann Minds and universes where there is just one mind surrounded by zombies. I don't think the Traveling Forms is well enough known to have its own critique, but I expect someone will come up with one at some point.

I found an excellent study of this topic in this book: http://www.michalpaszkiewicz.co.uk/blog/reviewnapocs/index.html

$\endgroup$
  • 1
    $\begingroup$ I don't find it problematic to ask how a popular interpretation is compatible with a specific phenomenon. Your overview of other interpretations is actually nice, but that's not asked here? $\endgroup$ – M. Stern Jan 11 at 22:19
  • $\begingroup$ Thanks for the comment - the OP mentioned 2 different interpretations (Copenhagen and Bohmian) and also didn't specify that there was a particular desire for an answer for the Copenhagen interpretation - so I wasn't sure of how many interpretations the OP was aware of and thought it needed a more generalised answer. $\endgroup$ – Michal Paszkiewicz Jan 12 at 10:15
1
$\begingroup$

As others have mentioned, your definition of observer seems to have mislead you.

Take the double slit experiment for instance. In this case, the observer which forces the wave function to collapse is the screen, not the person looking at the screen. The results would be the same without a person looking at the screen.

$\endgroup$
  • $\begingroup$ So how large does this screen need to be in order to count as an "observer"? What if you isolate the screen very well, would you still call it an observation? This approach has some problems if you think about it. There are similar problems with highly upvoted answers, to be fair... $\endgroup$ – M. Stern Jan 15 at 17:41
0
$\begingroup$

It's an interesting question...
I find your question interesting, because of the following thoughts, where I'm trying to stick with the spirit of the Copenhagen Interpretation: The collapse of the wavefunction describes the change of knowledge of an observer. The fact whether an observer is able to get the information matters (quantum eraser). Whether a collapse happens or not can make a difference, think of quantum computation or zeno effect. So us looking at the universe now might well be the reason for a collapse of the wavefunction back then. It's a bit weird in the sense that we can only observe because of the collapse, a funny causal relation.

..with no answer
That said, your question involves aspects of general relativity and of quantum theory. As of now, we just don't have a theory for this situation, so anything might have happened. At least quantum theory probably does not apply.

$\endgroup$

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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