Abstract
Nonlocality is the most characteristic feature of quantum mechanics, but recent research seems to suggest the possible existence of nonlocal correlations stronger than those predicted by theory. This raises the question of whether nature is in fact more nonlocal than expected from quantum theory or, alternatively, whether there could be an as yet undiscovered principle limiting the strength of nonlocal correlations. Here, I review some of the recent directions in the intensive theoretical effort to answer this question.
Main
Quantum mechanics is, without any doubt, our best theory of nature. Apart from gravity, quantum mechanics explains virtually all known phenomena, from the structure of atoms, the rules of chemistry and properties of condensed matter to nuclear structure and the physics of elementary particles. And it does all this to an unprecedented level of accuracy. Yet, almost 80 years since its discovery, there is a general consensus that we still lack a deep understanding of quantum mechanics. Indeed, novel, puzzling and even paradoxical situations are frequently discovered. And I'm not talking about the wellknown interpretational puzzles related to the measurement problem, but about a variety of quantum effects, from the Aharonov–Bohm effect^{1}, which was hidden in plain sight, to Bell inequality violations^{2}, to the multitude of strange effects related to entanglement and quantum information. They are all puzzling and paradoxical only because we do not yet have the intuition and understanding that would allow us to predict and expect them.
Surprisingly however, with very few notable exceptions, for many years research on the fundamental aspects of quantum mechanics was put on the back burner; there seemed to always be more important, pressing issues. During the past couple of years, however, there has been a strong renewed interest in the subject and there seems to be hope that we will finally reach a much deeper understanding of the nature of quantum mechanics. In what follows, I will describe a small part of this research.
As was noted long ago, the axioms of quantum mechanics are far less natural, intuitive and 'physical' than those of other theories, such as special relativity. Special relativity can be completely deduced from two axioms: (1) all inertial frames of reference are equivalent and (2) there is a finite maximum speed for propagations of signals. Contrast these with the very mathematical and physically obscure axioms of quantum mechanics: every state is a vector in a complex Hilbert space, every observable corresponds to a Hermitian operator acting on that Hilbert space, and so on. Furthermore, when trying to make physical statements about nature, they are all sort of negative: nature is uncertain, we cannot predict the result of a measurement, if we measure this we disturb that, and so on. Clearly there is no way to reconstruct the whole theory from such physical statements. Yet, there is a glimmer of hope. As both Aharonov (ref. 3 and personal communication) and Shimony^{4} independently noticed, the fundamental nondeterminism of quantum mechanics, one of the most unpleasant aspects of the theory and the very subject of Einstein's famous complaint “God doesn't play dice”, actually plays a positive role: it opens the window to a new phenomenon — nonlocality. And Aharonov even went a step further (ref. 3 and personal communication). He remarked that it is possible, in principle, to have a theory that is nondeterministic without being nonlocal. On the other hand, it is impossible to have a nonlocal theory that respects relativistic causality but is deterministic. Indeed, very roughly speaking, if by moving something here, something else instantaneously wiggles there, the only way in which this doesn't lead to instantaneous communication is if that 'wiggling thing' is uncertain and the wiggling can be only spotted a posteriori. The bottom line, therefore, is that if we take nonlocality to be the starting point, then fundamental nondeterminism — the most characteristic property of quantum mechanics — immediately follows as a consequence. Hence, we should consider nonlocality and not nondeterminism as a basic axiom of quantum mechanics.
In the years following Aharonov and Shimony's suggestion and due to the advent of quantum information and the extremely intense study of entanglement in particular, nonlocality came indeed to be appreciated as a fundamental property of nature. Yet, there is an even more interesting twist in the story. Rohrlich and I^{5} took the Aharonov–Shimony suggestion seriously and investigated whether or not quantum mechanics can be deduced from the axioms of (1) relativistic causality and (2) the existence of nonlocality. In other words, we asked: “Is quantum mechanics the unique theory that allows for nonlocal phenomena consistent with special relativity?” Surprisingly, we discovered that this is not the case: nature could be even more nonlocal than that quantum mechanics predicts, yet be fully consistent with relativity! This immediately raises two questions. Perhaps nature is indeed more nonlocal than is described in quantum mechanics says, but we haven't yet observed such a situation experimentally. Alternatively, if such stronger nonlocal correlations do not exist, why don't they? Is there any deep principle that allows for nonlocality but limits its strength? This Review is dedicated to reporting the very intense present research into this question.
Before going forward, I want to reiterate that the scope of this Review is, by necessity, very limited and what is presented here is only a small part of a much larger effort to understand the foundations of quantum mechanics that is going on at present. To start with, I would like to mention the intensive work in characterizing quantum nonlocality itself^{6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25}, where not even the simple algebraic question — of fundamental importance — of which nonlocal correlations can be obtained from quantum mechanics is yet completely solved; see seminal works by Tsirelson^{26,27,28,29} (Cirel'son) as well as others^{30,31}. Another interesting direction is that of generalized probabilistic theories^{32,33,34,35,36,37,38}. I also cannot cover the fascinating flow of ideas back from this research into quantum information theory, where it has led to a variety of new ideas, concepts and applications, out of which I would like to mention the newly emerged area of deviceindependent physics (including deviceindependent key distribution^{39,40,41,42,43,44,45,46,47,48,49} and deviceindependent randomness generation^{50,51,52,53,54,55,56,57,58,59,60}). A recent review article^{61} covers these results and many more in detail. Further afield, I would like to single out the intense activity in searching for natural axioms of quantum mechanics along the lines initiated by Hardy^{62,63,64,65,66,67,68,69}. Finally, I would like to mention a completely different type of nonlocality, namely dynamic nonlocality^{70}.
Modelindependent statements about physics
Physics is usually discussed in very concrete terms, indicating the systems of interest and the specific interactions between them. A very important recent development, however, was the realization that physics can also be presented in a 'model independent' way, that is, in a way that is largely independent of the details of the specific underlying theories; this allows one to compare various possible theories.
For our purpose, it is very convenient to view experiments as input–output blackbox devices. Every experiment can be viewed as a 'black box'. For example, suppose Alice has a box that accepts inputs x and yields outputs a (Fig. 1). One can imagine that inside the box there is an automated laboratory, containing particles, measuring devices, and so on. The laboratory is prearranged to perform some specific experiments; the input x simply indicates which experiment is to be performed. Suppose also that for every measurement we know in advance the set of the possible outcomes; the output a is simply a label that indicates which of the results has been obtained. In this framework, the entire physics is encapsulated in P(ax), the probability that output a occurs given that measurement x was made.
In our discussion, we are interested in the constraints that relativistic causality imposes on experiments carried out by two parties, Alice and Bob, who are situated far from each other. The physics is encapsulated in P(a,bx,y), the joint probability that Alice obtains a and Bob obtains b when Alice inputs x and Bob inputs y. We are interested in the case when the experiments of Alice and Bob are spacelike separated, that is, each experiment takes place before any information about the other's input and output could reach it. We allow, however, the boxes to have been prepared long in advance, so that they could have been prepared in some correlated way, and they may also be connected by radios, telephone cables and so on. Also, obviously, to determine the joint probability, we need time to collect the entire information in one place.
Nonlocality
Consider the simple case when x, y, a and b have only two possible values, conventionally denoted 0 and 1. Suppose that Alice and Bob would like to construct some boxes that will yield outputs a and b such that:
where ⊕ denotes addition modulo 2 (that is, a ⊕ b = 0 if a = 0 and b = 0 or a = 1 and b = 1 and a ⊕ b = 1 if a = 0 and b = 1 or a = 1 and b = 0). In simple terms, what the above equation says is that when the inputs are x = y = 1, the outputs must be different from each other, whereas for any other pair of inputs, the outputs must be equal to each other. The question is, how well can they succeed?
Suppose, without loss of generality, that Alice and Bob prearrange that if x = 0 then Alice's box yields a = 0. Now, to ensure that they win the game when the inputs are x = 0 and y = 0, they obviously must arrange that when y = 0 Bob's box should yield b = 0. Furthermore, to ensure success when x = 0 and y = 1, they must also arrange that when y = 1 Bob's box must also yield b = 0. Now, as Bob's box will yield b = 0 when y = 0, to ensure success if the inputs are x = 1 and y = 0, Alice's box must be such that it yields a = 0 when x = 1. But by now we have fixed the behaviour of both boxes for all the inputs. And we have a problem: if x = 1 and y = 1 the outputs will be a = 0 and b = 0, which constitutes a failure. Hence for one in four inputs, Alice and Bob fail. If the inputs x and y are given at random, 0 and 1 with equal probability, then Alice and Bob's probability of success is at most 3/4.
Of course, if the boxes could communicate with each other, then they could always succeed: Alice's box tells Bob's something like, “My input was x = 0, I output a = 0, take care what you do!”. But, the whole point of the setup was that Alice and Bob's experiments are spacelike separated from each other, so any such signal would have to propagate faster than light. The upper bound of 3/4 on the probability of success derives from 'locality' (that is, no superluminal communication between the boxes), and it is called a Bell inequality^{2}. There are many different Bell inequalities, describing constraints derived from locality in similar tasks; the particular one discussed here is the Clauser–Horne–Shimony–Holt (CHSH) inequality^{71}.
John Bell's seminal discovery^{2} was that if the boxes contain quantum particles prepared in an appropriate entangled quantum state, and if appropriate measurements are performed on them, one can arrange a situation such that the probability of success of the above game is larger than 3/4.
Quantum particles, therefore, somehow communicate with each other superluminally. One could wonder if this doesn't immediately contradict Einstein's relativity. Here is precisely where the probabilistic nature of quantum mechanics comes into play. All that Alice and Bob can immediately see are the probabilities of their experiments; to learn the joint probabilities takes time. Suppose, for example, that for Alice the outcomes a = 0 and a = 1 are equally probable, regardless of what x and y are. Then Bob has no way of signalling superluminally to Alice: all he can do is to choose the value of y, but this doesn't affect the probabilities of Alice's outcomes. Similarly, Alice could also be prevented from signalling to Bob. So, in Shimony's words, the probabilistic nature of quantum mechanics allows for the “peaceful coexistence of relativity and nonlocality”: the particles could communicate to each other superluminally, but the experimentalists cannot use them to communicate superluminally with each other.
Nonlocality beyond quantum mechanics
As discussed above, quantum mechanics allows for a probability of success larger than 3/4 in the correlation game, meaning that the boxes (or the particles contained within) somehow communicate superluminally with each other. This is now recognized as being one of the most important aspects of quantum mechanics. However, quantum mechanics cannot always win in the game — the quantum probability of success is at most (2 + √2)/4, as proved by Cirel'son^{29}. That this is the case is a simple consequence of the Hilbertspace structure of quantum mechanics. But the deeper question is, why?^{5} Is there a deep principle of nature that limits the amount of nonlocality?
The first guess was that stronger nonlocal correlations would be forbidden by relativistic causality; perhaps the randomness that provides the umbrella under which nonlocality can coexist with relativistic causality is not enough to allow for stronger nonlocality. So the very first question to ask is: could — theoretically — nonlocal correlations stronger than quantum mechanical ones exist, without violating relativity? When Bell discovered nonlocality, the problem was not formulated in a modelindependent way but by using the specific language of quantum mechanics: entangled quantum states, Hermitian operators, eigenvalues and so on. From this point of view, the very question of whether or not nonlocal correlations stronger than the quantum mechanical ones could exist was very difficult to even envisage, let alone to answer. In the above box framework, however, the question and its answer are almost trivial: as long as locally a and b are 0 or 1 with equal probability, there is nothing that prevents the game from being won with certainty. These particular correlations are now known as Popescu–Rohrlich (PR) boxes^{5,72}.
Superquantum correlations
The existence of superquantum nonlocal correlations shows that quantum mechanics cannot be deduced from the two axioms of (1) relativistic causality and (2) the existence of nonlocal correlations. Something else is needed. But what? What could be a supplementary, very natural, axiom that could rule out such correlations?
The statement that superquantum correlations could — in principle — exist is very far from a fully fledged physical theory. Therefore, it may seem very unlikely that one could make further progress in answering the above question before such a full theory, which could explain all the known results — hydrogen atoms and so on — but also incorporate superquantum correlations, is formulated. Surprisingly enough, it turns out that there is a lot one can do with even just the above particular example. Help came at first from computer science, and now this is one of the hottest areas in the foundations of physics. Various very interesting situations have been discussed, including communication complexity^{73,74}, nonlocal computation^{75}, information causality^{76}, macroscopic locality^{77}, local orthogonality^{78} and nonlocality swapping^{79}. In this Review, I will discuss only a few examples.
Communication redundancy
Almost all of our communication is redundant, and that is not only because some of us like to talk too much, but also because it is a law of nature. Indeed, consider the following problem. Suppose Alice and Bob would like to meet, but are both very busy. They speak on the telephone and try to find a day this year when they could meet. To make the problem more interesting, suppose that they do not want to find out a precise day, but first they want to establish whether the number of days when they could meet is even or odd (zero counting as even). To make the problem simpler, suppose it is only Bob that sends information to Alice, and Alice has to decide the result. The question is, how much information must Bob send to Alice?
We have now a problem in which the result is a single bit, a single yes or no answer: yes = even, no = odd. On the other hand, it is obvious that Bob needs to inform Alice about the status of each day of the year in his calendar. Indeed, one of the possible situations is that Alice is free only one single day. To decide whether they can meet or not, she has to know whether Bob is free that day; as Bob doesn't know anything about Alice's calendar, he has to tell her about each of his days. He has therefore to send Alice 365 bits of information, a 'yes = I'm free' or 'no = I'm not free' for each day of the year; all this for Alice to find out a single bit of information. Very redundant indeed.
Clearly, in the process Alice learns much more than what she wanted to know. Indeed, not only will she find out if the total number of days when they could meet is even or odd, but also she will know the precise days they can meet. She didn't want to learn that, but there is no other way.
Wim van Dam^{73} observed in his PhD thesis, however, that if Alice and Bob have access to PR boxes, they could reduce the communication to a single bit, eliminating therefore the entire redundancy. They can do this by not attempting to directly communicate information about their calendars, but using this as input to their boxes and communicating information about their outputs.
In particular, all Alice and Bob have to do is to associate with each day i a variable x_{i} (y_{i}) that is equal to 0 if the day is busy and to 1 if the day is free and use them as inputs for their PR boxes. The sum of their outputs is even (odd) if the number of days when they can meet is even (odd). For Alice to find out whether the sum of their outputs is even or odd, Bob only needs to inform her whether the sum of his outputs is even or odd, that is, a single bit of communication (Box 1).
The result is particularly important, as the above calendar problem is not just some silly communication task; it is in fact the most difficult communication task possible (technically called the 'inner product' problem). Indeed, every other communication problem can be mapped onto this one, so removing the redundancy from this calendar problem means removing the redundancy from all communication problems.
Crucially, quantum mechanical nonlocal correlations cannot help with this task^{80} (though they can help in easier communication problems^{81}), hence, they cannot eliminate all redundancy from communication. Quantum nonlocal correlations (Box 2) are therefore dramatically different from PR boxes.
Prompted by the above result, Brassard et al.^{74} raised a tantalizing possibility: maybe not only the perfect PR boxes, which are the strongest nonlocal correlations possible, but all superquantum nonlocal correlations could eliminate all redundancy from communication. If that were the case, it would single out quantum mechanics as the maximal nonlocal theory that doesn't make all communication efficient.
Brassard et al.^{74} took the first steps towards answering their question. Recall that quantum mechanical boxes can yield outputs a and b such that a ⊕ b = xy with a probability of success of at most (2 + √2)/4 ≈ 0.85, whereas perfect PR boxes have a probability of success of 1. Using errorcorrection techniques, they showed that even imperfect PR boxes can eliminate all communication redundancy, as long as their probability of success is larger than approximately 0.91. However, there is still a gap, from 0.85 to 0.91, about which we know nothing. Hence, we don't know yet if the task of eliminating communication redundancy can single out quantum mechanics.
Nonlocal computation
While the status of communication complexity (as the above general problem is technically known) versus quantum mechanics is yet unsettled, a different task, nonlocal computation^{75}, has for the first time singled out the quantum–superquantum transition.
Consider an ordinary computation in which the input consists of N bits, z_{1}, ..., z_{N} and the output is a single bit, c = f(z_{1}, ..., z_{N}). To this computation we can associate a 'nonlocal computation' in the following way. The computation is carried out by two devices, one at Alice's location and one at Bob's. To each bit z_{i} of the original computation we associate two bits, x_{i} given to Alice and y_{i} given to Bob, such that z_{i} = x_{i} ⊕ y_{i}. For each value of z_{i}, there are two possible combinations of x_{i} and y_{i}: x_{i} = 0, y_{i} = 0 and x_{i} = 1, y_{i} = 1 for z_{i} = 0 and x_{i} = 1, y_{i} = 0 and x_{i} = 0, y_{i} = 1 for z_{i} = 1. For a given value of z_{i}, each combination is selected with equal probability. As a consequence, by looking only at their own variables, neither Alice nor Bob can determine the original variables z_{i}. Alice is required to output a bit a and Bob a bit b such that a ⊕ b = c = f(z_{1}, ..., z_{N}). Alice and Bob know what the function f is and are allowed to communicate in advance and set up their devices in a correlated manner; they only don't know what the values of the inputs will be. The question is, how well can they succeed? More precisely, if the values of the original inputs z_{i} are chosen at random, what is the probability that the nonlocal computation gives the correct result?
An important notion is that of the 'best linear approximation' of a computation. To each function f(z_{1}, ...., z_{N}) we associate a linear function f_{L}(z_{1}, ..., z_{N}) = α_{1}z_{1} ⊕ α_{2}z_{2} ⊕ ... ⊕ α_{N}z_{N}, where α_{i} are constants equal to 0 or 1. In other words, f_{L} is the sum of a subset of the original variables. The function f_{L} is chosen in such a way that it is equal to f for as many inputs as possible. For example, if f is the logical AND function, that is, the product f^{AND}= z_{1}z_{2}, the best linear approximation is f_{L}^{AND} = 0. Indeed, by always yielding 0, f_{L}^{AND} = f^{AND} in 3 out of 4 cases, the exception being z_{1} = z_{2} = 1.
Nonlinearity is the core of computation, so, in some sense, a linear approximation means no computation at all. Now, it turns out that if Alice and Bob have at their disposal only devices functioning according to the laws of classical physics, the best they can do is the best linear approximation of the desired computation. Even more surprisingly, although quantum mechanical nonlocal correlations are, in general, stronger than the classical ones — after all, this is the whole point of nonlocality — these correlations do not help nonlocal computation: quantum devices cannot do better than the best linear approximation either. On the other hand, the very moment we allow for superquantum correlations, we can do nonlocal computation better than the best linear approximation. Hence, as far as nonlocal computation is concerned, there is a sharp transition between quantum and superquantum correlations.
Information causality
Suppose Alice sends to Bob a message consisting of a single binary digit (0 or 1). By this procedure, Alice cannot send Bob more than one bit of classical information, even if they also share some nonlocal particles and perform measurements on them according to the information they wish to transmit or receive. Indeed, if by such a procedure Alice could communicate to Bob more than one bit of information, they could also communicate superluminally. This is easy to prove — Bob wouldn't actually need to wait for Alice's message; he could simply guess it, perform his measurements according to the guess, simultaneously with those of Alice, and learn, with a success probability of 1/2, more than one bit of information. This can then easily be converted into learning some information with certainty.
An interesting possibility emerges, however. Suppose Alice has two bits that she wants to communicate to Bob. Even though by sending a onebit message she cannot communicate both bits to Bob, perhaps Bob could choose which bit to learn, even though he can make the decision at the last moment, long after Alice has already sent her message. Surprisingly, if Alice and Bob share a PR box, this is possible. Indeed, let x_{0} and x_{1} be Alice's two bits. She inputs x = x_{0} ⊕ x_{1} into her box and sends Bob the message m = x_{0} ⊕ a. If Bob wants to learn x_{0} he inputs in his box y = 0, whereas if he wants to learn x_{1} he inputs y = 1. Bob then calculates m ⊕ b. He obtains m ⊕ b = x_{0} ⊕ a ⊕ b = x_{0} ⊕ xy = x_{0} ⊕ (x_{0} ⊕ x_{1})y. It is easy to see that if y = 0 then m ⊕ b = x_{0} and if y = 1 then m ⊕ b = x_{1}.
On the other hand, one may feel uneasy with this result. Indeed, although Bob cannot find both x_{0} and x_{1}, one may consider that even the ability of Bob to choose which bit to learn should be unphysical. Indeed, the message sent by Alice consists of just one binary digit; how can it allow Bob to retrieve information about two bits, even if he cannot read both of them? Imposing the restriction that this is impossible yields a new principle, which was proposed by Pawlowski et al.^{76} and called 'information causality'.
As shown above, PR boxes violate information causality. However, it turns out that both classical physics and quantum mechanics obey information causality. And here comes the really exciting thing: for a restricted class of nonlocal correlations (namely the unbiased ones, where the local probabilities of all outcomes are equal), information causality breaks exactly at the boundary between quantum and superquantum nonlocal correlations. That is, suppose we make the PR boxes weaker by adding white noise until they become only as strong as quantum mechanical correlations. Exactly here information causality ceases to be violated. Information causality is, therefore, yet another example that singles out part of the quantum–superquantum boundary.
Quantum mechanics is special (or maybe not)
So what is the status of this research now? In this Review, I have discussed only a few examples; there is, however, intense, ongoing effort along similar lines^{82,83,84,85,86,87,88,89,90,91,92,93,94,95,96}.
Although it is early days, one can already see that quantum mechanics is special. Starting from various completely unrelated tasks that have nothing to do with the dynamics of microscopic particles, but are general purpose questions, such as nonlocal computation, information causality, macroscopic locality, the possibility of nonlocality swapping and so on, quantum mechanics emerges. It is precisely at the boundary between quantum mechanical and superquantum correlations that qualitative changes in the performance of the above tasks occur. True, these are only glimpses — there is no known task yet that completely differentiates quantum correlations from superquantum ones; only part of the boundary has emerged so far. Indeed, it is now known that any task that would be able to completely single out quantum mechanics has to be multipartite, as opposed to the bipartite tasks discussed here^{19}. However, it is remarkable that parts of the quantum boundary appeared at all — there was no a priori reason whatsoever for this to happen. Yet, quantum mechanics starts to appear from the fog. That quantum mechanics has special significance in at least some of such tasks means that quantum mechanics is special, and one should not expect that the ultimate theory of nature should be some slight deviation from quantum mechanics — there are basic statements about nature that have to be changed. It also means that quantum mechanics is probably here to stay — at least much longer than one would have imagined.
At the same time, one can legitimately question the relevance of such computer scienceinspired tasks in the grand scheme of things. Why should we care about such things as communication complexity, nonlocal computation or information causality? Why should we let our quest for a new theory of nature — or the justification for the present one — be guided by such ideas?
The very first indication that this line of thought is good is the simple fact that it seems to work. Quantum mechanics appears unexpectedly in various contexts. The fact that it does so is fascinating, and certainly nontrivial.
Second, whereas the tasks discussed here may appear quite random and completely insignificant from the point of view of hardcore physics — certainly they tell us nothing about the spectra of atoms or about phase transitions — from the point of view of information theory they are actually fundamental. (A pair of perfect PR boxes is a device that transforms the basic nonlinear function, the product, into a linear one, xy = a ⊕ b. At the same time, it can be viewed as the maximal zerocapacity communication channel.)
Yet again, it might not be quantum mechanics that we see emerging, but something altogether different. A few years ago Navascués and collaborators^{30} discovered a hierarchy of sets of 'selfconsistent' nonlocal correlations, each set is larger than quantum mechanics, but their boundaries coincide with quantum mechanics in some places. Maybe it is one of these sets that we are starting to see. These sets were discovered based on some rather obscure mathematical considerations, going opposite to the direction of considering natural tasks, which was the whole point of the research discussed above.
But recently, quantum gravity led to a tantalizing result: motivated by considerations of quantum gravity, a class of generalized theories was proposed by GellMann and Hartle^{97,98}, which was further developed by Sorkin^{99}. And in a very recent (yet unpublished) paper^{100}, it was shown that these theories lead to strongerthanquantum correlations, namely to the NavascuésPironioAcín set known as Q(1+AB), which is known to coincide with quantum mechanics in most of the places where the information tasks indicated quantum mechanics. Hence, maybe what those tasks indicate is Q(1+AB), not quantum mechanics. The jury is still out.
To conclude, all the above is great fun. Each answer raises new questions, completely different in nature from the ones one started with; this, more than anything else, indicates that finally we might be on the right track.
References
 1
Aharonov, Y. & Bohm, D. Significance of electromagnetic potentials in the quantum theory. Phys. Rev. 115, 485–491 (1959).
 2
Bell, J. S. On the EinsteinPodolskyRosen paradox. Physics 1, 195–200 (1964).
 3
Aharonov, Y. & Rohrlich, D. Quantum Paradoxes: Quantum Theory for the Perplexed (WileyVCH, 2005).
 4
Shimony, A. in Proc. Int. Symp. Foundations of Quantum Mechanics (eds Kamefuchi, S. et al.) 225–230 (Physical Society of Japan, 1983).
 5
Popescu, S. & Rohrlich, D. Causality and nonlocality as axioms for quantum mechanics. Found. Phys. 24, 379–385 (1994).
 6
Horodecki, R., Horodecki, P., Horodecki, M. & Horodecki, K. Quantum entanglement. Rev. Mod. Phys. 81, 865–942 (2009). Extensive review of the characterization of quantum nonlocality.
 7
Gisin, N. Bell inequality holds for all nonproduct states. Phys. Lett. A 154, 201–202 (1991).
 8
Popescu, S. & Rohrlich, D. Generic quantum nonlocality. Phys. Lett. A 166, 293–297 (1992).
 9
Barrett, J., Collins, D., Hardy, L., Kent, A. & Popescu, S. Quantum nonlocality, Bell inequalities and the memory loophole. Phys. Rev. A 66, 42111 (2002).
 10
Brunner, N., Gisin, N. & Scarani, V. Entanglement and nonlocality are different resources. New J. Phys. 7, 88 (2005).
 11
Cerf, N. J., Gisin, N., Massar, S. & Popescu, S. Simulating maximal quantum entanglement without communication. Phys. Rev. Lett. 94, 220403 (2005).
 12
Almeida, M., Pironio, S., Barrett, J., Tóth, G. & Acín, A. Noise robustness of the nonlocality of entangled quantum states. Phys. Rev. Lett. 99, 040403 (2007).
 13
Brunner, N., Gisin, N., Popescu, S. & Scarani, V. Simulation of partial entanglement with nosignaling resources. Phys. Rev. A 78, 52111 (2008).
 14
Allcock, J., Brunner, N., Pawlowski, M. & Scarani, V. Recovering part of the boundary between quantum and nonquantum correlations from information causality. Phys. Rev. A 80, 040103(R) (2009).
 15
Acín, A. et al. Unified framework for correlations in terms of local quantum observables. Phys. Rev. Lett. 104, 140404 (2010).
 16
Barnum, H., Beigi, S., Boixo, S., Elliott, M. B. & Wehner, S. Local quantum measurement and no signaling imply quantum correlations. Phys. Rev. Lett. 104, 140401 (2010).
 17
Oppenheim, J. & Wehner, S. The uncertainty principle determines the nonlocality of quantum mechanics. Science 330, 1072–1074 (2010).
 18
Barrett, J. & Gisin, N. How much measurement independence is needed to demonstrate nonlocality? Phys. Rev. Lett. 106, 100406 (2011).
 19
Gallego, R., Würflinger, L. E., Acín, A. & Navascués, M. Quantum correlations require multipartite information principles. Phys. Rev. Lett. 107, 210403 (2011).
 20
Bancal, JD. et al. Quantum nonlocality based on finitespeed causal influences leads to superluminal signalling. Nature Phys. 8, 867–870 (2012).
 21
Liang, YC., Masanes, L. & Rosset, D. All entangled states display some hidden nonlocality. Phys. Rev. A 86, 052115 (2012).
 22
Palazuelos, C. Superactivation of quantum nonlocality. Phys. Rev. Lett. 109, 190401 (2012).
 23
Vértesi, T. & Brunner, N. Quantum nonlocality does not imply entanglement distillability. Phys. Rev. Lett. 108, 030403 (2012).
 24
Cavalcanti, D., Acín, A., Brunner, N. & Vértesi, T. All quantum states useful for teleportation are nonlocal resources. Phys. Rev. A 87, 042104 (2013).
 25
Hirsch, F., Quintino, M. T., Bowles, J. & Brunner, N. Genuine hidden quantum nonlocality. Phys. Rev. Lett. 111, 160402 (2013).
 26
Khalfin, L. A. & Tsirelson, B. S. Quantum/classical correspondence in the light of Bell's inequalities. Found. Phys. 22, 879–948 (1992).
 27
Tsirelson, B. S. Some results and problems on quantum Belltype inequalities. Hadronic J. Suppl. 8, 329–345 (1993).
 28
Tsirelson, B. S. Quantum analogues of the Bell inequalities. The case of two spatially separated domains. J. Sov. Math. 36, 557–570 (1987).
 29
Tsirelson, B. S. Quantum generalizations of Bell's inequality. Lett. Math. Phys. 4, 93–100 (1980).
 30
Navascués, M., Pironio, S. & Acín, A. Bounding the set of quantum correlations. Phys. Rev. Lett. 98, 10401 (2007).
 31
Navascués, M., Pironio, S. & Acín, A. A convergent hierarchy of semidefinite programs characterizing the set of quantum correlations. New J. Phys. 10, 73013 (2008).
 32
Barnum, H., Barrett, J., Leifer, M. S. & Wilce, A. Generalized nobroadcasting theorem. Phys. Rev. Lett. 99, 240501 (2007).
 33
Barrett, J. Information processing in generalized probabilistic theories. Phys. Rev. A 75, 32304 (2007).
 34
Barrett, J. & Short, A. J. Strong nonlocality: a tradeoff between states and measurements. New J. Phys. 12, 33034 (2010).
 35
Barnum, H. et al. Entropy and information causality in general probabilistic theories. New J. Phys. 12, 033024 (2010).
 36
Short, A. J. & Wehner, S. Entropy in general physical theories. New J. Phys. 12, 33023 (2010).
 37
Gross, D., Mueller, M., Colbeck, R. & Dahlsten, O. C. O. All reversible dynamics in maximally nonlocal theories are trivial. Phys. Rev. Lett. 104, 80402 (2010).
 38
Janotta, P., Gogolin, C., Barrett, J. & Brunner, N. Limits on nonlocal correlations from the structure of the local state space. New J. Phys. 13, 063024 (2011).
 39
Barrett, J., Hardy, L. & Kent, A. No signaling and quantum key distribution. Phys. Rev. Lett. 95, 10503 (2005).
 40
Acín, A., Gisin, N. & Masanes, L. From Bell's theorem to secure quantum key distribution. Phys. Rev. Lett. 97, 120405 (2006).
 41
Scarani, V. et al. Secrecy extraction from nosignaling correlations. Phys. Rev. A 74, 42339 (2006).
 42
Acín, A. et al. Deviceindependent security of quantum cryptography against collective attacks. Phys. Rev. Lett. 98, 230501 (2007).
 43
Hänggi, E. & Renner, R. Deviceindependent quantum key distribution with commuting measurements. Preprint at http://arxiv.org/abs/1009.1833 (2010).
 44
Franz, T., Furrer, F. & Werner, R. F. Extremal quantum correlations and cryptographic security. Phys. Rev. Lett. 106, 250502 (2011).
 45
Masanes, L., Pironio, S. & Acín, A. Secure device independent quantum key distribution with causally independent measurement devices. Nature Commun. 2, 238 (2011).
 46
Barrett, J., Colbeck, R. & Kent, A. Unconditionally secure deviceindependent quantum key distribution with only two devices. Phys. Rev. A 86, 062326 (2012).
 47
Vazirani, U. & Vidick, T. Certifiable quantum dice. Phil. Trans. R. Soc. A. 370, 3432–3448 (2012).
 48
Barrett, J., Colbeck, R. & Kent, A. Memory attacks on deviceindependent quantum cryptography. Phys. Rev. Lett. 110, 010503 (2013).
 49
Huber, M. & Pawlowski, M. Weak randomness in deviceindependent quantum key distribution and the advantage of using highdimensional entanglement. Phys. Rev. A 88, 032309 (2013).
 50
Colbeck, R. Quantum And Relativistic Protocols For Secure MultiParty Computation PhD thesis, Univ. Cambridge (2007).
 51
Pironio, S. et al. Random numbers certified by Bell's theorem. Nature 464, 1021–1024 (2010).
 52
Colbeck, R. & Kent, A. Private randomness expansion with untrusted devices. J. Phys. A 44, 095305 (2011).
 53
Acín, A., Massar, S. & Pironio, S. Randomness versus nonlocality and entanglement. Phys. Rev. Lett. 108, 100402 (2012).
 54
Colbeck, R. & Renner, R. Free randomness can be amplified. Nature Phys. 8, 450–454 (2012).
 55
Vazirani, U. & Vidick, T. in Proc. 44th Symp. Theory Comput. STOC 2012 (eds Pitassi, T. et al.) 61–76 (ACM Press, 2012).
 56
Gallego, R. et al. Full randomness from arbitrarily deterministic events. Nature Commun. 4, 2654 (2013).
 57
Pironio, S. & Massar, S. Security of practical private randomness generation. Phys. Rev. A 87, 012336 (2013).
 58
Rabelo, R., Ho, M., Cavalcanti, D., Brunner, N. & Scarani, V. Deviceindependent certification of entangled measurements. Phys. Rev. Lett. 107, 050502 (2011).
 59
Bancal, JD., Gisin, N., Liang, YC. & Pironio, S. Deviceindependent witnesses of genuine multipartite entanglement. Phys. Rev. Lett. 106, 250404 (2011).
 60
Moroder, T., Bancal, JD., Liang, YC., Hofmann, M. & Gühne, O. Deviceindependent entanglement quantification and related applications. Phys. Rev. Lett. 111, 030501 (2013).
 61
Brunner, N., Cavalcanti, D., Pironio, S., Scarani, V. & Wehner, S. Bell nonlocality. Preprint at http://arxiv.org/abs/1303.2849 (2013).
 62
Hardy, L. Quantum theory from five reasonable axioms. Preprint at http://arxiv.org/abs/quantph/0101012 (2001).
 63
Chiribella, G., D'Ariano, G. M. & Perinotti, P. Probabilistic theories with purification. Phys. Rev. A 81, 062348 (2010).
 64
Paterek, T., Dakić, B. & Brukner, C. Theories of systems with limited information content. New J. Phys. 12, 053037 (2010).
 65
Chiribella, G., D'Ariano, G. M. & Perinotti, P. Informational derivation of quantum theory. Phys. Rev. A 84, 012311 (2011).
 66
Masanes, L. & Müller, M. P. A derivation of quantum theory from physical requirements. New J. Phys. 13, 063001 (2011).
 67
de la Torre, G., Masanes, L., Short, A. J. & Müller, M. P. Deriving quantum theory from its local structure and reversibility. Phys. Rev. Lett. 109, 090403 (2012).
 68
Masanes, L., Müller, M. P., Augusiak, R. & PérezGarcía, D. Existence of an information unit as a postulate of quantum theory. Proc. Natl Acad. Sci. USA. 110, 16373–16377 (2013).
 69
Müller, M. P. & Masanes, L. Threedimensionality of space and the quantum bit: an informationtheoretic approach. New J. Phys. 15, 053040 (2013).
 70
Popescu, S. Dynamical quantum nonlocality. Nature Phys. 6, 151–153 (2010).
 71
Clauser, J. F., Horne, M. A., Shimony, A. & Holt, R. A. Proposed experiment to test local hiddenvariable theories. Phys. Rev. Lett. 23, 880 (1969).
 72
Barrett, J., Linden, N., Massar, S., Pironio, S., Popescu, S. & Roberts, D. Nonlocal correlations as an information theoretic resource. Phys. Rev. A 71, 22101 (2005).
 73
van Dam, W. Implausible consequences of superstrong nonlocality. Preprint at http://arxiv.org/abs/quantph/0501159 (2005).
 74
Brassard, G. et al. Limit on nonlocality in any world in which communication complexity is not trivial. Phys. Rev. Lett. 96, 250401 (2006).
 75
Linden, N., Popescu, S., Short, A. J. & Winter, A. Quantum nonlocality and beyond: limits from nonlocal computation. Phys. Rev. Lett. 99, 180502 (2007).
 76
Pawlowski, M. et al. Information causality as a physical principle. Nature 461, 1101–1104 (2009).
 77
Navascués, M. & Wunderlich, H. A glance beyond the quantum model. Proc. R. Soc. A 466, 881–890 (2010).
 78
Fritz, T. et al. Local orthogonality as a multipartite principle for quantum correlations. Nature Commun. 4, 2263 (2013).
 79
Skrzypczyk, P., Brunner, N. & Popescu, S. Emergence of quantum correlations from nonlocality swapping. Phys. Rev. Lett. 102, 110402 (2009).
 80
Cleve, R., van Dam, W., Nielsen, M. & Tapp, A. Quantum Computing and Quantum Communication, Lecture Notes in Computer Science Vol. 1509 (Springer, 1999).
 81
Cleve, R. & Buhrman, H. Substituting quantum entanglement for communication. Phys. Rev. A 56, 1201–1204 (1997).
 82
Barrett, J. & Pironio, S. PopescuRohrlich correlations as a unit of nonlocality. Phys. Rev. Lett. 95, 140401 (2005).
 83
Methot, A. A. & Scarani, V. An anomaly of nonlocality. Quantum Inform. Compu. 7, 157–170 (2007).
 84
Forster, M., Winkler, S. & Wolf, S. Distilling nonlocality. Phys. Rev. Lett. 102, 120401 (2009).
 85
Brunner, N. & Skrzypczyk, P. Nonlocality distillation and postquantum theories with trivial communication complexity. Phys. Rev. Lett. 102, 160403 (2009).
 86
Pawlowski, M. & Brukner, C. Monogamy of Bell's inequality violations in nonsignaling theories. Phys. Rev. Lett. 102, 30403 (2009).
 87
Cavalcanti, D., Salles, A. & Scarani, V. Macroscopically local correlations can violate information causality. Nature Commun. 1, 136 (2010).
 88
Fitzi, M., Hänggi, E., Scarani, V. & Wolf, S. The nonlocality of n noisy Popescu–Rohrlich boxes. J. Phys. A 43, 465305 (2010).
 89
Brunner, N., Cavalcanti, D., Salles, A. & Skrzypczyk, P. Bound nonlocality and activation. Phys. Rev. Lett. 106, 20402 (2011).
 90
Gallego, R., Würflinger, L. E., Acín, A. & Navascués, M. Operational framework for nonlocality. Phys. Rev. Lett. 109, 70401 (2012).
 91
Borsten, L., Bradler, K. & Duff, M. J. Tsirelson's bound and supersymmetric entangled states. Preprint at http://arxiv.org/abs/1206.6934 (2012).
 92
Short, A. J., Popescu, S. & Gisin, N. Entanglement swapping for generalized nonlocal correlations. Phys. Rev. A 73, 12101 (2006).
 93
Branciard, C., Gisin, N. & Pironio, S. Characterizing the nonlocal correlations created via entanglement swapping. Phys. Rev. Lett. 104, 170401 (2010).
 94
Bancal, JD., Brunner, N., Gisin, N. & Liang, YC. Detecting genuine multipartite quantum nonlocality: a simple approach and generalization to arbitrary dimensions. Phys. Rev. Lett. 106, 020405 (2011).
 95
Branciard, C., Rosset, D., Gisin, N. & Pironio, S. Bilocal versus nonbilocal correlations in entanglement swapping experiments. Phys. Rev. A 85, 032119 (2012).
 96
Bancal, JD., Barrett, J., Gisin, N. & Pironio, S. Definitions of multipartite nonlocality. Phys. Rev. A 88, 014102 (2013).
 97
GellMann, M. & Hartle, G. B. in Complexity, Entropy and the Physics of Information: SFI Studies in the Sciences of Complexity Vol VIII (ed. Zurek, W.) 150–173 (Addison Wesley, 1990).
 98
Hartle, J. B. in Proc. 1992 Les Houches Summer School on Gravitation and Quantization (eds ZinnJustin, J. & Julia, B.) 285–480 (NorthHolland, 1995).
 99
Sorkin, R. D. Quantum mechanics as quantum measure theory. Mod. Phys. Lett. A 9, 3119–3128 (1994).
 100
Dowker, F. Henson, J. & Wallden, P. A histories perspective on characterising quantum nonlocality. Preprint at http://arxiv.org/abs/1311.6287 (2013).
Author information
Affiliations
Corresponding author
Ethics declarations
Competing interests
The author declares no competing financial interests.
Rights and permissions
About this article
Cite this article
Popescu, S. Nonlocality beyond quantum mechanics. Nature Phys 10, 264–270 (2014). https://doi.org/10.1038/nphys2916
Received:
Accepted:
Published:
Issue Date:
Further reading

Quantum Black Holes as Solvents
Foundations of Physics (2021)

Constraints on nonlocality in networks from nosignaling and independence
Nature Communications (2020)

Nonlocal correlations and noise in different settings of a twoplayer game
Quantum Information Processing (2020)

Noisy threeplayer dilemma game: robustness of the quantum advantage
Quantum Information Processing (2020)

The computational landscape of general physical theories
npj Quantum Information (2019)