A new physical principle behind quantum physics

Tomasz Paterek, Dagomir Kaszlikowski, Valerio Scarani and Andreas Winter from CQT, together with co-workers of the University of Gdansk (Poland), propose a new physical principle called "information causality" in a paper published in Nature. If this principle is enforced, the number of theories that can describe our world is drastically reduced. This might explain why no phenomenon has ever been observed that would go beyond quantum physics.

Quantum Mechanics is the most successful physical theory we have. It's predictive power is impressive, both for pure science (think of particle physics) and for applications (the laser, semiconductors). A significant amount of public and private funding goes into research in quantum physics, including here in Singapore.

Naturally, one would like to know what quantum physics is actually about. But this query normally receives the frustrating answer "Hmmm, well, it's complicated!" - which nobody doubted. However, relativity is also complicated, and still physicists can tell you that it is based on a simple physical principle, namely the constancy of the speed of light. In the same vein, should not physicists state the principle on which quantum physics is built? The problem is that nobody has ever been able to formulate such a principle; so much so, that most physicists just gave up the quest for it. The present work revives this quest by focusing on a physical principle that had never been noticed before. This principle, called information causality, is indeed satisfied in nature; moreover, it comes very close to singling out quantum physics as the only possible description of natural phenomena.

Admittedly, the principle of information causality is still more complicated than the simple principle of the constancy of the speed of light. An example is probably the best way of approaching it.

Alice and Bob had a nice holiday together in Singapore: they shared experiences, pictures, restaurant bills... For our purpose, they might even have shared quantum objects. Now they are both back in their respective countries. In the UK, Alice discovers a CD with 10 songs, which she would like to share with Bob. Quite clearly, Alice has to send Bob some information: Bob will not find the songs among whatever was shared in the past. Suppose now that each song occupies 10Mbits and that Alice can send only 10Mbits: information causality says that the best thing Alice can do, is to choose one of the songs and send that one to Bob.

Isn't this intuitive? It is, and this is why it can be proposed as a physical principle. But isn't it actually trivial? No, it is not trivial. Remember that the goal is to single out quantum physics, so one has to study what would happen if one could go beyond what quantum physics allows. In the example, one then assumes that Alice and Bob have shared some hypothetical "more-than-quantum objects". Then the following becomes possible: Alice sends 10Mbits in a way that allows Bob to recover any one of the songs - not all of them, just one, but he can choose which one at a later stage. In this scenario, all the 100Mbits of the songs are encoded in 10Mbits, in such a way that each song, though only one, can be reconstructed perfectly. This is what information causality forbids: Alice cannot encode 100Mbits in 10Mbits, in such a way that Bob can reconstruct perfectly any 10Mbits of his choice.

In the light of this example, information causality can be summarized as follows. In a scenario with bounded communication, information cannot be coded in such a way, that the receiver can choose later which part to retrieve: the sender must choose which part to send.

The fact that quantum physics does not break information causality, while anything beyond quantum physics would break it, shows that this principle is not trivial and may actually be one of the defining features of our universe.
 

Reference: M. Pawlowski, T. Paterek, D. Kaszlikowski, V. Scarani, A. Winter, M. Zukowski, Information causality as a physical principle, Nature 461, 1101 (2009).