Abstract

Why don't we see large macroscopic objects in entangled states? Even if the particles composing the object were all entangled and insulated from the environment, we shall still find it almost always impossible to observe the superposition. The reason is that as the number of particles n grows, we need an ever more careful preparation, and an ever more carefully designed experiment, in order to recognize the entangled character of the state of the object. An observable W that distinguishes all the unentangled states from some entangled states is called a witness. We consider witnesses on n quantum bits (qbits), and use the following normalization: A witness W satisfies |tr(Wr)|<= 1 for all separable states r, while ||W|| >1, with the norm being the maximum among the absolute values of the eigenvalues of W. Although there are n-qbit witnesses whose norm is exponential in n, we conjecture that for a large majority of such witnesses ||W||<=O[(nlogn)^1/2]. We prove this conjecture for the family of extremal witnesses introduced by Werner and Wolf (Phys. Rev. A 64, 032112 (2001)). Assuming the conjecture is valid we argue that multiparticle entanglement can be detected only if a system has been carefully prepared in a very special state. Otherwise, multiparticle entanglement lies below the threshold of detection, even if it exists, and even if decoherence has been ``turned off''.