We analyze symmetries of the 1-loop effective action of φ 4 noncommutative field theory. It is shown, that despite the twisted Poincaré invariance of the classical noncommutative action, its 1-loop quantum counterpart lacks this invariance. Though Noether analysis of the model is somewhat obscure, it is still possible to interpret this symmetry breaking as a quantum anomaly due to inappropriate choice of the quantization method. * apinzul@unb.br 1
IntroductionBased on a quite general argument [1], it is believed that one or another form of a noncommutative (NC) space-time should emerge as a quasi-classical approximation to quantum gravity in a sense that quantum fields on such a space-time will probe the quantum gravity scale through noncommutativity. This observation, as well as the appearance of NC field theory in super strings [2], sparked considerable interest in noncommutative physics. Most of the efforts have been devoted to the analysis of theories on a NC space-time with constant noncommutativity, a so-called Moyal space. (Literature on the subject is vast, for a review see, e.g. [3]) This is the simplest possible type of noncommutativity, defined by the following commutation relation between the coordinatesHere θ µν gives the scale of quantum gravity, usually it is thought to be related to the Planck scale. It is clear that if θ µν is some non-dynamical constant, which is the case in string theory, then (1) breaks Lorentz invariance.The study of Lorentz-breaking theories has a long history by now [4]. Up to the first order in θ µν , NC quantum field theory with constant noncommutativity is just a special case of the general situation (e.g., see discussion in [5]). The quantization of NC field theories (NCFT) had been performed within this framework, i.e quantization of a Lorentz violating theory. Perturbative analysis led to some noncommutative corrections but also, what is more important, to a quite universal property of NCFT -UV/IR mixing [6]. (Though, see some recent proposals on models without UV/IR mixing [7][8][9].) This property, which is due to the non-local nature of noncommutativity, put an end to hopes that noncommutativity (at least in the simple form of a constant θ µν ) would serve as a UV regulator, which was the major motivation for one of the first works on a NC space-time [10].After works [11,12], another possibility to look at NCFT had become possible: instead of thinking in terms of broken Lorentz invariance, one could now think about NCFT as a theory with preserved, but twisted Lorentz symmetry.1 On the classical level, this is just a point of view that does not add anything new to the picture. But the transition to quantization changes the situation drastically: one has to maintain this new twisted symmetry on quantum level. Essentially, there exist two approaches to quantization of NCFT with twisted Lorentz symmetry:• The quantization should be done according to the usual rules. In other words, the usual Feynman rules, derived from the path integral with the standard quant...