On the nonlinear extension of quantum superposition and uncertainty principles (Q1299126)

It was shown in [\textit{R. Cirelli}, \textit{A. Mania}, and \textit{L. Pizzocchero}, J. Math. Phys. 31, No. 12, 2891-2897 (1990), ibid. 2898-2903 (1990)] that a strong form of the Heisenberg uncertainty principle and a carefully selected set of reasonable mathematical hypotheses on the algebraic structure of the set of observables (associativity and algebraic closure) are sufficient to force the choice of Projective Quantum Mechanics. This expression will denote in this work quantum mechanics (QM) seen as a geometry of the projective Hilbert space. This theory, a synthetical description of which may be found in [loc. cit.], has the same physical content of the ordinary Hilbertian formulation of QM without superselection rules, but it avoids the annoying and conceptually misleading ``tip to phase'' language. In spite of the fundamental role played by the observable algebra in the \(C^*\)-algebraic formulation of QM, we give up in this work the algebraic structure of the observable set and only take as fundamental the quantum superposition principle (QSP) and the Heisenberg uncertainty principle (HUP). By abstracting from the usual geometric formulation of quantum mechanics, we select a minimal geometric formalism allowing -- the formulation of QSP, -- a complete characterization of observables as those respecting superpositions, -- a complete characterization of dynamical vector fields as those whose flows respect superpositions (isometricity), and -- the formulation of HUP. As we will see, a considerable part of the quantum formalism, which is necessary to reproduce the algebraic structure of QM, plays no role in the description of quantum fundamental principles. We only need a connection (to describe superpositions), a Finslerian structure (to define dispersions) and a Poissonian structure (to formulate HUP). This work is the first step towards a geometrical foundation of nonlinear quantum mechanics (NLQM). Here wo just identify some geometric structures which have to be (and can be) saved in attempts to delinearize ordinary quantum mechanics. However, superposition and uncertainty principles and the probabilistic content of QM are strictly linked to spectral theory of observables. So we have to study also a suitable spectral theory for nonlinear observables. This will be the argument of a forthcoming publication. Of course, at this stage of development of the delinearization program we do not claim a complete geometrical framework for nonlinear extensions of quantum mechanics.