Local observables of an infinite system are described by a quasilocal C*-algebra A. Traditionally classical observables are introduced as parameters that can be assigned to certain states on the local observables. As fundamental theories should be essentially free of parameters this can hardly be regarded as satisfactory. In the present paper classical observables are introduced explicitly as operators by a transitive system of imprimitivity of a kinematical group in the center of the W*-algebra π(A)″ belonging to a representation π of A. The states of the system are represented by linear functionals on π(A)″, which are dispersion-free on the center of π(A)″. For classical observables related to a kinematical group a W*-description—with algebra π(A)″ and its associated linear functionals that are dispersion-free on the center and normal on the noncentral part of π(A)″—is equivalent to a C*-description—with algebra A and a distinguished family of linear functionals on A, determined uniquely by π. The implications of these results on the interpretation of quantum mechanics of infinite systems are discussed.
Local observables of an infinite system are described by a quasilocal C*-algebra A. Traditionally classical observables are introduced as parameters that can be assigned to certain states on the local observables. As fundamental theories should be essentially free of parameters this can hardly be regarded as satisfactory. In the present paper classical observables are introduced explicitly as operators by a transitive system of imprimitivity of a kinematical group in the center of the W*-algebra π(A)″ belonging to a representation π of A. The states of the system are represented by linear functionals on π(A)″, which are dispersion-free on the center of π(A)″. For classical observables related to a kinematical group a W*-description—with algebra π(A)″ and its associated linear functionals that are dispersion-free on the center and normal on the noncentral part of π(A)″—is equivalent to a C*-description—with algebra A and a distinguished family of linear functionals on A, determined uniquely by π. The implications of these results on the interpretation of quantum mechanics of infinite systems are discussed.
Following the work of P. Bona [J. Math. Phys. 29, 2223 (1988)], a class of infinite mean-field quantum lattice systems is considered within the framework of algebraic quantum mechanics, and the following question is discussed: Which are the possible dynamical descriptions of the system, i.e., in which representations π of the system’s quasilocal C*-algebra 𝒜 do the local time evolutions converge in the thermodynamic limit to an automorphism group of a certain enlarged C*-algebra 𝒞π containing π(𝒜) and (some of) the classical observables of the system in π. To this end we associate a ‘‘classical state space’’ Eπ to any representation π (which is a subset of the phase-space E⊇RL defined by Bona for the largest possible representation), and show how it is characterized by the π-normal states, thereby obtaining a way to determine Eπ explicitly in many cases. The answer to the question posed then reads: The limiting (Heisenberg) dynamics exists in π on 𝒞π if and only if the set Eπ is invariant under the flow φQt on E which describes the time evolution of classical observables. Conversely, to any invariant, closed subset B of E, there exists a dynamical description of the quantum system with classical state space B.
A certain infinite-dimensional Lie group, obtained originally by regularizing and then exponentiating a singular local current algebra constructed from canonical fields, has been shown over several decades to describe a wide variety of quantum systems, via its continuous unitary representations. The group is a semidirect product of Schwartz space functions under pointwise addition with a diffeomorphism group under composition. Previously unsuspected possibilities were predicted this way, including anyons and nonabelian anyons in two-dimensional space. Last year David Sharp and I offered a much simpler, direct construction. We proposed fundamental reasons why this group serves as a universal kinematical group for quantum systems with mass in an arbitrary physical space. This paper beyond summarizing that construction, discusses more extensively what it means to obtain quantum mechanics without actually quantizing classical observables in phase space, and answers some fundamental questions. Directions are suggested for obtaining still greater generality of kinematical and dynamical possibilities by modifying the initial assumptions.
This thesis is an examination of the Algebraic Approach to Quantum Mechanics and its use in the description of infinite Quantum systems.This approach is based on the postulate that it is possible to set up a correspondence between Quantum systems and the mathematical theory of C*algebras.In Chapter 1 the theory of C*aIgebras and their representations is discussed.The main result isthe G.N.S. construction.In Chapter 2 the methods developed in Chapter 1 are applied to Quantum spin sytems.The C*alebra corresponding to a single spin is constructed and analyzed, and the G.N.S. representation for the state corresponding to a canonical ensemble with definite temperature is constructed.The discussion is then generalized to a two spin system, a multi spin system, and finally to a system consisting of an infinite number of spins.It is shown that the G.N.S. representations corresponding to different temperatures are unitarily equivalent for the finite system, while for the infinite system these representations become unitarily inequivalent.This example illustrates that classical variables, in this case the temperature, arise in quantum systems as labels that distinguish between different inequivalent representations of the Quantum algebra in the case of inifinitely many degrees of freedom.In Chapter 3 the algebraic description of a non-interactihg Bose gas is examined.The C.C.R. algebra is constructed.Methods developed by Araki and Woods are used to construct the state of the infinite Bose gas that has density p and in 0 which all particles have zero momentum.The representation constructed by Araki andWoods is shown to be the G.N.S. representation for this state.It is shown that for the infinite gas the GN.S. representations corresponding to different densities are unitarily inequivalent.Again a classical variable, the density in this case, arises as a label that distinguishes the different inequivalent representations.
The algebra of polynomials in operators that represent generalized coordinate and momentum and depend on the Planck constant is defined. The Planck constant is treated as the parameter taking values between zero and some nonvanishing h0. For the second of these two extreme values, introduced operatorial algebra becomes equivalent to the algebra of observables of quantum mechanical system defined in the standard manner by operators in the Hilbert space. For the vanishing Planck constant, the generalized algebra gives the operator formulation of classical mechanics since it is equivalent to the algebra of variables of classical mechanical system defined, as usually, by functions over the phase space. In this way, the semiclassical limit of kinematical part of quantum mechanics is established through the generalized operatorial framework.
This is an up-to-date survey of the p-mechanical construction (see funct-an/9405002, quant-ph/9610016, math-ph/0007030, quant-ph/0212101, quant-ph/0303142), which is a consistent physical theory suitable for a simultaneous description of classical and quantum mechanics. Observables in p-mechanics are defined to be convolution operators on the Heisenberg group H^n. Under irreducible representations of H^n the p-observables generate corresponding observables in classical and quantum mechanics. p-States are defined as positive linear functionals on p-observables. It is shown that both states and observables can be realised as certain sets of functions/distributions on the Heisenberg group. The dynamical equations for both p-observables and p-states are provided. The construction is illustrated by the forced and unforced harmonic oscillators. Connections with the contextual interpretation of quantum mechanics are discussed. Keywords: Classical mechanics, quantum mechanics, Moyal brackets, Poisson brackets, commutator, Heisenberg group, orbit method, deformation quantisation, symplectic group, representation theory, metaplectic representation, Berezin quantisation, Weyl quantisation, Segal--Bargmann--Fock space, coherent states, wavelet transform, Liouville equation, contextual interpretation, interaction picture, forced harmonic oscillator.
The work can be considered as an essay on mathematical and conceptual structure of nonrelativistic quantum mechanics (QM) which is related here to some other (more general, but also to more special and “approximative”) theories. QM is here primarily reformulated in an equivalent form of a Poisson system on the phase space consisting of density matrices, where the “observables”, as well as “symmetry generators” are represented by a specific type of real valued (densely defined) functions, namely the usual quantum expectations of corresponding selfadjoint operators. It is shown in this paper that inclusion of additional (“nonlinear”) symmetry generators (i.e. “Hamiltonians”) into this reformulation of (linear) QM leads to a considerable extension of the theory: two kinds of quantum “mixed states” should be distinguished, and operator ‐ valued functions of density matrices should be used in the r ˆ ole of “nonlinear observables”. A general framework for physical theories is obtained in this way: By different choices of the sets of “nonlinear observables” we obtain, as special cases, e.g. classical mechanics on homogeneous spaces of kinematical symmetry groups, standard (linear) QM, or nonlinear extensions of QM; also various “quasiclassical approximations” to QM are all subtheories of the presented extension of QM - a version of the extended quantum mechanics (EQM). A general interpretation scheme of EQM extending the usual statistical interpretation of QM is also proposed. Eventually, EQM is shown to be (included into) a C -algebraic (hence linear) quantum theory. Mathematical formulation of these theories is presented. The presentation includes an analysis of problems connected with differentiation on infinite ‐ dimensional manifolds, as well as a solution of some problems connected with the work with only densely defined unbounded real‐valued functions on the (infinite dimensional) “phase space” which correspond to unbounded operators (generators) and to their nonlinear generalizations. Also “nonlinear deformations” of unitary representations of kinematical symmetry Lie groups are introduced. Possible applications are briefly discussed, and some specific examples are presented. The text contains also brief reviews of Hamiltonian classical mechanics, as well as of QM. Mathematical appendices make the paper nearly selfcontained.