Towards Space from Hilbert Space: Finding Lattice Structure in Finite-Dimensional Quantum Systems

TL;DR

This paper investigates how the tensor product lattice structure of finite-dimensional Hilbert spaces in quantum theories can emerge from Hilbert-space considerations alone, highlighting limitations and conditions for such structures.

Contribution

It introduces the concept of direct-sum locality and explores the emergence of lattice structures in finite-dimensional Hilbert spaces, with implications for quantum field theory and cosmology.

Findings

Most finite-dimensional Hilbert spaces cannot be decomposed into tensor products like lattice theories.

A notion of direct-sum locality is defined to characterize compatible states and decompositions.

Toy model illustrates the concepts in a quantum-mechanical double-well potential.

Abstract

Field theories place one or more degrees of freedom at every point in space. Hilbert spaces describing quantum field theories, or their finite-dimensional discretizations on lattices, therefore have large amounts of structure: they are isomorphic to the tensor product of a smaller Hilbert space for each lattice site or point in space. Local field theories respecting this structure have interactions which preferentially couple nearby points. The emergence of classicality through decoherence relies on this framework of tensor-product decomposition and local interactions. We explore the emergence of such lattice structure from Hilbert-space considerations alone. We point out that the vast majority of finite-dimensional Hilbert spaces cannot be isomorphic to the tensor product of Hilbert-space subfactors that describes a lattice theory. A generic Hilbert space can only be split into a direct sum corresponding to a basis of state vectors spanning the Hilbert space; we consider setups in which the direct sum is naturally decomposed into two pieces. We define a notion of direct-sum locality which characterizes states and decompositions compatible with Hamiltonian time evolution. We illustrate these notions for a toy model that is the finite-dimensional discretization of the quantum-mechanical double-well potential. We discuss their relevance in cosmology and field theory, especially for theories which describe a landscape of vacua with different spacetime geometries.