Source-linked AI summary
Quantum from principles
Giulio Chiribella, Giacomo Mauro D'Ariano, Paolo Perinotti
TL;DR
The paper asks whether quantum theory can be reconstructed from first principles rather than ad hoc or opaque axioms. It formulates six information-processing principles in operational-probabilistic theories and shows that they identify quantum theory, including its Hilbert-space framework and characteristic information features. The central conclusion is that Purification characterizes quantum theory as allowing maximal control of randomness.
Problem
The paper addresses the need for an insightful reconstruction explaining what distinguishes quantum theory from alternative theories and why it should be preferred.
Method
The paper formulates six principles in operational-probabilistic theories, an extension of probability theory in which events connect into circuits.
Results
The six principles identify quantum theory, reconstruct quantum information features, and ultimately recover the Hilbert-space framework.
Takeaways & Limitations
Purification is the quantum principle: it enables every state preparation to be simulated through a pure bipartite preparation and expresses maximal control of randomness.
Takeaways & Limitations
The reconstruction assumes finite-dimensional systems, non-determinism, and closure under operational limits, while Ideal Compression is restricted to single-shot, zero-error scenarios.
Abstract
from arXiv · showhide
Quantum theory was discovered in an adventurous way, under the urge to solve puzzles-like the spectrum of the blackbody radiation-that haunted the physics community at the beginning of the 20th century. It soon became clear, though, that quantum theory was not just a theory of specific physical systems, but rather a new language of universal applicability. Can this language be reconstructed from first principles? Can we arrive at it from logical reasoning, instead of ad hoc guesswork? A positive answer was provided in Refs. [1, 2], where we put forward six principles that identify quantum theory uniquely in a broad class of theories. We first defined a class of "theories of information", constructed as extensions of probability theory in which events can be connected into networks. In this framework, we formulated the six principles as rules governing the control and the accessibility of information. Directly from these rules, we reconstructed a number of quantum information features, and eventually, the whole Hilbert space framework. In short, our principles characterize quantum theory as the theory of information that allows for maximal control of randomness.
1. Introduction
The paper seeks an insightful, operationally grounded reconstruction of quantum theory rather than specialized logical axioms. It reviews six principles in theories of information, with Purification uniquely selecting quantum theory and expressing maximal control of randomness.
- Motivation: Quantum logic derived much of the quantum framework from logical axioms, but its axioms were considered insufficiently insightful and difficult to interpret.The motivation is to explain what distinguishes quantum theory from alternatives and why it should be preferred.
- Motivation: Quantum information suggested reversing the usual implication: deriving quantum mathematics from operational consequences such as key distribution, algorithms, no-cloning, teleportation, and dense coding.
- Motivation: Operational advantages over classical information made quantum theory a more promising target for axiomatization than a merely impoverished classical theory.
- Approach: The framework combines circuit-connected events with probability theory to form theories of information suitable for reconstructing quantum theory.
- Contribution: Five principles define a standard theory of information, while Purification uniquely identifies quantum theory by allowing random preparations to be simulated using pure system-environment preparations.The six principles are Causality, Purity of Composition, Local Discriminability, Perfect State Discrimination, Ideal Compression, and Purification.
2. Operational-probabilistic theories
Operational-probabilistic theories extend probability theory with composable events, systems, tests, and circuits. Their sequential and parallel compositions form a strict symmetric monoidal framework for describing information-processing processes.
- Framework: Operational-probabilistic theories combine categorical circuit structure with elementary probability theory so events can be connected into networks.
- Systems and events: Systems label the inputs and outputs that determine how events can be connected, with composite systems formed using tensor composition.
- Systems and events: Events represent transformations from input system A to output system B, including preparations I → A and observations A → I.
- Composition: Sequential composition connects matching event types, while parallel composition combines events on separate systems.
- Composition: Associativity, identities, and swap operations make the event structure a strict symmetric monoidal category.
- Tests: A test is a finite-outcome collection of alternative events, with deterministic tests containing a single event and preparation- and observation-tests representing nondeterministic preparation and measurement.
2.2. Probabilistic structure
The probabilistic structure assigns probabilities to closed experiments while preserving consistency and independence, then identifies events with identical circuit statistics. This quotient yields a statistically relevant operational theory represented through ordered vector spaces and positive maps.
- Probabilistic structure: A probabilistic structure assigns probabilities to scalar events produced by composing all tests in a closed experiment.
- Probabilistic structure: Consistency normalizes the probabilities of outcomes in every test, while independence factorizes probabilities for independent scalar events.
- Probabilistic structure: An operational-probabilistic theory consists of an operational structure together with a compatible probabilistic structure.
- Statistical equivalence: Statistical equivalence identifies events that produce the same probabilities in every circuit involving arbitrary preparations, observations, and ancillary systems.
- Statistical equivalence: Equivalence classes of events and tests preserve sequential and parallel composition, forming strict symmetric monoidal structures.
- Quotient theory: The quotient theory discards distinctions without probabilistic consequences and is subsequently treated as the default operational-probabilistic theory.
- Vector-space representation: The quotient structure supports ordered vector spaces for states and effects, positive maps for events, and does not assume the No-Restriction Hypothesis.Without Local Tomography, the transformation-to-linear-map correspondence need not be one-to-one; Local Tomography removes this issue.
3. Background of the quantum reconstruction
The reconstruction uses finite-dimensional, nondeterministic theories closed under operational limits and introduces operational notions for signaling, side information, tomography, and state discrimination. These notions provide the setting and tasks motivating the axioms.
- 3.1. Standing assumptions: The framework assumes finite-dimensional systems, non-determinism, and closure under operational limits.Finite dimensionality permits state identification from finitely many finite-outcome measurements, while non-determinism requires at least one experiment with non-predetermined outcomes.
- Operational tasks: Signaling asks whether changing one test alters another node’s marginal outcome distribution, including the possibility of signaling from the future to the past.
- Operational tasks: A refined test can expose side information unavailable to the original agent, motivating pure transformations as those that leak no useful side information.
- Operational tasks: Pure transformations are defined by the requirement that side information held by another agent is uncorrelated with the transformation occurring in the first agent’s laboratory.
- Operational tasks: State tomography determines whether a restricted set of observations distinguishes every pair of states through their measurement statistics.
- Operational tasks: Local tomography holds when product measurements on component systems are tomographically complete for the composite system.Checking bipartite systems suffices to establish local tomography for arbitrary multipartite systems.
4. The principles
The paper presents five axioms and one postulate for standard operational-probabilistic theories, with Purification supplying the distinctive non-classical principle. Together, these principles constrain information processing and characterize quantum theory.
- Principles: The framework consists of five axioms and one postulate, distinguished by their roles in reconstructing quantum theory.The axioms, background assumptions, and postulate are mathematically treated alike, but the names emphasize their different conceptual roles.
- Principles: The five axioms require causality, purity of composition, local tomography, perfect state discrimination, and ideal compression.These requirements cover signaling, compositional side information, local state determination, discrimination, and state compression.
- Principles: Purification requires every preparation to be simulable via a pure preparation and is presented as a radically non-classical feature.The paper states that this feature singles out quantum theory uniquely among standard operational-probabilistic theories.
- Causality: Without interaction between systems A and B, tests on B cannot influence the probability distribution of a test on A.The paper identifies this as an important consequence of Causality.
- Causality: Causality means that future choices cannot affect past outcome probabilities and is equivalent to each system having a unique deterministic effect.The unique deterministic effect also permits canonical marginals and represents a unique way to discard each physical system.
- Causality: The positive reformulation of Causality enables conditional tests, in which later tests depend on outcomes of earlier tests.The conditional collection is required to constitute a valid test.
1. C′ implies C,
The reconstruction derives standard information-processing properties and distinctive quantum features from operational principles. Purification is singled out as the principle that supports entanglement, reversible dilation, and the Hilbert-space framework.
- C′ implies C,: Convexity follows from non-determinism, Causality, and closure of the probabilities generated by tests, rather than being assumed initially.Conditional tests generate convex combinations of transformations.
- C′ implies C,: Purity of Composition ensures that sequential composition of pure transformations remains pure.For quantum transformations, composing single-Kraus-operator maps preserves the single-Kraus form.
- C′ implies C,: Local Tomography makes transformations invertibly identifiable with linear maps determined by preparing inputs and measuring outputs.The principle also permits local measurements to extract information from entangled states.
- C′ implies C,: Perfect State Discrimination lets every normalized non-internal state be paired with another state to encode a bit without errors.In quantum theory, non-internal density matrices have a kernel containing a perfectly distinguishable state.
- C′ implies C,: Ideal Compression is a single-shot, zero-error property, distinct from asymptotic Shannon and Schumacher compression.It applies when the source is used once and decoding errors are disallowed.
- C′ implies C,: Purification enables maximal control of an environment, from which entanglement, Hilbert-space structure, and reversible dilation of irreversible processes follow.Under Causality, Local Tomography, and Purification, a theory is either deterministic or entangled; under non-determinism, entanglement is necessary.
5. The reconstruction of Quantum Theory
The reconstruction proceeds by deriving quantum features from first principles before introducing Hilbert spaces. It is organized into six stages, from elementary facts through the density matrix.
- 5. The reconstruction of Quantum Theory: The reconstruction aims to derive the Hilbert-space framework and key quantum features directly from axioms, postponing Hilbert spaces until the end.The stated scope includes rebuilding quantum features rather than deriving only the mathematical framework.
- 5. The reconstruction of Quantum Theory: The six stages cover elementary facts, correlation structures, distinguishability structures, their interaction, qubit features, and the density matrix.These subsections define the chapter’s progression from operational structures to the quantum state representation.
5.1. Elementary facts
Local Tomography supplies structural consequences for composite systems, while Purification makes pure states reversibly connected and supports minimal purifications.
- 5.1. Elementary facts: Local Tomography implies that products of pure states are pure and that a composite state with a pure marginal is a product state.It also preserves internality under tensor products and factors unique invariant states.
- 5.1. Elementary facts: Purification makes every pair of pure states of a system connected by a reversible transformation.The reversible group acts transitively on the manifold of pure states.
- 5.1. Elementary facts: Combining Ideal Compression and Purification yields a minimal purification for every state.A minimal purification is obtained by compressing the purifying system until its marginal is internal.
5.2. Correlation structures
Pure Steering yields operational correlation structures, including faithful states, a state-transformation isomorphism, and entanglement-based protocols such as swapping and teleportation.
- Pure Steering: Pure Steering remotely generates every ensemble decomposition of a marginal state through a measurement on its purification.
- Faithful states: A purification of an internal state is tomographically faithful, while a minimal purification is faithful on both component systems.
- State-transformation isomorphism: The Bell state induces a one-to-one state-transformation isomorphism, mapping transformations to Choi states and identifying the full theory through normalized states.
- Entanglement protocols: Pure Steering enables conclusive entanglement swapping, which is equivalent through the isomorphism to conclusive teleportation.
- Dimension bounds: Teleportation and distinguishability structures provide bounds on the dimension of the state space.
5.3. Distinguishability structures
The principles constrain distinguishability through nondisturbing measurements, pure maximal sets, state-effect duality, spectral structure, and orthomodular faces and projections.
- Nondisturbance: A measurement that extracts no information from a face can be implemented without disturbing that face.
- Maximal distinguishability: Every pure state belongs to a maximal set of perfectly distinguishable pure states, and all such sets for a system have the same cardinality.
- State-effect duality: Pure normalized states correspond one-to-one with pure normalized effects, and Pure Steering connects this distinguishability duality to correlations.
- Informational dimension: The common cardinality d_A, called informational dimension, equals the number of classical messages encodable and perfectly decodable in system A.
- Composite systems: Informational dimension is multiplicative for composites: d_A⊗B = d_A d_B.
- Spectral and lattice structure: The axioms imply spectral decomposition and make the faces of each state space an orthomodular lattice with unique orthogonal projections.
5.4. Interaction between correlation and distinguishability structures
Combining correlation and distinguishability structures produces operational Schmidt decompositions, conjugate-system dimension relations, teleportation bounds, and the state-space dimension equality.
- Schmidt structure: Pure Steering combined with Spectral Decomposition provides an operational analogue of Schmidt bases for pure bipartite states.
- Conjugate systems: Applying this structure to Bell states shows that conjugate systems have the same informational dimension.
- Teleportation bounds: Teleportation constructions relate the maximum Bell-state weight in an invariant-state decomposition to teleportation probability and dimension bounds.
- Dimension reconstruction: State-effect duality and orthogonal representations supply a teleportation lower bound that, combined with earlier bounds, yields the final state-space dimension equality.
5.5. Qubit structures
The principles recover qubit geometry in two dimensions and extend it to superposition, purification constraints, coherent transformations, and equivalence of systems with equal dimension.
- Qubit geometry: All two-dimensional systems have qubit state spaces: their deterministic states form a three-dimensional Euclidean ball.
- Density-matrix representation: The qubit representation yields density matrices, positive effects bounded by the identity, and the Born rule for probabilities.
- Superposition: Ideal Compression extends the qubit superposition principle from two-dimensional systems to systems of arbitrary dimension.
- Purification: A mixture of r perfectly distinguishable pure states has a purification with purifying system B if and only if d_B ≥ r.
- Transformation superposition: Superposition extends to pure transformations with orthogonal supports through the state-transformation isomorphism.
- System equivalence: All systems of the same informational dimension are operationally equivalent through a reversible transformation.
5.6. The density matrix
The reconstruction maps each system’s real state space linearly and one-to-one onto Hermitian matrices, with normalized states corresponding to density matrices. Pure states correspond to rank-one projectors, and the resulting state and test spaces match quantum theory.
- Linear representation: Each system admits a one-to-one linear map from its real state space to d_A × d_A Hermitian matrices.The construction represents real state vectors as real matrices and then transforms them into complex Hermitian matrices.
- Density-matrix construction: The image of normalized states is the set of nonnegative matrices with unit trace, namely density matrices.The representation enforces unit trace through the chosen maximal pure set and defines off-diagonal entries using encoded qubit states.
- Pure states: Pure states are shown to correspond exactly to rank-one projectors.The argument first establishes that pure states yield rank-one projectors and then uses the superposition principle to prove the converse.
- Completion of the reconstruction: Convexity then establishes a one-to-one correspondence between all states and quantum density matrices.The reconstruction also invokes a theorem identifying the theory’s tests with the tests allowed by quantum theory.
6. Conclusions
Quantum theory is reconstructed from six information-processing principles rather than assumptions about particular physical systems. Five principles define standard information theories, while Purification uniquely selects quantum theory and enables control over randomness without inaccessible side information.
- Six principles: The reconstruction starts from six principles governing how information can be processed, not from properties of specific physical systems.The principles are presented as rules concerning information rather than particles or waves.
- Standard information theory: Causality, Purity of Composition, Local Tomography, Perfect State Discrimination, and Ideal Compression define a standard theory of information.These five requirements are shared by quantum and classical information theory.
- Purification: Purification uniquely identifies quantum theory among standard theories of information.It allows every state to be simulated by preparing a pure bipartite state.
- Quantum consequences: Purification brings features such as entanglement, no-cloning, and teleportation while giving the agent control over otherwise hidden side information.The reconstruction’s conclusion characterizes quantum theory as standard information theory with maximal control of randomness.