I have heard that both non-quantum and quantum physics theories can be mathematically…
- User
I have heard that both non-quantum and quantum physics theories can be mathematically formulated using C-star algebras (the non-quantum ones happen to be commutative). This applies both to fundamental interactions and to imprecise measurements and subjective lack of information. Let us try to investigate that claim. In non-quantum physics, there is an important process of Bayesian update, when prior information is transformed into posterior information, based on likelihoods of observations. Describe this transformation as formulas using elements of some C-star algebra, then drop the assumption on commutativity. Do we get a unique quantum transformation, or are multiple formulas possible (only becoming equivalent in commutative case)? Does the quantum transformation correspond to something known and specific (for example wave function collapse), or are additional philosophical assumptions needed (for example related to interpretations of quantum mechanics)?
- Assistant
This is a sophisticated question that touches on fundamental connections between classical and quantum theories through the lens of operator algebras. Let me work through this systematically.
Classical Bayesian Update in C*-Algebra Framework
In classical probability theory, we can formulate Bayesian updating using commutative C*-algebras. Consider:
- Algebra: A = L∞(Ω, μ), the algebra of essentially bounded measurable functions on a probability space
- States: Positive linear functionals ρ: A → ℂ with ρ(1) = 1 (representing probability measures)
- Events: Projections (characteristic functions) E ∈ A
The Bayesian update formula becomes:
ρ_posterior(a) = ρ(E · a · E) / ρ(E)This gives the conditional expectation E[a|E] in the classical sense.
Quantum Generalization: Multiple Pathways
When we drop commutativity and move to non-commutative C*-algebras (typically B(H), bounded operators on Hilbert space), we encounter multiple possible generalizations, not a unique one:
1. Lüders Rule (von Neumann-Lüders)
ρ → √E ρ √E / Tr(√E ρ √E)where E is a positive operator (effect) and ρ is a density operator.
2. Born Rule with Projection
For projective measurements with projection P:
ρ → P ρ P / Tr(P ρ P)3. General POVM Updates
For measurement operators {M_k}:
ρ → M_k ρ M_k† / Tr(M_k ρ M_k†)4. Other Possibilities
- Holevo-type updates
- Continuous monitoring formulations
- Information-theoretic approaches
Key Insights
Non-uniqueness: Unlike the classical case, there isn't a single canonical quantum generalization. The "correct" choice depends on the physical context and measurement apparatus.
Physical Correspondence: These formulations do correspond to known quantum phenomena:
- The projection-based updates relate to what's often called "wave function collapse"
- POVM updates describe generalized measurements
- Different rules can give different predictions for sequential measurements
Interpretational Considerations
The connection to "wave function collapse" requires additional philosophical assumptions:
- Copenhagen interpretation: The mathematical update represents literal physical collapse
- Many-worlds: The update represents decoherence and branching, not collapse
- Consistent histories: The update reflects the choice of consistent framework
- QBism: The update represents subjective belief revision of the agent
The mathematical formalism itself is interpretation-neutral - it describes how to update quantum states upon measurement, but doesn't specify the ontological meaning of this update.
Bottom Line
The non-commutative generalization yields multiple valid quantum update rules that become equivalent in the commutative limit. While these correspond to known quantum phenomena, their physical interpretation depends critically on one's chosen quantum mechanical worldview. The mathematics underdetermines the physics and philosophy.
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.