Composite Systems and Entanglement Basics

Quantum composition is a new postulate, not merely a larger list of classical variables. If systems $A$ and $B$ have Hilbert spaces $\mathcal H_A$ and $\mathcal H_B$, the composite system is represented on $\mathcal H_A\otimes\mathcal H_B$. That tensor-product structure permits product states but also entangled states that cannot be assigned separate pure states for the two subsystems.

This chapter introduces the formal grammar needed elsewhere in Core Formalism. The detailed theory of multipartite systems, identical particles, Fock space, entanglement measures, nonlocality, and operational resource theory belongs to Composite Systems and Entanglement.

Canonical Ownership

Core Formalism is the canonical home for the first composition rule and standard bipartite notation through Composite Systems and Bipartite Systems.

The other Core articles are concise bridges. Their deeper canonical homes are

• Tensor Products of Hilbert Spaces;
• Product States and Entangled States;
• Reduced Density Operators and Partial Trace;
• Schmidt Decomposition;
• Local and Global Observables.

The index below establishes the conceptual sequence and points to those canonical treatments rather than reproducing them.

Tensor-Product Composition

For distinguishable subsystems,

$$\mathcal H_{AB} = \mathcal H_A\otimes\mathcal H_B.$$

If $\{\lvert i\rangle_A\}$ and $\{\lvert j\rangle_B\}$ are orthonormal bases, then

$$\left\{ \lvert i\rangle_A\otimes\lvert j\rangle_B \right\}_{i,j}$$

is a product basis for the composite space. In finite dimensions,

$$\dim\mathcal H_{AB} = (\dim\mathcal H_A) \,(\dim\mathcal H_B).$$

Dimensions multiply because direct sums and tensor products encode different physical structures. A direct sum usually describes alternatives or sectors; a tensor product describes jointly present subsystems or degrees of freedom.

The ordering convention matters. With $A$ listed before $B$, a product basis is ordered according to a declared rule such as lexicographic order. Matrix representations, bit strings, Kronecker products, and partial-trace code must all use the same convention. See Tensor-Product Ordering.

Product and Entangled Pure States

A pure product state factors as

$$\lvert\psi\rangle_{AB} = \lvert a\rangle_A \otimes \lvert b\rangle_B.$$

A pure state is entangled across the chosen bipartition when no such factorization exists. Entanglement is therefore relative to a subsystem decomposition; the same abstract vector space can admit different physically meaningful tensor-product structures.

For a general bipartite pure state,

$$\lvert\psi\rangle_{AB} = \sum_{i,j} C_{ij} \lvert i\rangle_A \lvert j\rangle_B.$$

In finite dimensions, the state is a product state exactly when the coefficient matrix $C$ has rank one. Local basis changes transform $C$ on its two indices but do not change its rank.

The Bell state

$$\lvert\Phi^+\rangle = \frac{ \lvert0\rangle_A\lvert0\rangle_B + \lvert1\rangle_A\lvert1\rangle_B }{\sqrt2}$$

has coefficient-matrix rank two and is entangled. It cannot be factored even though it is written as a superposition of product-basis vectors. This illustrates a crucial distinction: every vector can be expanded in a product basis, but not every vector is itself one product.

Entanglement is not a force or a signal. It is nonfactorization of the state relative to a composition structure, with observable consequences for correlations and local reduced states.

Subsystem Observables

An observable $A$ acting only on subsystem $A$ is represented on the composite space by

$$A\otimes I_B.$$

Similarly, a $B$-local observable is $I_A\otimes B$, while a product observable is $A\otimes B$. Operators on distinct subsystems commute:

$$[A\otimes I_B,I_A\otimes B] = 0.$$

Local measurements can nevertheless have correlated outcomes when the state is not a product. Operator locality does not imply statistical independence.

An interacting Hamiltonian usually has the form

$$H_{AB} = H_A\otimes I_B + I_A\otimes H_B + H_{\mathrm{int}}.$$

The interaction term can generate entanglement from an initially factorized state. The tensor-product rule tells us where these operators act; model-specific dynamics determines what correlations they produce.

Reduced States and Partial Trace

For a composite density operator $\rho_{AB}$, the state that reproduces every $A$-local expectation value is

$$\rho_A = \operatorname{Tr}_B(\rho_{AB}).$$

Its defining property is

$$\operatorname{Tr}_{AB} \!\left[ \rho_{AB}(A\otimes I_B) \right] = \operatorname{Tr}_A(\rho_AA)$$

for every $A$-local operator $A$. In a basis $\{\lvert j\rangle_B\}$,

$$\rho_A = \sum_j {}_B\langle j\rvert \rho_{AB} \lvert j\rangle_B.$$

The resulting operator is independent of which orthonormal basis is used for the calculation.

For the Bell state, tracing out either qubit gives

$$\rho_A = \rho_B = \frac{I_2}{2}.$$

Thus a globally pure state can have mixed subsystem states. For a globally pure bipartite state, mixedness of either reduced state is equivalent to entanglement across that bipartition. For a globally mixed state, a mixed marginal alone does not diagnose entanglement; even a product mixed state can have mixed marginals.

Tracing out a subsystem is not a claim that the subsystem ceased to exist. It is the operation that produces the correct statistics for observables restricted to the retained subsystem.

Schmidt Structure

Every finite-dimensional bipartite pure state has a Schmidt decomposition

$$\lvert\psi\rangle_{AB} = \sum_{r=1}^{R} \sqrt{\lambda_r} \lvert u_r\rangle_A \lvert v_r\rangle_B,$$

where $\lambda_r>0$, $\sum_r\lambda_r=1$, and the two sets of Schmidt vectors are orthonormal. The number $R$ is the Schmidt rank.

The state is a product exactly when $R=1$ and entangled when $R>1$. Its reduced states are

$$\rho_A = \sum_r\lambda_r \lvert u_r\rangle\langle u_r\rvert,$$ $$\rho_B = \sum_r\lambda_r \lvert v_r\rangle\langle v_r\rvert.$$

They therefore share the same nonzero spectrum. The Schmidt coefficients are basis independent, although Schmidt vectors within degenerate coefficient subspaces are not unique. Schmidt decomposition is a bipartite pure-state theorem; it does not directly provide a comparable single-list normal form for arbitrary multipartite or mixed states.

Local Operations and Signaling

A trace-preserving operation performed locally on $B$, with its outcome ignored, cannot change $A$‘s reduced state. A selected outcome on $B$ can change the state assigned to $A$ conditional on that outcome, but identifying the selected subensemble requires the classical record. These statements are the beginning of the no-signaling structure, not a full treatment of Bell nonlocality.

Entanglement can produce correlations stronger than those allowed by local hidden-variable models in suitable experiments, but entanglement and Bell-nonlocal correlations are not synonymous for all mixed states and measurement scenarios. The dedicated composite-systems and foundations volumes own those distinctions.

Page Map

Question	Core page	Canonical scope
How are quantum systems composed?	Composite Systems	canonical Core composition rule
What does the tensor product construct?	Tensor Products	bridge to the dedicated tensor-product chapter
Which pure states factor?	Product States	bridge to separable pure states
Which pure states do not factor?	Entangled States	bridge to canonical entanglement theory
What notation organizes two-part systems?	Bipartite Systems	canonical Core notation and first calculations
What state describes one subsystem?	Reduced States	bridge to reduced density operators
How is a subsystem traced out?	Partial Trace: First Encounter	practical bridge to the canonical derivation
What normal form diagnoses bipartite pure entanglement?	Schmidt Decomposition Overview	theorem preview and route onward
How do local operators act?	Subsystems and Local Observables	bridge to local and global operator theory

These nine articles form the planned chapter.

Suggested Routes

First systematic pass

Read composite systems, tensor products, product states, entangled states, and bipartite systems. Then add reduced states and the partial-trace guide.

Structural route

Read bipartite systems, reduced states, and the Schmidt overview, then continue to Schmidt rank and entanglement entropy in the dedicated volume.

Quantum-information route

Pair the Bell-state examples with Measurement in a Chosen Basis, then continue to qubits, gates, channels, and entanglement protocols.

Many-particle route

After mastering distinguishable subsystems, continue to Indistinguishability and the symmetrization postulate. Do not impose bosonic or fermionic exchange symmetry on the distinguishable-subsystem examples in this chapter.

Composition Sanity Checks

• Declare subsystem order and basis order before forming Kronecker products.
• Check that finite dimensions multiply.
• Verify normalization of composite states and trace one for density operators.
• Test pure-state factorization by coefficient-matrix rank or Schmidt rank.
• Check partial-trace output dimension, Hermiticity, positivity, and trace.
• Confirm local expectation values from both the global and reduced descriptions.
• Distinguish conditioning on a remote outcome from ignoring that outcome.
• For numerical tensor networks or truncated spaces, verify that the chosen factorization matches the physical subsystems.

Common Mistakes

• Using a direct sum where composition requires a tensor product. The two constructions encode different physical structures.
• Changing subsystem order midway through a calculation. Basis indices, bit strings, and Kronecker products then refer to different vectors.
• Calling every product-basis expansion a product state. Factorization, not expansion, is the criterion.
• Equating entanglement with any correlation. Separable mixed states can be correlated.
• Assuming a mixed marginal proves a mixed global state. A pure entangled global state has mixed reduced states.
• Assuming a mixed marginal proves entanglement of a mixed global state. Product mixed states provide counterexamples.
• Treating the partial trace as deleting a physical system. It constructs the state relevant to local observables.
• Using Schmidt decomposition as a universal multipartite or mixed-state normal form. Its simple form is bipartite and pure-state specific.
• Interpreting conditional remote-state changes as faster-than-light signaling. The conditioning record is classical information.

Cross-Links

• States and Representations
• Density Operators and Mixed States
• Measurement and State Update
• Composite Systems and Entanglement
• Tensor-Product Ordering
• Bell States
• Symmetrization Postulate
• Partial Trace formula entry

References

• A. Peres, Quantum Theory: Concepts and Methods, Kluwer, 1995.
• M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information, 10th anniversary ed., Cambridge University Press, 2010.
• J. Watrous, The Theory of Quantum Information, Cambridge University Press, 2018.
• I. Bengtsson and K. Życzkowski, Geometry of Quantum States, 2nd ed., Cambridge University Press, 2017.
• R. Horodecki, P. Horodecki, M. Horodecki, and K. Horodecki, “Quantum entanglement,” Reviews of Modern Physics 81, 865–942, 2009, doi:10.1103/RevModPhys.81.865.