Why Continuum QFT Does Not Factorize Naively

A sharp spatial cut in continuum QFT usually does not produce $\mathcal H=\mathcal H_A\otimes\mathcal H_{A^c}$ with $\mathfrak A(A)=\mathcal B(\mathcal H_A)$. The obstruction is structural: local algebras are typically type III and ultraviolet correlations persist across an arbitrarily sharp boundary. Tensor products remain legitimate after a regulator, for selected modes, or through a split inclusion with a nonzero collar, but each construction answers a different question.

Required background. Review tensor products and local observable nets. Helpful background. Reflection-positive lattice regulators supply a controlled type-I comparison, and restricted states separates restriction from partial trace.

Why the naive tensor factor fails

If a bipartition were represented by

$$\mathcal H=\mathcal H_A\otimes\mathcal H_B, \qquad \mathfrak A(A)=\mathcal B(\mathcal H_A)\otimes1,$$

then $\mathfrak A(A)$ would be a type-I factor: it would possess minimal projections and the ordinary operator trace. Relativistic local algebras in the vacuum representation are instead generically type III. They have no nonzero finite projections and no faithful normal trace. This conclusion was already visible in Araki’s analysis of free fields Araki 1964, pp. 956–965.

The same mismatch appears physically. A sharp boundary admits field modes at arbitrarily short wavelength on both sides. Vacuum correlations across the cut generate a regulator-dependent area divergence in the entropy. If an exact local tensor factor existed with an ordinary reduced density operator, this ultraviolet structure would have to be accommodated by a trace-class object; the type-III algebra says that the intrinsic continuum description is different.

This does not mean that a region has no state or no entanglement. State restrictions, relative entropy, modular data, mutual information for separated regions, and operational correlation tests remain meaningful.

The structural map places Why Continuum QFT Does Not Factorize Naively among sharp local algebras, split inclusions, and regulated or operational substitutes.

A causally complete region determines a sharp local algebra, usually type III. A nonzero split collar or an explicit cutoff, mode selection, or probe model can instead supply a type-I realization; these alternatives enable ordinary density matrices but retain different physical approximations. Schematic, not to scale.

Three legitimate substitutes

1. Lattice or mode regulator. A finite set of oscillators factorizes. Quantities such as $S(\rho_A)$ are defined at cutoff $a$, and their divergences, universal combinations, or limits must be analyzed before a continuum claim is made.
2. Split inclusion. For $O_1\Subset O_2$, a type-I factor $\mathfrak N$ may interpolate as $\mathfrak A(O_1)\subset\mathfrak N\subset\mathfrak A(O_2)$. The collar $O_2\setminus\overline O_1$ prevents a sharp cut and permits a tensor-product realization.
3. Selected wavepackets or detectors. A finite collection of modes or probes defines a finite operational subsystem. Its algebra is not the full algebra of a geometric region.

These are not interchangeable. A lattice factorization depends on discretization and edge prescription; a split factor depends on a collar and is generally noncanonical; a mode subsystem can be nonlocal in position space.

Oscillator-chain diagnostic

For a chain with spacing $a$, the Hilbert space factorizes by sites and a block $A$ has a reduced covariance matrix. Taking $a\to0$ at fixed physical length can yield convergent correlators for smeared observables while $S_A(a)$ diverges. The controlled statement is therefore

$$\bigl(\mathcal H_A(a),\rho_A(a),S_A(a);\ \text{geometry, boundary rule, limit}\bigr),$$

not an unqualified continuum $\rho_A$. Varying $a$ and an independent boundary buffer distinguishes cutoff artifacts from split-distance effects.

A decisive check

Ask which operations survive removal of the regulator. Expectation values of appropriately smeared local observables may converge. Partial traces, basis-dependent mode entropies, and minimal projections generally do not become intrinsic operations on the sharp local algebra. Any continuum statement must name the surviving algebraic quantity or controlled universal combination.

Before applying this result, use the validity map to keep the algebra, state, operation class, resources, and approximation fixed.

Every local-information claim must specify the represented algebra and state, the allowed operations and resource support, and any split collar or regulator. The dashed lower boxes show what fails when the algebra, protocol, or limiting prescription is left implicit. Schematic.

References

• Araki, Huzihiro. “Type of von Neumann Algebra Associated with Free Field.” Progress of Theoretical Physics 32 (1964): 956–965. DOI.

Further reading

• Buchholz, Detlev, Claudio D’Antoni, and Klaus Fredenhagen. “The Universal Structure of Local Algebras.” Communications in Mathematical Physics 111 (1987): 123–135. DOI.