Isotony, Additivity, Duality, and Primitive Causality

Isotony, additivity, Haag duality, causal completion, and primitive causality constrain different parts of a local net. Only a few implications hold without extra assumptions. In particular, Einstein causality gives $\mathfrak A(O')\subseteq\mathfrak A(O)'$, whereas Haag duality requires equality; topological flux operators in free Maxwell theory make the distinction concrete.

Required background. Haag–Kastler Nets and Locality supplies the net and causal complement. Quasilocal C*-Algebras and Inductive Limits supplies the generated global algebra.

Helpful background. The Reeh–Schlieder Theorem supplies cyclic and separating vectors. Microcausality and Relativistic Compatibility distinguishes causal commutation from signaling, and Additivity, Haag Duality, and Information Completeness develops the subsystem consequences.

Five distinct net properties

For a represented net $O\mapsto\mathfrak A(O)$, the properties are:

• Isotony: $O_1\subset O_2$ implies $\mathfrak A(O_1)\subset\mathfrak A(O_2)$.
• Additivity: if $O=\bigcup_i O_i$ for an admitted cover, then $\mathfrak A(O)=\bigvee_i\mathfrak A(O_i)$. Weak and strong versions differ in their covers and translated families.
• Einstein causality: $\mathfrak A(O')\subseteq\mathfrak A(O)'$.
• Haag duality: $\mathfrak A(O')=\mathfrak A(O)'$, equivalently $\mathfrak A(O)=\mathfrak A(O')'$ when the region class is closed under causal complement.
• Primitive causality or the time-slice property: if $N$ contains a Cauchy surface for a globally hyperbolic region $D(N)$, then $\mathfrak A(N)=\mathfrak A(D(N))$.

Causal completion is the condition $\mathfrak A(O)=\mathfrak A(O'')$. It concerns propagation within a causal development, whereas duality compares a region with the commutant of its complement. The original net axioms and their independence are discussed in Haag and Kastler 1964, §§ 2–3, pp. 850–856; the free-scalar time-slice construction is explicit in Fewster and Rejzner 2020, § 4.2, pp. 15–17.

Valid implications and missing converses

Assume isotony and that the stated complements remain in the region class. Haag duality implies Einstein causality: if $O_1\subset O_2'$, then

$$\mathfrak A(O_1) \subseteq \mathfrak A(O_2') = \mathfrak A(O_2)'.$$

Duality also implies causal completion wherever duality is available for $O$ and $O''$, because $(O'')'=O'$. Neither converse holds in general. Locality supplies only the inclusion, while causal completion says nothing about whether all operators commuting with the complement were generated locally.

Additivity does not imply duality: smaller contractible regions may generate an algebra that omits a topological operator associated with the union. Conversely, duality does not fix which covers satisfy additivity. Primitive causality is also independent of spacelike commutativity; it is a hyperbolic propagation or time-slice assertion, not a commutator identity.

The dual net

$$\mathfrak A^d(O)=\mathfrak A(O')'$$

exposes a duality defect through $\mathfrak A(O)\subseteq\mathfrak A^d(O)$. Replacing a net by its dual can improve duality on a selected region class, but locality and additivity of the new assignment still require proof. Dualization is not an automatic equivalence of theories.

First application: Maxwell flux and a strict dual inclusion

Electric and Magnetic One-Form Symmetries supplies the physical interpretation of the electric and magnetic flux operators.

Let $R$ be a causally completed ring-like region in four-dimensional Minkowski spacetime, and let $\mathfrak A_{\mathrm{add}}(R)$ be generated by free Maxwell field strengths smeared in contractible subregions of $R$. A regularized electric or magnetic flux can depend only on the boundary of a spanning surface. It commutes with all observables in the causal complement, yet need not be generated by those contractible subregions. Consequently,

$$\mathfrak A_{\mathrm{add}}(R) \subsetneq \mathfrak A_{\mathrm{add}}(R')'.$$

For linked spanning discs $D$ and $\widehat D$ with disjoint thickened boundary tori, one finds

$$4\pi i[ E_\rho(D),H_\sigma(\widehat D)] = \left(\int\rho(\mathbf x)\,d^3x\right) \left(\int\sigma(\mathbf x)\,d^3x\right).$$

This nonzero linking pairing identifies the missing topological class. The surfaces may be deformed while keeping their boundaries, so contractible local generation cannot assign both fluxes in the naive way. The explicit construction and the strict inclusion are given in Schroer 2015, § 3, equations (11)–(15). This establishes a free-field duality failure for multiply connected regions; it is not a claim that Haag duality fails for every region or every gauge net.

Failure test: deriving duality from commutativity

Starting from $[\mathfrak A(O),\mathfrak A(O')]=0$ and replacing the resulting inclusion by equality silently assumes maximality. In the Maxwell ring, the flux operator belongs to $\mathfrak A(O')'$ but not to the additive $\mathfrak A(O)$, so the equality is false even though Einstein causality survives.

The strongest conclusion licensed by locality is

$$\mathfrak A(O)\subseteq\mathfrak A(O')'.$$

To strengthen it, one must specify the region topology, the observable net—including extended or flux operators—and the representation, then prove duality for that assignment.

Independent check

First compute $O'$, $O''$, and the topology of each region. Next generate $\mathfrak A_{\mathrm{add}}(O)$ from the admitted subregions and test every proposed extra operator against $\mathfrak A(O')$. Finally test whether that operator can be generated inside $O$. In the Maxwell example, deformation invariance of the flux and the linked-flux commutator provide independent checks of membership in the dual algebra and nonmembership in the additive algebra.

Exercises

Assuming isotony and Haag duality, prove Einstein causality for $O_1\subset O_2'$.

Solution

Isotony gives $\mathfrak A(O_1)\subseteq\mathfrak A(O_2')$. Haag duality for $O_2$ gives $\mathfrak A(O_2')=\mathfrak A(O_2)'$. Hence every element of $\mathfrak A(O_1)$ commutes with every element of $\mathfrak A(O_2)$. Equality was used only in the second step; locality alone would not recover it.

References

• Fewster, Christopher J., and Kasia Rejzner. “Algebraic Quantum Field Theory—an Introduction.” In Progress and Visions in Quantum Theory in View of Gravity, 1–61. Cham: Birkhäuser, 2020. DOI; Open PDF.
• Haag, Rudolf, and Daniel Kastler. “An Algebraic Approach to Quantum Field Theory.” Journal of Mathematical Physics 5 (1964): 848–861. DOI.
• Schroer, Bert. “Peculiarities of Massive Vector Mesons and Their Zero Mass Limits.” European Physical Journal C 75 (2015): article 365. DOI.