nLab Haag-Kastler axioms



Algbraic Quantum Field Theory

algebraic quantum field theory (perturbative, on curved spacetimes, homotopical)



field theory:

Lagrangian field theory


quantum mechanical system, quantum probability

free field quantization

gauge theories

interacting field quantization



States and observables

Operator algebra

Local QFT

Perturbative QFT


physics, mathematical physics, philosophy of physics

Surveys, textbooks and lecture notes

theory (physics), model (physics)

experiment, measurement, computable physics



The Haag–Kastler axioms (Haag-Kastler 64) (sometimes also called Araki–Haag–Kastler axioms) try to capture in a mathematically precise way the notion of quantum field theory (QFT), by axiomatizing how its algebras of quantum observables should depend on spacetime regions, namely as local nets of observables.

The main point of these axioms is to say that

  1. to every causally closed subset 𝒪X\mathcal{O} \subset X of spacetime XX there is associated a C*-algebra 𝒜(𝒪)\mathcal{A}(\mathcal{O});

  2. for every inclusion 𝒪 1𝒪 2\mathcal{O}_1 \hookrightarrow \mathcal{O}_2 of such spacetime regions there is a corresponding inclusion 𝒜(𝒪 1)𝒜(𝒪 2)\mathcal{A}(\mathcal{O}_1) \hookrightarrow \mathcal{A}(\mathcal{O}_2)

  3. such that this respects composition of inclusions (hence such that 𝒜\mathcal{A} is a functor from the poset of causally closed subsets to C*-algebras)

  4. (causal locality) and such that whenever 𝒪 1,𝒪 2𝒪X\mathcal{O}_1, \mathcal{O}_2 \subset \mathcal{O} \subset X are spacelike separated, then the elements of the corresponding algebras of observables (graded-)commute with each other:

    [𝒜(𝒪 1),𝒜(𝒪 2)]=t0AAAA𝒜(𝒪). [\mathcal{A}(\mathcal{O}_1), \mathcal{A}(\mathcal{O}_2)] = t 0 \phantom{AAAA} \in \mathcal{A}(\mathcal{O}) \,.

Moreover one wants these assignments to behave well with spacetime symmetry.

The formulation of quantum field theory via these axioms has come to be known as algebraic quantum field theory or AQFT, for short. There are various further axioms in the list such as the time slice axiom. The precise details of the list of axioms is in flux as the theory develops.

The Haag-Kastler axioms in their original form aim for the description of non-perturbative quantum field theory on Minkowski spacetime. They could be used to give rigorous proof of structural statements of QFT such as the spin-statistics theorem or the PCT theorem, but examples of interacting field theories in dimensions 4\geq 4 are missing.

There are however variants:

The observation that causal perturbation theory yields quantum observables that satisfy the Haag-Kastler axioms, except that C*-algebras are replaced by formal power series algebras, is due to (Il’in-Slavnov 78, Brunetti-Fredenhagen 00, Brunetti-Dütsch-Fredenhagen 09). See at S-Matrix – Causal Locality and Quantum Observables.


The approch to quantum field theory based on these axioms is often called AQFT : either for axiomatic quantum field theory (since it was among the first attempts to put the edifice of QFT on solid axiomatic grounds) or algebraic quantum field theory (since it amplifies the algebras of local observables over the spaces of states). Neither of these terms is very descriptive. First there is another, dual, axiomatization which does axiomatize the propagation of states – see FQFT – which, second, is also “algebraic” in some sense, even though algebras of observables to not appear directly.

Another common term for these axioms is local quantum field theory (see the title of the standard textbook (Haag)) since, as becomes clear below, they are focused on encoding the locality properties of QFT in terms of the algebras of observables. However, also the core aspect of extended FQFT is all about the notion of locality of QFT.

Therefore neither of the traditional terms for QFT as axiomatized by Haag-Kastler is truly descriptive in that it genuinely distinguishes from the other, the Atiyah-Segal axiomatization by FQFT. What does distinguish the two approaches may be characterized in traditional terminology of quantum theory as follows (Schreiber, SatiSchreiber):


A central difference between the Haag-Kastler axioms and traditionally more widespread formulations of QFT (usually far from being formalized in any way) is the emphasis of the algebra of observables of a QFT (Heisenberg picture) and the de-emphasis of the (Hilbert) spaces of states (Schrödinger picture). This emphasis receives motivation from the the fact that many technical problems of QFT simply disappear when one is not trying to form its spaces of states, while at the same time no real information about the theory is lost.

Examples of technical problems that formulation in terms of spaces of states bring with them are the following:


We formulate the ideas of the core axioms of Haag-Kastler, and their intended physical meaning. For more details see local net of observables.

  1. spacetime locality

    Since the fields in quantum field theory (such as the electromagnetic field) exhibit and are characterized by their local excitations (for instance the value of the electric/magnetic field strength at any point) having effects only locally (the field excitations at two points a finite distance apart do not directly influence each other) the fields over any region of spacetime form a subsystem of the fields of any larger region and in particular of the total system.

    If “quantum mechanical system” is formalized as “C-star algebra” (of observables) then “subsystem” translates to “sub-C *C^*-algebra”. Therefore the above sentence translates into: quantum fields form a copresheaf of C-star algebras on spacetimes whose co-restriction morphisms are monomorphisms.

    In AQFT such is called an isotonic net of algebras .

    There are different approaches to define what kind of spacetime regions the algebras of observables are assigned to, hence different approaches as to what exactly the site is on which the co-presheaf is defined. A common approach is to take all bounded open subsets of Minkowski spacetime. For more general setups see AQFT on curved spacetimes.

  2. causal locality

    If two regions of spacetime are spacelike separated, then there can be no influence between them whatsoever. Not only do the field excitations in one of the two regions not directly influence those in the other region (as per item 1), but they do not even influence indirectly : no waves of excitations (for instance electromagnetic waves: light) can run from one region to a spacelike separated region. Therefore the two subsystems constituted by these two regions accordording to the first point are even independent subsystems . This is called causal locality.

    The formalization of “two independent subsystems” in quantum mechanics is: two subalgebras that commute with each other inside the larger C-star algebra. (And usually one adds: and such that the algebra they generate in the larger algebra is isomorphic to their tensor product.)

    Therefore this translates into the axiom: quantum fields on a spacetime form an isotonic copresheaf of algebras such that the algebras assigned to any two spacelike separated regions commute with each other inside the algebra assigned to any larger region containing these two regions.

  3. spacetime covariance

    The geometric symmetry operations map the algebra of a region onto the algebra of the transformed region.

    (this is not an extra axiom if one defines the site of spacetime regions general enough…)

    In Minkowski spacetime the geometric symmetry group is usually taken to be the Poincaré group, but note that some authors consider subgroups of the full Poincaré group, like the subgroup of translations (Borchers: “Translation group and particle representations in quantum field theory”).

  4. positivity of energy

    An axiom is needed to ensure that only nonnegative energies occur – one possibility is the “spectrum condition”, which says that the spectrum (to be more precise: the support of the spectral measure) of the operator associated with a translation is contained in the closed forward light cone, for all translations.


Structural theorems

It is possible to prove both a spin-statistics theorem and a PCT theorem in the Haag-Kastler approach. The mathematically precise, model independent statements and their proofs are considered to be a major breakthrough of the theory.

Relation to Wightman axioms

Unlike the Wightman axioms, the Haag–Kastler axioms do not need the notion of “field”: the fields in the Wightman axioms are – from the Haag–Kastler point of view – only necessary to describe how the algebras of observables are constructed; any way to consistently construct the net of algebras would suffice.


The original article that introduced these axioms is


  • Rudolf Haag, Discussion des “axiomes” et des propriétés asymptotiques d’une théorie des champs locales avec particules composées, Les Problèmes Mathématiques de la Théorie Quantique des Champs, Colloque Internationaux du CNRS LXXV (Lille 1957), CNRS Paris (1959), 151.

translated to English as:

Textbook accounts include

precursors include

  • Bogolyubov, Logunov, Oksak, Todorov: General principles of quantum field theory. (ZMATH entry)

An online reference page is here:

See also the references at AQFT and at perturbative AQFT.

One of the founding fathers of perturbative quantum field theory wrote:

  • Freeman DysonMissed opportunities, Bulletin of the AMS, Volume 78, Number 5, September 1972 (pdf)

    These axioms, taken together with the axioms defining a C-algebra are a distillation into abstract mathematical language of all the general truths that we have learned about the physics of microscopic systems during the last 50 years. They describe a mathematical structure of great elegance whose properties correspond in many respects to the facts of experimental physics. In some sense, the axioms represent the most serious attempt that has yet been made to define precisely what physicists mean by the words “observability, causality, locality, relativistic invariance,” which they are constantly using or abusing in their everyday speech. [[]] I therefore propose as an outstanding opportunity still open to the pure mathematicians, to create a mathematical structure preserving the main features of the Haag-Kastler axioms but possessing E-invariance instead of P-invariance.

    (on the latter see AQFT on curved spacetimes)

The observation that causal perturbation theory yields quantum observables that satisfy the Haag-Kastler axioms, except that C*-algebras are replaced by formal power series algebras is due to

An introduction into Tomita-Takesaki modular theory:

  • Stephen J. Summers: “Tomita-Takesaki Modular Theory” (arXiv)

and on its role in AQFT:

  • Hans-Jürgen Borchers, On Revolutionizing of Quantum Field Theory with Tomita’s Modular Theory, ESI Preprint 773 (1999) [pdf]

A discussion of how the Haag-Kastler axioms (those concerning locality) follow from an extended FQFT with Lorentzian structure is in

  • Urs Schreiber, AQFT from nn-functorial QFT , Comm. Math. Phys., Volume 291, Issue 2, pp.357-401 (pdf)

A discussion putting the state of the art of the AQFT-axiomatization in context with that of the FQFT-axiomatization is in

A discussion of perturbation theory and renormalization in terms of the Haag-Kastler axioms is in

Haag's theorem and its meaning and implication is discussed thoroughly in

  • John Earman, Doreer Fraser, Haag’s theorem and its implications for the foundations of quantum field theory (pdf)

Last revised on March 30, 2023 at 07:30:34. See the history of this page for a list of all contributions to it.