Lawson Self-Duality in Domain Theory

• Lawson self-duality is a duality phenomenon where a domain, enriched category, or frame is transformed via a contravariant functor that exchanges compact/open data and recovers the original structure after a double dual.
• It unifies various settings including classical continuous dcpos with Scott-open filters, Hofmann–Lawson duality of frames and spaces, and enriched frameworks using quantale categories.
• Recent developments extend the concept to graded Stone-type dualities and geometric contexts, showcasing mirror mechanisms and refined self-dual structures in homological theories.

Lawson self-duality denotes a family of duality phenomena in which a Lawson-type structure is sent contravariantly to a dual object of the same or closely related kind, and a second application returns the original object up to natural isomorphism. In its classical domain-theoretic form, the category of continuous dcpos with Scott-open filter reflecting maps is self-dual; in later work the same pattern is extended to quantale-enriched categories, continuous frames and locally compact sober spaces, two-sorted pointfree structures, graded Stone-type dualities, and Lawson compact algebraic L-domains [(Hofmann et al., 2010); (Bezhanishvili et al., 2022); (Abbadini et al., 24 Jul 2025)].

1. Classical domain-theoretic form

In the classical setting, a dcpo $X$ is continuous when every $y\in X$ is the directed supremum of the elements way below it, where $x\ll y$ means that whenever $D\subseteq X$ is directed and $\bigvee D\ge y$, some $d\in D$ satisfies $x\le d$. The Scott topology $\sigma(X)$ consists of upsets inaccessible by directed joins, and the Lawson topology $\lambda(X)$ is the join of the Scott topology with the lower topology; for continuous dcpos it is compact Hausdorff. Against this background, Lawson’s theorem states that the category of continuous dcpos with Scott-open filter reflecting maps is self-dual (Hofmann et al., 2010).

A standard reformulation replaces a continuous domain $D$ by the dcpo of its Scott-open filters, ordered by reverse inclusion. The relation

$y\in X$0

packages the duality into a two-sorted order-theoretic object. In this form, Lawson self-duality exchanges the domain and its Scott-open filters, and the Lawson topology appears as the topological expression of the simultaneous presence of Scott-open and compact-saturated data (Abbadini et al., 24 Jul 2025).

The classical theorem is therefore not merely a duality between a space and an algebra of opens. It is a self-duality of a category whose morphisms are chosen to reflect Scott-open filters, so that the approximation structure encoded by $y\in X$1 is recovered on the dual side (Hofmann et al., 2010).

2. Quantale-enriched generalization

A major extension replaces ordered sets by $y\in X$2-categories, where $y\in X$3 is a commutative unital quantale and a $y\in X$4-category $y\in X$5 carries hom-values $y\in X$6 satisfying enriched reflexivity and transitivity. For a saturated class $y\in X$7 of modules, a $y\in X$8-category is $y\in X$9-cocomplete when the Yoneda embedding $x\ll y$0 has a left adjoint $x\ll y$1, and it is $x\ll y$2-continuous when $x\ll y$3 itself has a left adjoint. The corresponding way-below module $x\ll y$4 is the enriched analogue of the usual way-below relation (Hofmann et al., 2010).

The decisive construction is the $x\ll y$5-category of open modules,

$x\ll y$6

and the associated evaluation map

$x\ll y$7

For $x\ll y$8-cocomplete, $x\ll y$9-continuous, open module determined $D\subseteq X$0-categories, the assignment $D\subseteq X$1 defines a contravariant functor, and Theorem 3.9 proves that the category $D\subseteq X$2-Dom is self-dual, with $D\subseteq X$3 giving a natural isomorphism $D\subseteq X$4 (Hofmann et al., 2010).

This framework subsumes several familiar cases. When $D\subseteq X$5 and $D\subseteq X$6, $D\subseteq X$7-Dom is exactly the category of continuous dcpos with Scott-open filter reflecting maps, so the enriched theorem reduces to classical Lawson self-duality. The same formalism also covers completely distributive complete lattices, Yoneda-complete quasi-metric spaces, generalized metric and ultrametric examples, and the absolute case of totally continuous cocomplete $D\subseteq X$8-categories (Hofmann et al., 2010).

3. Frames, spaces, and two-sorted reformulations

A parallel line of development situates Lawson self-duality within pointfree topology. Hofmann–Lawson duality identifies continuous frames with locally compact sober spaces; one formulation is that

$D\subseteq X$9

and that this duality can be derived from Priestley duality by passing through suitable Priestley-space models of frames (Bezhanishvili et al., 2022). In this setting, continuous, stably continuous, stably compact, and compact regular frames correspond respectively to locally compact sober, stably locally compact, stably compact, and compact Hausdorff spaces (Bezhanishvili et al., 2022).

Bice’s unification places Hofmann–Lawson duality inside a broader dual equivalence between $\bigvee D\ge y$0-bases of core compact sober spaces and $\bigvee D\ge y$1-distributive $\bigvee D\ge y$2-predomains. Here the way-below or compact-containment relation is abstracted to an auxiliary, interpolative, approximating, $\bigvee D\ge y$3-preserving relation $\bigvee D\ge y$4. Hofmann–Lawson duality reappears as the special case in which the algebraic object is a complete lattice and $\bigvee D\ge y$5 is the way-below relation $\bigvee D\ge y$6 (Bice, 2020).

A more explicitly self-dual account is provided by the theory of ko-spaces and bi-dcpos. A ko-space $\bigvee D\ge y$7 treats compact saturated sets and open sets as primitive data, subject to compactness and cocompactness axioms. Its de Groot dual swaps the order and exchanges $\bigvee D\ge y$8 with complements of $\bigvee D\ge y$9. On the pointfree side, a bi-dcpo $d\in D$0 is a two-sorted polarity in which $d\in D$1 has codirected meets, $d\in D$2 has directed joins, and the relation $d\in D$3 satisfies double compactness, weakening, and extensionality. Swapping the two sorts gives another bi-dcpo, and for bicontinuous bi-dcpos this self-duality specializes to classical Lawson self-duality for continuous domains (Abbadini et al., 24 Jul 2025).

The same paper proves that distributive bi-dcpos, distributive embedded bi-dcpos, and ko-spaces form equivalent categories, each equipped with a self-duality. It follows that Lawson self-duality and de Groot self-duality are two manifestations of the same open/compact symmetry, rather than unrelated constructions (Abbadini et al., 24 Jul 2025).

A locally small variant of Hofmann–Lawson duality replaces ordinary topological spaces by locally small spaces $d\in D$4, where $d\in D$5 is a smopology and $d\in D$6 its generated topology. The spectral adjunction sends $d\in D$7 to $d\in D$8 and $d\in D$9 to $x\le d$0; restricted to locally compact sober locally small spaces and continuous frames with a distinguished sup-generating sublattice, this yields a Hofmann–Lawson duality for locally small spaces (Piękosz, 2020).

4. Stone-type and graded extensions

Two recent directions push Lawson duality into more explicitly Stone-like settings. The first is graded Lawson–Stone duality. For a group $x\le d$1, the category $x\le d$2 of $x\le d$3-graded-Boolean inverse $x\le d$4-semigroups is dually equivalent to the category $x\le d$5 of $x\le d$6-graded Hausdorff ample topological groupoids. The duality is implemented by the ultrafilter groupoid $x\le d$7 and the homogeneous compact-slice semigroup $x\le d$8, with natural isomorphisms

$x\le d$9

When the grading is trivial, this recovers Lawson’s ungraded noncommutative Stone duality (Hazrat et al., 7 Oct 2025).

The second is a Stone duality for Lawson compact algebraic $\sigma(X)$0-domains. This work introduces finitely disjunctive distributive lattices (FDD-lattices), characterized by the facts that every element is a finite join of co-primes and that the meet of any two co-primes decomposes into a finite disjoint join of co-primes. The resulting categories

$\sigma(X)$1

of FDD-lattices and Lawson compact algebraic $\sigma(X)$2-domains with spectral maps are dually equivalent. The functors are

$\sigma(X)$3

and the bidual identifications are given by

$\sigma(X)$4

Here the Lawson topology is essential: compactness of the Lawson topology is characterized by mub-completeness and finiteness of minimal upper bounds of compact elements, and these finiteness conditions are mirrored exactly by the FDD axioms on the lattice side (Hou et al., 13 Feb 2026).

These Stone-type results show that Lawson self-duality often takes the form of a dual equivalence between two categories together with canonical double-dual isomorphisms, rather than an endofunctor on a single category (Hazrat et al., 7 Oct 2025, Hou et al., 13 Feb 2026).

5. Other Lawson-named self-dual structures

The phrase also appears in distinct geometric and homological settings. For abelian varieties, Lawson homology admits a Fourier–Mukai transform

$\sigma(X)$5

satisfying the inversion theorem

$\sigma(X)$6

Together with the eigenspace decomposition

$\sigma(X)$7

this yields isomorphisms

$\sigma(X)$8

and Proposition 8.1 shows that Friedlander–Lawson duality respects the same $\sigma(X)$9-grading. In this context, “Lawson self-duality” refers to a refined self-dual structure on Lawson homology and morphic cohomology, parallel to Beauville’s decomposition on Chow groups (Hu, 2011).

In $\lambda(X)$0-geometry, a Harvey–Lawson configuration $\lambda(X)$1 with $\lambda(X)$2 determines canonical vector fields

$\lambda(X)$3

When the orthogonal six-plane fields are integrable, the hypersurfaces $\lambda(X)$4 and $\lambda(X)$5 inherit almost Calabi–Yau structures related by explicit identities exchanging symplectic and complex data. The exposition describes this as a Lawson self-dual mechanism in which a Harvey–Lawson 3-fold inside a $\lambda(X)$6-manifold canonically generates a mirror pair of Calabi–Yau manifolds (Akbulut et al., 2015).

These uses are mathematically separate from the domain-theoretic Lawson dualities. They share the themes of canonical pairing, mirror exchange, and double-dual recovery, but they do not rely on the Lawson topology or on Scott-open filters [(Hu, 2011); (Akbulut et al., 2015)].

6. Conceptual scope and common misconceptions

Across the literature, “Lawson self-duality” does not denote a single theorem. It may mean a contravariant self-equivalence of a category, as in continuous dcpos or $\lambda(X)$7-Dom; a dual equivalence between two different categories equipped with natural biduality maps, as in graded Lawson–Stone duality or the FDD-lattice/Lawson-compact-domain correspondence; or a refined self-dual structure internal to a geometric or homological theory [(Hofmann et al., 2010); (Hazrat et al., 7 Oct 2025); (Hou et al., 13 Feb 2026); (Hu, 2011)].

The most stable structural motif is the presence of two complementary forms of data—Scott-open filters and elements, opens and compact saturated sets, open modules and enriched objects, compact-open lattices and point spaces—together with a contravariant passage that exchanges them and a canonical comparison with the double dual. In the bi-dcpo framework this symmetry becomes explicit as the passage

$\lambda(X)$8

and on the spatial side it appears as de Groot duality for ko-spaces (Abbadini et al., 24 Jul 2025).

The following table summarizes the main settings recorded in the cited works.

Setting	Objects	Duality statement
Classical Lawson duality	Continuous dcpos with Scott-open filter reflecting maps	Self-dual category (Hofmann et al., 2010)
Enriched Lawson duality	$\lambda(X)$9-Dom	Self-dual category via $D$0 (Hofmann et al., 2010)
Hofmann–Lawson duality	Continuous frames / locally compact sober spaces	Dual equivalence (Bezhanishvili et al., 2022)
Bi-dcpo symmetry	Bicontinuous bi-dcpos	Self-duality by swapping $D$1 and $D$2 (Abbadini et al., 24 Jul 2025)
Graded Lawson–Stone duality	$D$3 and $D$4	Dual equivalence with double-dual isomorphisms (Hazrat et al., 7 Oct 2025)
Lawson compact algebraic $D$5-domains	$D$6 and $D$7	Dual equivalence with $D$8 and $D$9 (Hou et al., 13 Feb 2026)

A plausible implication is that the unifying content of Lawson self-duality is not a fixed formula but an invariant pattern: approximation data and compact/open data are organized so that a dual object can be formed functorially, and the original object is recovered canonically after a second passage to the dual. In the domain-theoretic and pointfree-topological literature, this pattern is the one most directly associated with the name “Lawson self-duality” [(Hofmann et al., 2010); (Abbadini et al., 24 Jul 2025)].