Real Lagrangian Floer Homology

• Real Lagrangian Floer homology is a Floer-theoretic invariant that studies Lagrangian submanifolds with a real structure induced by anti-symplectic involutions.
• It employs a localization spectral sequence to relate ordinary Floer homology with its real counterpart, yielding dimension inequalities similar to Smith theory.
• Applications extend to real Heegaard Floer homology and knot theory, providing new insights into strongly invertible knots and associated torsion invariants.

Real Lagrangian Floer homology is a Floer-theoretic invariant associated with Lagrangian submanifolds carrying a compatible real structure, typically encoded by an anti-symplectic involution. In the exact setting considered by Guth and Manolescu’s framework and its subsequent localization theory, one starts with an exact symplectic manifold $(M,\omega)$, a pair of compact exact Lagrangians $L_0,L_1\subset M$, and an anti-symplectic involution $R:M\to M$ with $L_1=R(L_0)$. The resulting invariant is defined by

$HFR(M,L_0,L_1):=HF(M,L_0,M^R),$

where $M^R$ is the fixed set of $R$, itself a Lagrangian, and $HF(\cdots)$ denotes standard Lagrangian Floer homology (Hendricks, 5 Aug 2025). This formulation places real structures directly inside Lagrangian Floer theory and, in recent work, admits a localization spectral sequence relating ordinary Floer homology to its real counterpart (Hendricks, 5 Aug 2025).

1. Definition and geometric framework

The basic exact setup consists of an exact symplectic manifold $(M,\omega)$, compact exact Lagrangians $L_0,L_1\subset M$, and an anti-symplectic involution $L_0,L_1\subset M$0 such that $L_0,L_1\subset M$1. In this setting, the fixed locus $L_0,L_1\subset M$2 is a Lagrangian, and the real Lagrangian Floer homology is defined by

$L_0,L_1\subset M$3

(Hendricks, 5 Aug 2025). The defining feature is that the second Lagrangian is replaced by the fixed Lagrangian of the involution, so the real structure is incorporated at the level of the ambient symplectic manifold rather than only through a symmetry on a chain complex.

The terminology “real Lagrangian” is also used more broadly for fixed loci of complex conjugation or antiholomorphic involutions. In a toric symplectic manifold $L_0,L_1\subset M$4, the real part $L_0,L_1\subset M$5 is the fixed locus under complex conjugation and is a Lagrangian submanifold (Haug, 2011). In the Fermat quintic threefold, the set of real points

$L_0,L_1\subset M$6

is the fixed locus of complex conjugation and forms a Lagrangian submanifold (Alston, 2010). In monotone Hermitian symmetric spaces of compact type, real forms are fixed point sets of antiholomorphic involutive isometries and are totally geodesic Lagrangian submanifolds (Iriyeh et al., 2011). These examples establish the geometric breadth of real Lagrangians across exact, monotone, toric, Calabi–Yau, and symmetric-space settings.

A broader survey perspective identifies “the real locus of a real algebraic variety” as a basic example of a Lagrangian submanifold (Fukaya, 2011). This suggests that real Lagrangian Floer homology sits naturally at the intersection of symplectic topology, real algebraic geometry, and equivariant or involutive Floer theory.

2. Localization and the spectral sequence

A central recent development is the existence of a localization spectral sequence for real Lagrangian Floer homology in the exact anti-symplectic setting. The main theorem yields a spectral sequence with

$L_0,L_1\subset M$7

and

$L_0,L_1\subset M$8

where $L_0,L_1\subset M$9 is a degree-one variable, identified as the generator of $R:M\to M$0 (Hendricks, 5 Aug 2025).

The key geometric step is that the anti-symplectic involution $R:M\to M$1 is replaced by a symplectic involution on $R:M\to M$2, given by

$R:M\to M$3

which enables the application of the Seidel-Smith/Large localization theorem (Hendricks, 5 Aug 2025). This converts the real problem into a form accessible to existing symplectic-equivariant localization machinery.

At the chain level, the spectral sequence arises from the double complex

$R:M\to M$4

where $R:M\to M$5 is the involution induced from $R:M\to M$6, and the first differential is

$R:M\to M$7

(Hendricks, 5 Aug 2025). The formal resemblance to equivariant constructions is explicit, but the target is real Floer homology rather than only an equivariant refinement of ordinary Floer theory.

The localization viewpoint has immediate structural consequences. The paper states that the spectral sequence provides dimension inequalities between ordinary and real Floer homology, analogous to the classical Smith theory for topological group actions (Hendricks, 5 Aug 2025). This suggests that real Lagrangian Floer homology functions as a localized or fixed-locus-sensitive refinement of standard Lagrangian Floer theory.

3. Real Heegaard Floer homology as a specialization

The same localization construction specializes to real Heegaard Floer homology. Guth and Manolescu constructed $R:M\to M$8 for a $R:M\to M$9-manifold $L_1=R(L_0)$0 equipped with an orientation-preserving involution $L_1=R(L_0)$1 whose fixed set has codimension two; the construction passes to symmetric products of Heegaard surfaces, where $L_1=R(L_0)$2 induces an anti-symplectic involution $L_1=R(L_0)$3 interchanging the $L_1=R(L_0)$4- and $L_1=R(L_0)$5-curves (Hendricks, 5 Aug 2025).

In this setting the localization spectral sequence takes the form

$L_1=R(L_0)$6

and

$L_1=R(L_0)$7

where the sum runs over real $L_1=R(L_0)$8 structures associated to $L_1=R(L_0)$9 (Hendricks, 5 Aug 2025). For branched double covers $HFR(M,L_0,L_1):=HF(M,L_0,M^R),$0 of knots with covering involution $HFR(M,L_0,L_1):=HF(M,L_0,M^R),$1, this reduces to

$HFR(M,L_0,L_1):=HF(M,L_0,M^R),$2

(Hendricks, 5 Aug 2025).

Under the additional hypothesis that there exists a Heegaard diagram with an $HFR(M,L_0,L_1):=HF(M,L_0,M^R),$3-symmetric family of almost complex structures achieving transversality, the first differential is identified explicitly as

$HFR(M,L_0,L_1):=HF(M,L_0,M^R),$4

where $HFR(M,L_0,L_1):=HF(M,L_0,M^R),$5 is the conjugation-induced involution and $HFR(M,L_0,L_1):=HF(M,L_0,M^R),$6 is induced by the involutive diffeomorphism (Hendricks, 5 Aug 2025). The paper emphasizes that this differential differs from previous localization spectral sequences in the “symplectic involution” case by preserving, rather than reversing, the $HFR(M,L_0,L_1):=HF(M,L_0,M^R),$7 structure, allowing for nontrivial spectral sequences in all $HFR(M,L_0,L_1):=HF(M,L_0,M^R),$8 gradings (Hendricks, 5 Aug 2025).

The Heegaard Floer specialization shows that real Lagrangian Floer homology is not confined to abstract symplectic topology. It also enters low-dimensional topology through branched double covers, involutions of $HFR(M,L_0,L_1):=HF(M,L_0,M^R),$9-manifolds, and the structure of $M^R$0-graded Floer invariants.

4. Knot-theoretic extension: strongly invertible knots

The construction extends further to strongly invertible knots in $M^R$1, namely knots admitting an orientation-preserving involution that fixes an axis intersecting the knot in two points (Hendricks, 5 Aug 2025). In that setting one constructs a real Heegaard diagram with two basepoints, both on the fixed set, and defines real knot Floer homology by

$M^R$2

There is then a localization spectral sequence

$M^R$3

converging to

$M^R$4

with

$M^R$5

(Hendricks, 5 Aug 2025).

The paper states that for strongly invertible knots the spectral sequence splits along Alexander gradings and is computable in many cases (Hendricks, 5 Aug 2025). It also records applications to branched double covers and to strongly invertible knots, including new torsion-type knot invariants, $M^R$6-space property, and “Miyazawa-type” invariants (Hendricks, 5 Aug 2025). Since these claims are formulated in the source as consequences of the localization package, they identify knot theory as one of the principal application domains of real Lagrangian Floer methods.

A plausible implication is that the real formalism captures symmetry data not seen by ordinary knot Floer homology alone, because the differentials explicitly involve involutive and conjugation maps rather than only the standard Floer differential.

5. Computational models, clean intersections, and orientations

Real Lagrangian Floer homology draws on computational methods developed in ordinary Lagrangian Floer theory, especially in Morse-Bott and clean-intersection settings. For closed monotone Lagrangians intersecting cleanly, there are two spectral sequences constructed using a Morse-Bott version of Floer homology (Schmäschke, 2016). If $M^R$7 decomposes into clean connected components $M^R$8, the local spectral sequence has first page

$M^R$9

and a global spectral sequence converging to $R$0 (Schmäschke, 2016). The same paper states that for symmetric or real situations, for example when $R$1 is the fixed locus of an anti-symplectic involution, the careful tracking of orientation includes real structure and the effect of conjugation on moduli spaces, and that the general clean-intersection framework specializes directly to make the real theory effectively computable (Schmäschke, 2016).

Orientation issues are a recurrent technical theme. In the clean-intersection framework, full integer-valued Floer homology is defined when $R$2 is a relative spin pair (Schmäschke, 2016). A separate integral construction for oriented closed exact Lagrangians in a Liouville domain removes the standard relatively Pin restriction by replacing global orientations with coefficient systems built from twisted loop spaces (Rezchikov, 2019). That theory applies to real loci such as $R$3, and the paper states that it “applies robustly to real loci” and enables refined computations for real Lagrangians (Rezchikov, 2019). Although this is not itself the Guth–Manolescu notion of real Lagrangian Floer homology, it provides a coefficient-level framework relevant whenever real Lagrangians fail standard orientability hypotheses.

Analytically, gluing theory underlies the construction of Floer differentials and continuation maps. A Hardy space approach proves that $R$4 in the monotone case with minimal Maslov number at least three, proves invariance, and develops the Lagrangian-Floer-Donaldson functor and Seidel homomorphism (Simcevic, 2014). The same source notes that these techniques are directly applicable to real settings, provided suitable transversality is achieved (Simcevic, 2014). This suggests that the modern real theory depends not only on equivariant constructions but also on the same gluing, compactness, and orientation infrastructure that supports standard Floer theory.

6. Examples, calculations, and related real invariants

Several concrete families of real Lagrangians serve as testing grounds for Floer-theoretic calculations. In Fano toric manifolds with minimal Chern number at least $R$5, the real part $R$6 is Lagrangian, and its quantum homology satisfies

$R$7

as $R$8-modules, while also being isomorphic as a ring to the quantum homology of the ambient symplectic manifold: $R$9 (Haug, 2011). The paper states that the wide property implies that the Floer homology of the real Lagrangian $HF(\cdots)$0 is non-vanishing and computable, and mirrors its classical topology (Haug, 2011). These results concern quantum homology rather than the exact localization theory of $HF(\cdots)$1, but they identify a class of real loci where Floer-theoretic behavior is especially rigid.

In monotone Hermitian symmetric spaces of compact type, if $HF(\cdots)$2 are real forms intersecting transversally and both have minimal Maslov number at least $HF(\cdots)$3, then

$HF(\cdots)$4

(Iriyeh et al., 2011). The calculation uses a free $HF(\cdots)$5-action on moduli spaces of holomorphic strips induced by geodesic symmetry, forcing the Floer differential to vanish (Iriyeh et al., 2011). The same paper derives a generalized Arnold-Givental inequality for arbitrary pairs of real forms and proves that a totally geodesic Lagrangian sphere in the complex hyperquadric is globally volume minimizing under Hamiltonian deformations (Iriyeh et al., 2011). These are not framed as real Lagrangian Floer homology in the exact anti-symplectic-localization sense, but they show how real symmetry can simplify the differential and sharpen intersection bounds.

In the Fermat quintic threefold, the real locus $HF(\cdots)$6 and its $HF(\cdots)$7 images under coordinatewise multiplication by fifth roots of unity yield a family of real Lagrangians (Alston, 2010). The calculations proceed via Bott–Morse Floer theory and a spectral sequence whose $HF(\cdots)$8-page is built from the cohomology of clean intersection components, with differentials determined by counting lowest-energy holomorphic strips (Alston, 2010). The real structure is described as crucial for several techniques: a formula for the Maslov index, a formula for the obstruction bundle, a relation between holomorphic strips, discs, and holomorphic spheres, and sign cancellations induced by conjugation (Alston, 2010). The same source states that there is no disc bubbling for these real Lagrangians, so they are unobstructed in the sense of Floer theory (Alston, 2010).

Taken together, these examples show that “real” can enter Floer theory in several non-equivalent but overlapping ways: as a fixed-locus definition $HF(\cdots)$9, as a symmetry on moduli spaces of strips, as a source of vanishing differentials, and as a constraint on orientations and local systems. The recent localization spectral sequence (Hendricks, 5 Aug 2025) unifies part of this landscape by giving a direct passage from ordinary Floer homology to a real variant in the presence of an anti-symplectic involution.

7. Relation to the broader Floer-theoretic landscape

Real Lagrangian Floer homology belongs to a wider network of Floer-theoretic constructions. Standard Lagrangian Floer homology is generated by intersection points and counts pseudo-holomorphic strips, with bounding cochains and $(M,\omega)$0-structures entering in the obstructed case (Fukaya, 2011). The general survey literature explicitly lists the real locus of a real algebraic variety among the fundamental examples of Lagrangian submanifolds (Fukaya, 2011). This places real Lagrangians within the same foundational framework as Fukaya categories, potential functions, and mirror-symmetry constructions.

Other related variants emphasize different geometric questions. Lagrangian Rabinowitz Floer homology studies characteristic chords and relative leaf-wise intersection points, and the source explicitly notes its relevance for “real loci, intersection phenomena, and the effects of twisting and non-exactness on Lagrangian Floer-type invariants” (Merry, 2010). Sheaf-quantization results for exact Lagrangians in cotangent bundles construct complexes of sheaves whose filtered sheaf cohomology coincides with filtered Floer cohomology (Viterbo, 2019). A plausible implication is that analogous sheaf-theoretic descriptions could be sought for real or involutive Floer theories, although such a statement is not made explicitly in the cited sources.

The current state of the subject, as documented by the localization theorem for $(M,\omega)$1, indicates a shift from isolated computations of real loci toward a systematic comparison between ordinary and real theories (Hendricks, 5 Aug 2025). In this formulation, real Lagrangian Floer homology is not merely Floer homology for a special class of Lagrangians; it is a functorial construction that packages anti-symplectic symmetry, fixed-point geometry, and localization into a new invariant with applications to Heegaard Floer homology, branched double covers, and strongly invertible knots (Hendricks, 5 Aug 2025).