2.6 H F ∗ and D b ℱ ( M ) for immersed Lagrangians [03NI]
Original official author HTML, exact retained edition. Historical TeX conversion verdicts remain unchanged. Cited-edition alignment and mathematical self-containment are not assessed.
For the programme of §3, it will be necessary to enlarge the derived Fukaya category of a Calabi–Yau -fold to include immersed Lagrangians. As a first step in doing this, Akaho and the author [2] explain how to generalize the Lagrangian
Floer cohomology of Fukaya, Oh, Ohta and Ono [20] from
embedded Lagrangians to immersed Lagrangians with transverse
self-intersections. We now explain some of the main ideas
in [2].
Let be a Calabi–Yau -fold, and a Lagrangian brane in . As in §2.5, in the embedded case [20], a bounding cochain for is a singular -chain (or equivalence class of such chains), satisfying an equation (2.15) involving virtual chains for moduli spaces of -holomorphic discs in with boundary in .
In the immersed case [2], if has transverse
self-intersections, a bounding cochain for consists of
two pieces of data: a chain in
as above, and also, for each point at which two local
sheets of intersect transversely with
, an element
(2.17)
where we write for the restriction of to the local sheets . These must satisfy equations involving virtual chains for moduli spaces of -holomorphic discs in with boundary in , but now these -holomorphic discs can be polygons with ‘corners’ at self-intersection points of .
For example, suppose are embedded, transversely
intersecting Lagrangian branes in . Then is an immersed Lagrangian brane in . A bounding cochain for could consist of , where for are embedded bounding cochains for , together with elements in (2.17) for with or which encode how the objects in are glued together to make . For instance, if we have a distinguished triangle in
then the for with
form a chain in representing , and otherwise.
Note that is represented by with , but to define a bounding cochain we require that . This can be achieved by multiplying by for , which does not change up to isomorphism in .
Figure 2.3: -holomorphic ‘teardrop’ making immersed obstructed
The new cause of obstructions to for immersed Lagrangians
with transverse self-intersections is ‘teardrop-shaped’
-holomorphic discs of the form shown in Figure
2.3, with one corner at , and with
, where are the local sheets of
intersecting at . As the moduli space of
such discs has virtual dimension 0. Such only obstruct
if they have ‘small area’ (that is, is
smaller than the areas of other relevant curves with boundary in
). Thus we deduce an analogue of Lemma 2.21:
Lemma 2.23.
Suppose is a Calabi–Yau -fold and
is an immersed Lagrangian brane in with only
transverse self-intersections. If and has no
self-intersection points with or
where are the local sheets of at then has
unobstructed.
In §2.5 we explained that if is a smooth family of embedded Lagrangian branes with the Hamiltonian isotopic and the locally constant in , and is a bounding cochain for , then extends to bounding cochains for by a kind of ‘parallel transport’, and the isomorphism class of in is independent of .
In the immersed case, things are more complicated. Firstly, there
are two notions of Hamiltonian isotopy. Let for
be a smooth family of compact, immersed Lagrangians in
, where we also write as . We call the family
globally Hamiltonian isotopic if for
is Hamiltonian flow by for some smooth
. We call the family locally Hamiltonian
isotopic if for is Hamiltonian
flow by some smooth , where there may exist with but , so that does not descend from to .
There is a notion of ‘parallel transport’ for bounding cochains along such local Hamiltonian isotopies, but it does not work all the time. Suppose for simplicity that has only transverse self-intersections for all . Then the self-intersection points of in depend smoothly on , so we can write for the intersection of local sheets , of for , where depend smoothly on . Then is independent of .
Let be a bounding cochain for depending
smoothly on , with in . Then evolves in time by a kind of ‘parallel transport’. Let
be as above with . As above, includes an element . Since the local systems are locally constant in , we can identify the fibres for , and the fibres for , and so regard as being independent of . Then is not constant, but
evolves by
(2.18)
Integrating this over gives
(2.19)
Suppose , and write with , , and .
Then
(2.20)
Thus , required for to
be a bounding cochain by (2.17), if and only if
(2.21)
Hence we have the following situation, which will be important in §3.4. Let be a local Hamiltonian isotopy of Lagrangian branes in , and a bounding cochain for . We may extend to a family of bounding cochains for for some , so that in . But at time we may cross a ‘wall’ when the l.h.s. of (2.21) becomes zero, and we cannot define for . Either for may have obstructed, or a bounding cochain may exist but in .
The Lagrangian -principle, due to Gromov [23, p. 60-61]
and Lees [46], says that two Lagrangians are locally
Hamiltonian isotopic in if and only if they are homotopic
in a weak sense, which can be well understood using homotopy theory,
and is weaker than isomorphism in . So we should expect local Hamiltonian isotopies to connect Lagrangians with unobstructed and with obstructed, or to connect non-isomorphic Lagrangians in .
Remark 2.24.
As in §2.5, in the embedded case, the Fukaya category has objects for an embedded Lagrangian brane and a bounding cochain, but the derived Fukaya category has objects twisted complexes, consisting of objects in together with for satisfying an equation.
In the immersed case, we can regard such a twisted complex as a single object in , where is the disjoint union , considered as a single immersed Lagrangian, , and is a bounding cochain for built from and for . Thus there is no need to add twisted complexes, and we can suppose all objects of are of the form .
The idempotent completion of as in §2.5 could still include objects which are direct summands of some , but do not have a good geometric interpretation. However, in §3.1 we will conjecture that in the situation we are interested in, is already idempotent complete, so that we can take all objects of to be of the form .