4 MARKUS BANAGL AND DOMINIK J. WRAZIDLO
In [6], the first author derived Verdier-Riemann-Roch type formulae for the Gysin restric-
tion of both the topological characteristic classes L∗of Goresky and MacPherson [21] and the
Hodge-theoretic intersection Hirzebruch characteristic classes IT1∗of Brasselet, Sch¨
urmann
and Yokura [9]. For an introduction to characteristic classes of singular spaces via mixed
Hodge theory in the complex algebraic context see Sch¨
urmann’s expository paper [32]. The
formulae in [6] have the potential to yield new evidence for the equality of the characteristic
classes L∗and IT1∗for pure-dimensional compact complex algebraic varieties, as conjectured
by Brasselet, Sch¨
urmann and Yokura in [9, Remark 5.4]. Cappell, Maxim, Sch¨
urmann and
Shaneson proved the conjecture in [11, Cor. 1.2] for orbit spaces X=Y/G, with Ya projec-
tive G-manifold and Ga finite group of algebraic automorphisms. They also showed the con-
jecture for certain complex hypersurfaces with isolated singularities [10, Theorem 4.3]. The
conjecture holds for simplicial projective toric varieties as shown by Maxim and Sch¨
urmann
[29, Corollary 1.2(iii)]. Furthermore, the conjecture was established by the first author in [5]
for normal connected complex projective 3-folds Xthat have at worst canonical singularities,
trivial canonical divisor, and dimH1(X;OX)>0. Generalizing the above cases, Fern´
andez de
Bobadilla and Pallar´
es [17] proved the conjecture for all compact complex algebraic varieties
that are rational homology manifolds. In ongoing work with J¨
org Sch¨
urmann, we apply the
methods developed in the present paper to prove the ambient version of the conjecture for
a certain class of subvarieties in Grassmannians. This class includes all Schubert varieties.
Since the homology of Schubert varieties injects into the homology of ambient Grassmanni-
ans, this would imply the conjecture for all Schubert varieties. Furthermore, we shall clarify
how other algebraic characteristic classes such as Chern classes fit into the framework.
Acknowledgement. We thank J¨
org Sch¨
urmann and Laurentiu Maxim for insightful com-
ments on an earlier version of the paper.
Notation. (Co)homology groups will be with rational coefficients unless otherwise stated.
Complex algebraic varieties are not assumed to be irreducible. If Xand Yare subvarieties of
a smooth variety W, then the symbol X∩Ydenotes the set theoretic intersection of Xand Y
in W. Given a set S, the Kronecker delta of two elements a,b∈Sis δab =1 for a=b, and
δab =0 else. The Kronecker product h−,−i is defined by hξ,xi=ε∗(ξ∩x)∈Qfor classes
ξ∈H∗(A)and x∈H∗(A)on a topological space A, where ε∗:H∗(A)→H∗(pt) = Qdenotes
the augmentation map induced by the unique map A→pt.
2. WHITNEY TRANSVERSALITY
Let Wbe a smooth manifold. Recall that a Whitney stratification of a closed subset Z⊂W
is a certain decomposition of Zinto locally closed smooth submanifolds of Wsuch that the
pieces of this decomposition fit together via Whitney’s conditions A and B (for details, see
e.g. [23, Section 1.2, p. 37]). Two Whitney stratified subsets Z1,Z2⊂Ware called transverse
if every stratum of Z1is transverse to every stratum of Z2as smooth submanifolds of W. The
transverse intersection Z1∩Z2then has a canonical Whitney stratification in Wwhose strata
are given by the intersections of the strata of Z1and Z2.
Pure-dimensional closed subvarieties of a nonsingular complex algebraic variety can al-
ways be Whitney stratified by strata of even codimension. In the context of this paper, we
call two such subvarieties Whitney transverse if they can be equipped with transverse Whit-
ney stratifications. The present section collects various results about Whitney stratified spaces
that will be applied to Whitney transverse subvarieties later on in the paper.
Lemma 2.1. Let Z1,Z2⊂W be Whitney stratified subspaces of a smooth manifold W that
are transverse to each other. If U ⊂W is a smooth submanifold that is open as a subset of
Z2, then every stratum of Z1is transverse to U in W .