On the Morse index with constraints: an abstract formulation (Q6097702)

In Morse theory, it is essential to study how negative is the symmetric bilinear form involved in the definition of the second variation. The Morse index is defined to be the maximum dimension of a subspace on which the form is negative definite. The authors study in this paper an abstract formulation of this concept in the context of a variational problem with constraints, like in iso-perimetric problems. They examine the index and nullity of a symmetric bounded bilinear form in a Hilbert space and determine quantitatively how they change when restricting to a subspace of a finite co-dimension. Their approach is to interpret constraints as linear functionals on a vector space and the subspace as the intersection of their kernels.