Knapp-Stein type intertwining operators for symmetric pairs. II. The translation principle and intertwining operators for spinors (Q2280423)

If \(\sigma\) is an involution of a group \(G\) and \(H\) is an open subgroup of the group \(G^\sigma\) of fixed points of \(\sigma\), then \((G,H)\) forms a symmetric pair. If \(P=MAN\) and \(P_H=M_HA_HN_H\) are parabolic subgroups of \(G\) and \(H\) respectively, \(\pi_\nu=\mathrm{Ind}_P^G(1\otimes e^\nu\otimes1)\) and \(\pi_{\nu'}=\mathrm{Ind}_{P_H}^H(1\otimes e^{\nu'}\otimes1)\) are the spherical principal series of \(G\) and \(H\) respectively, then, in Part I of this paper [Adv. Math. 294, 256--306 (2016; Zbl 1335.22014)], \textit{J. Möllers} et al. constructed explicitly a family of intertwining operators, symmetry breaking operators in \(\Hom_H(\pi_\nu|_H,\tau_{\nu'})\) as a generalization of the classical Knapp-Stein intertwining operators. They proved meromorphic dependence on the parameters in general and generic uniqueness in some special cases. In the paper under review, the authors further extend these results to the case of vector-valued principal series representations and establish many of the properties for arbitrary vector bundles, in particular the meromorphic dependence on the parameters. If \(G\) is a real reductive group and \(H\subseteq G\) is its reductive subgroup, then for parabolic subgroups \(P=MAN\subseteq G\) and \(P_H=M_HA_HN_H\subseteq H\) the authors form the generalized principal series representations \(\pi_{\xi,\lambda}=\mathrm{Ind}_P^G(\xi\otimes e^\lambda\otimes\mathbf{1})\) and \(\pi_{\eta,\nu}=\mathrm{Ind}_P^G(\eta\otimes e^\nu\otimes\mathbf{1})\), where \(\xi\) and \(\eta\) are finite-dimensional representations of \(M\) and \(M_H\), and \(\xi\) and \(\nu\) are elements of the complexified duals of the Lie algebras of \(A\) and \(A_H\). If \(\Hom_H(\pi_{\xi,\lambda},\pi_{\eta,\nu})\) is the space of continuous \(H\)-intertwining operators, \(\pi_{\xi,\lambda}\) and \(\tau_{\eta,\nu}\) are realized on the spaces of smooth sections of the vector bundles \(\mathcal{V}_\lambda=G\times(\xi\otimes e^{\lambda+\rho}\times\mathbf{1})\to G/P\) and \(\mathcal{W}_\nu=H\times_{P_H}(\xi\otimes e^{\eta+\rho_H}\times\mathbf{1})\to H/P_H\), where \(\rho\) and \(\rho_H\) are the half sums of positive roots, then continuous \(H\)-intertwining operators can be identified with their distribution kernels. The authors show that for every \(P_H\)-equivariant quotient \(p:E_{|P_H}\twoheadrightarrow E''\) with \(E''\simeq\tau''\otimes e^{\mu''}\) as \(P_H\)-representations, there is a unique linear map \(\Phi:\Hom_H(\pi_{\xi,\lambda},\pi_{\eta,\nu})\to\Hom_H(\pi_{\xi,\lambda,\tau',\mu'},\pi_{\eta,\nu,\tau'',\mu''})\), where \(\pi_{\xi,\lambda,\tau,\mu}=\mathrm{Ind}_P^G((\xi\otimes\tau)\otimes e^{\lambda+\mu}\otimes\mathbf{1})\), with the property that for every intertwining operator \(A\) with distribution kernel \(K^A\), the distribution kernel \(K^{\Phi(A)}\) of \(\Phi(A)\) is given by \(K^{\Phi(A)}=\varphi\otimes K^A\) as the multiplication of \(K^A\) with the smooth section \(\varphi\). This result is applied in two different situations. First, the authors construct and classify intertwining operators between principal series representations of \((G,H)=(\mathsf{O}(n+1,1),\mathsf{O}(n,1))\) induced from one-dimensional representations. Next, they construct intertwining operators between principal series representations of \((\widetilde{G},\widetilde{H})=(\mathsf{Pin}(n+1,1),\mathsf{Pin}(n,1))\) induced from spin-representations.