Is there support for pre emptive speech generation...