Configuring speech output with a specified rate of speaking