I'm pretty sure most distros default to espeak or espeak-ng for handling text-to-speech. Also, regardless of the TTS used on the back end, Orca uses Speech Dispatcher as a go between between itself and the TTS.