The Festival text-to-speech system is quite complex. One of the phenomenal contributions behind it, is the book Building Synthetic Voices by Black and Lenzo. This book explains the internals of Festival in a straightforward manner. With this resource, it should be possible to build small TTS systems for new languages.

While TTS these days is dominated by incredible deep learning results, building a TTS for an endangered language is much easier using Festival and its grammar-based tooling.

However, some of the people that can undertake that effort might not be able to speak English, but they might be bilingual in Spanish, French, Mandarin Chinese or an Arabic dialect. Translating Building Synthetic Voices to other languages can thus help make technology more inclusive.

As an added bonus, it might help encourage the authors to finish the book!