The Avesta, the Zoroastrian scripture, is preserved in a language that had no other literature and no living speakers by the time it was written down. It was carried orally by priests for a very long period, committed to writing comparatively late, and the manuscripts that survive descend from a small number of copies made after the tradition had already suffered heavy losses. Large portions of what the tradition itself remembered as scripture are gone, known only from summaries of their contents.
The community that carried it shrank drastically after the Arab conquest, and a significant part of it migrated to western India, where the Parsis preserved manuscripts and liturgy in circumstances quite different from those of the Iranian remnant. Modern scholarship depends on both branches, and where they differ the difference is itself evidence.
Outside the tradition, Greek writers reported on the magi from the fifth century BC onward, and their reports gave Europe both the word magic and a persistent image of Persian wisdom that owes as much to Greek expectations as to Iranian practice.
The twentieth century added an entirely unexpected source. Manichaean texts in several languages were recovered from the Turfan oasis and from Egypt, giving the first direct access to a religion previously known almost entirely through the polemics of its opponents, Augustine among them. It is the clearest case in this atlas of a tradition's own voice being restored by excavation.