Edition Corpus

These are canonical and collected volumes taken as electronic text rather than as scans: the Resources for Kanjur & Tanjur Studies (Vienna) publishes per-volume transcription for 27 editions of the Tibetan canon, and we pull it, split each volume at its text boundaries, and convert the Wylie transliteration to Tibetan script. What arrives is a text that can be searched, compared and translated — the rKTs e-text repository is the source, and it is published under CC BY 4.0.

Most of it is not new work. The Narthang, Stog and Phugbrag volumes largely reprint what the Derge corpus already holds, and we identify those by their opening syllables rather than by title — editions punctuate and spell titles differently, so a title match merges two different praises and splits one text in two. A text recognised as already present is marked Derge witness: it is kept for variant readings and for filling lacunae where Derge lost a folio, and it is not translated a second time. The Nyingma rgyud 'bum collections, by contrast, are largely outside the Tohoku catalogue altogether.

Transcription quality is not uniform, so each volume carries a score q — the share of its syllables attested in the Derge corpus. Curated transcriptions score 0.93–0.98. Volumes below 0.70 are printed as needs proofreading and are held out of the translation queue: a garbled transcription still converts to well-formed-looking Tibetan, so only a lexicon check sees it, and translating a corrupted witness would produce an English text that quietly misrepresents the source.

6,563texts in this category

Texts