Core Concepts

The Viewer and the Editor are two views onto the same body of data. This section explains the ideas they share — the model, rather than the buttons. Read it once; both guides assume it.

The short version

A text is one manuscript, with its photographs. An edition is one scholar's reading of that manuscript; a text can carry several, and nothing ranks them.

Within an edition, the reading is recorded in layers. Akṣaras are outlined on the image and spelled as sequences of graphemes. Tokens — words — point at those akṣaras, and carry a lemma and an inflection. Sequences group tokens into units of text and carry the translation and chāyā. Only the akṣara layer touches the image; every layer above it is defined by pointing at the layer below.

The Editor is where those records are made. The Viewer presents them. The IIIF endpoints publish them.

Three things that surprise people

Worth knowing before you go further, because each one accounts for a whole class of confusion:

• An akṣara is a sequence of graphemes, not a syllable. There is no śu in the catalogue; there is ś followed by u. See Akṣaras and Transliteration.
• Images belong to the text, not the edition. Every edition of a manuscript shares one set of photographs. See The Data Model.
• Lemmas are shared by everyone, not owned by your edition. Which is why approving one takes it away from you. See The Shared Lexicon.

In this section

• The Data Model — how the pieces relate, who owns them, and who can see them.
• Editions — text versus edition, and why a text carries several readings.
• Akṣaras and Transliteration — how the script is modelled, and how it sorts.
• The Shared Lexicon — lemmas and inflections as common property.
• Annotations — how each layer anchors to the one below it.
• IIIF and Deep-Zoom Images — how images are served and editions published.

Unfamiliar term? Check the glossary.