The index lists every genealogical unit Glottolog 5.3 records. Open one and its tree is drawn with the root at the left and descent running rightward. Depth is classification depth, not time: a language far to the right is deeply nested, not necessarily recent.

• The control bar sets how deep to draw. Showing every tip can produce a large tree; the shallower settings are usually the useful ones.
• Pick a single branch to give it the whole canvas.
• Colour is branch. Grey nodes are extinct, and a name in italic is a classificatory grouping rather than a language anyone spoke under that name.
• Filing buckets are not drawn as trees. Bookkeeping, Sign Language, Pidgin and the rest are shown flat. This view does not show their subgrouping.
• Dialects are omitted unless a million people or more speak them.
• Nothing moves. The layout is a deterministic function of the data, so the same settings always draw the same tree and the downloaded image matches the screen.

Classification follows Glottolog 5.3, which records attested languoids and no reconstructed proto-languages. Indo-European additionally keeps a curated tree that does include reconstructed stages.

Elsewhere: The Same Idea · The Missing N · Flat-Foot · Polite Company · Signs for Sounds · The Second Alphabet · Twenty-Two Letters · Brahmi's Daughters · The Same Word · Heartwood