Glossary of terms used

A UTF-8 encoded text file that has been imported into CLiC, invariably from the corpora repository.

A labelled portion of a book. Each region will have:

  • A name (or rclass), for example ‘chapter.sentence’. For all possible names, see /schema/10-rclass.sql.
  • A start and end character position within the full book text
  • (optionally) a number (or rvalue), for example it’s position within a chapter.

Regions are added by region tagger scripts, which are in clic.region.


A token is a labelled portion of a book that contains a single word. In the phrase 'For _more_!' said Mr. Limbkins., For, more, said, Mr, Limbkins would be tokens.

See clic.tokenizer.


A token has a type (thus ttype). This is a normalised form of the token. The tokens in the phrase 'For _more_!' said Mr. Limbkins. would have types for, more, said, mr, limbkins.

See clic.tokenizer.