Halliday & Matthiessen (2014: 247):
It might be argued … that since grammatical and semantic categories are not in one-to-one correspondence, then if we use grammatical terms that are semantic in import (as nearly all grammatical terms are) we cannot expect them to be appropriate for all instances. The reasoning is quite valid; grammatical labels are very rarely appropriate for all instances of a category — they are chosen to reflect its central or ‘core’ signification ( … ‘prototype’ …). These core areas are the central region for each process type … and the non-core areas lie on the borders between the different process types, shading into one another as the colours of a colour spectrum.