Martin (1992: 378):
Throughout this chapter, much more attention has had to be paid to register variables than in Chapter 2, 3 and 4. The discussion of IDEATION was undertaken with a view to textually mediating the realisation of field in lexicogrammar. During any such consideration of the realisation of one level of meaning in another, it needs to be kept in mind that as a theoretical construct realisation is not directional.
 This misrepresents field — tenor and mode were not discussed — as a register variable instead of contextual variable. No argument has been made for reinterpreting context as register; the claim has merely been repeated over and over.* The many reasons why modelling context as register is inconsistent with the meaning of both concepts will be outlined in the critique of Chapter 7: Context.
 Texts don't "mediate" realisation. Texts are instances of a systemic potential that is stratally organised in terms of realisation.
 'The realisation of field in lexicogrammar' misrepresents stratification. Field is realised in semantics, not lexicogrammar; it is semantics that is realised in lexicogrammar. However, since Martin's 'field' is actually semantic, there is an unintentional truth here.
 This continues the misunderstanding of stratification arising from confusing it with semogenesis, as previously explained. In SFL theory, the strata construe different levels of symbolic abstraction, only one of which, semantics, is the level of meaning. This is distinct from 'meaning-making' or semogenesis.
 Realisation is a relation of identity between two levels of symbolic abstraction — one higher (the Value), one lower (the Token). The lower Token (e.g. lexicogrammatical stratum) realises the higher Value (e.g. semantic stratum); the higher Value is realised by the lower Token.
*As Aldous Huxley (1991 : 38) described 'hypnopædia' in Brave New World:
Sixty-two thousand four hundred repetitions make one truth.