Martin (1992: 543):
Domestic sequences tend to be implicit — they are not usually written down, or taken notice of in any way by mature speakers, and are learned by doing, under the guidance of caregivers by children.
Here 'sequences' refers to (material order) behaviour, not (semiotic order) language, nor the context realised by language. The model being developed here is purported to be field, the ideational dimension of the culture as semiotic, realised in language, so the confusion is one of orders of experience.
The inconsistency is compounded by Martin's misconstrual of this behaviour as register, a subpotential of language.