Martin (1992: 507):
(v) Finally, and critically, there is the question of genre agnation. The argument here is that social processes are related in ways which complement the valeur determined by looking at them from the perspective of field, mode and tenor alone. Combinations of field, mode and tenor choices in other words enter into relationships with each other which are more than the sum of their parts; to some extent, genres have a life of their own (see 7.3.2 below). This argument can only be pursued by formulating networks for field, mode, tenor and genre and demonstrating that the genre network consists of different kinds of opposition to the field, mode and tenor ones. Consequently, exemplification will be reserved until after the presentation of preliminary field, mode and tenor options below.
 The claim here is that text types (genres) are insufficiently identified by the contextual features of field, tenor and mode — which Martin misconstrues as register — and require further specification through systems of a stratum that is more abstract than the culture as semiotic system, but which specifies text structure, which in SFL theory, is located two levels below, at the level of semantics, within language, rather than context.
Further critical analysis "will be reserved" until the networks are presented that claim to demonstrate "that the genre network consists of different kinds of opposition to the field, mode and tenor ones".
 To be consistent with Martin's argument, this claim should have been that genres are more than the sum of combinations of field, tenor and mode.