Martin (1992: 537):
The notion of activity sequence has obvious affinities with various concepts developed in artificial intelligence and cognitive psychology (e.g. the frames, scripts, scenarios and schemata reviewed in Brown & Yule 1983: 236-70). The most relevant articulation here however is that of Barthes (1966/1977: 101-4), of whose earlier work Brown and Yule make no mention.
 The frames, scripts, scenarios and schemata reviewed in Brown & Yule are ways of 'representing background knowledge', 'knowledge of the world' (1983: 236-7), with an emphasis on the way knowledge of the world is stored in memory (ibid.). That is, what, for SFL Theory is meaning potential, established ontogenetically, is construed, in such models, as extra-linguistic cognition.
As Martin conceives them, however, activity sequences are an entirely different matter. For Martin, activity sequences are an aspect of field, the ideational dimension of context. However, Martin's model is further complicated by the facts that
- context (culture) is misconstrued as register (subpotentials of language), and
- activity sequences are meanings (semantics) misconstrued as field (context).
 It is understandable that Brown & Yule (1983) did not review Barthes (1966/77) in this regard, since the former are concerned with models of extra-linguistic cognition, whereas the latter is concerned with a unit of language (the sequence).