Martin (1992: 123):
Forward reference between groups is cohesive and will be referred to as cataphora, narrowing Halliday & Hasan's (1976: 72) definition in order to oppose the term to esphora. This cohesive cataphora is realised by proximate demonstratives (this/these) and presumes text, not participants:
[3:55] This is what bothers me —you can't trust them.
 As previously explained here, Martin's theoretically unmotivated rebranding of structural cataphora as esphora is poor theorising on several counts, not least because, although structural cataphora is a type of endophora, the term 'esphora' is equivalent to its opposite, 'exophora' (the prefix es- derives from Latin ex-).
 According to the source of Martin's ideas, Halliday & Hasan (1976: 68), contrary to Martin's claim, the only demonstrative that makes cataphoric reference to extended text is the singular form this. On the other hand, the reader is invited to consider whether that can also serve this function, as in:
That is what bothers me —
you can't trust them.
 The opposition 'text vs participant' betrays Martin's misunderstanding of textual reference as ideational denotation of meanings (participants) outside language (texts). The inconsistencies here with SFL theory are thus metafunctional (textual vs ideational) and epistemological (immanent vs transcendent perspectives on meaning).