Martin (1992: 144):
Note that redundancy phoricity is not taken as contributing to the structure of reference chains, since it is not concerned with presuming the identity of participants, but simply with presuming some aspect of their experiential meaning. Substitution and ellipsis at group rank is thus more appropriately treated as an aspect of lexical cohesion…
 To be clear, redundancy phoricity is Martin's rebranding of Halliday & Hasan's (1976) cohesive system of substitution and ellipsis, misunderstood as a type of reference, and relocated from lexicogrammar to Martin's discourse semantics stratum.
 To be clear, here Martin proposes a system (paradigmatic axis) that has no structural realisation (syntagmatic axis).
 This is misleading, because it is manifestly untrue, since the substitution and ellipsis is of wording, not experiential meaning (Halliday & Matthiessen 2014: 635-6). Moreover, the elements subject to substitution and ellipsis are identified in terms of the interpersonal metafunction (ibid.). So, in terms of Martin's discourse semantics as 'meaning beyond the clause', this is neither meaning nor beyond the clause.
 On the one hand, as the terms 'group rank' and 'lexical cohesion' make clear, this is lexicogrammar misrepresented as discourse semantics. On the other hand, Martin recommends treating redundancy phoricity as the discourse semantic counterpart of grammatical reference (IDENTIFICATION) when the ellipsis–&–substitution occurs at clause rank, but as the discourse semantic counterpart of lexical cohesion (IDEATION) when the ellipsis–&–substitution occurs at group rank.
- redundancy phoricity is not the semantic counterpart of reference;
- redundancy phoricity has no structural realisation;
- redundancy phoricity is not meaning beyond the clause.
- substitution and ellipsis at clause rank is grammatical reference (his IDENTIFICATION);
- substitution and ellipsis at group rank is lexical cohesion (his IDEATION).