Martin (1992: 127, 128):
The nominal groups in [3:1] are listed below. Each is coded for the IDENTIFICATION choices made and the type of reference to the context where these choices are phoric. The analysis will be annotated for purposes of discussion, rather than presented in detail.
NOMINAL GROUP REFERENCE(terminal features) RETRIEVAL(where phoric) one day presenting…particular — I presuming…interlocutor anaphoric*
i. Anaphoric to the writer's name, which was written at the top of her story; pronominal references to the writer could in addition be coded as exophoric throughout [3:1] (the writer's name itself was homophoric, identifying the writer to her teacher and the class).
 This mistakes nominal groups for reference items; see previous posts.
 This misunderstands textual reference; all reference is 'phoric', by definition.
 This mistakes interpersonal deixis for textual reference. The non-specific Deictic 'one' does not mark 'day' as identifiable. The confusion is one of metafunction.
 To be clear, Martin's 'presenting' reference is neither reference nor textual; instead, it labels the first instantiation of an experiential participant in a text. The confusion is one of metafunction.
 This mistakes interpersonal deixis for textual reference. To be clear, only non-interactant (3rd person) pronouns and determiners function as personal reference items, since these alone mark identifiability. The identities of the interactants (1st & 2nd person) are given by their rôles in the speech event (Halliday & Matthiessen 2014: 606). The confusion is one of metafunction.