Martin (1992: 112):
The final set of distinctions relevant to presuming reference covers the of [sic] these have the function of combining relevance with reminding phoricity in the context of [undirected] groups. The classes of item in question are:
phoric Deictics: both, either, neither, each, whichordinative Numeratives: first, second, third etc.; next, last etc.superlative Epithets: biggest, most enormous etc.
Each of these items has the function of referring to a group of participants relevant to the participant being identified by virtue of including it as a member. So, within undirected reference there is a contrast between presuming the identity of the participant being realised and optionally presuming the identity of group [sic] of participants to which it belongs.
 Trivially, the typographical errors here suggest that the editor of the draft had, by this stage, given up trying to understand the text.
 This is a bare assertion, unsupported by argument or evidence.
 This confuses the experiential construal of participants, realised by nominal groups, with the textual system of reference (the purported concern of the chapter).
 This confuses reference in the transcendent sense of language referring to categories (meanings) outside language with reference in the sense of referring to elements in the co-text. Moreover, this again is a bare assertion, unsupported by argument or evidence.
 This misconstrues distinctions between potential referents as distinctions in the system that is the means of referring. Moreover, this again is a bare assertion, unsupported by argument or evidence.