David Rose wrote at 01:14 on 31/10/14 to the sys-func and sysfling email lists:
1. Isn't part of the notional problem a confusion of word rank lexical items with clause rank grammatical functions? E.g. a quoting verbal/mental process may be instantiated by a verb denoting behaviour, but does that make the clause a behavioural process?
Is this a classification of lexical verbs or of clause rank process types? What does this mean for the "lexis = delicate grammar" hypothesis?
2. What is the epistemological value of the behavioural process category? Some posts such as Yaegan's point to the value for students of distinguishing 'blurred' categories in text analysis, although Mick points to the cost in pedagogic labour. Other posts suggest its value for negotiating authority in the field, by defining the criteria for the category (or its absence). Is part of its value a relatively safe theoretical cul de sac for a good SFL argument?
 In SFL terms, Rose's nominal group 'word rank lexical items' confuses two distinct theoretical dimensions: the rank scale and delicacy. This is because it conflates the grammatical (word rank) with the lexical (lexical item). Halliday & Matthiessen (1999: 568):
The folk notion of the “word” is really a conflation of two different abstractions, one lexical [lexical item] and one grammatical [word rank]. The relation between a clause rank Process and a word rank verb is realisation (via group rank), not instantiation. Higher rank functions are realised by lower rank forms. Instantiation is the relation between the system as potential and a specimen of the system in an actual text.
 Deploying Functional Grammar (Martin, Matthiessen & Painter 2010) advises treating such quoting clauses as behavioural, despite the grammatical reactances; see discussion here.
 Epistemology is branch of philosophy that studies the nature of knowledge, its presuppositions and foundations, and its extent and validity.