2. Weber does not consider adjectives as a separate lexical class in Quechua. However, it appears justifiable, as I argue in section 3.3.3, to account for adjectives as a lexical class in its own right.
3. Notice, however, that when used predicatively, nouns and adjectives do not behave exactly as prototypical verbal predicates in that they are not assigned morphosyntactic specifications such as person marking which is, as mentioned in section 3.1.1, obligatory in verbs.
4. According to Weber (1989), the Actor argument of a passive clause may be demoted to an oblique role, in which case it is commonly followed by the ablative case marker –pita. However, it is more common to keep this argument unexpressed in spontaneous speech. What is more, Weber acknowledges that when asked, speakers do not always agree on the type of case marker to be used to mark the oblique role of the Actor argument, which may suggest that this language does not really allow for the Actor argument of a passive clause to be indicated. Further investigation is needed to find out if the passive-like phenomena Weber explains can be taken as evidence to account for the relevance of a Subject function in Quechua, as meant in FDG, when referring, for instance, to passivization in English as evidence of the relevance of the syntactic function Subject in this language.
5. A dative shift construction is one in which the Recipient lacks distinguishing morphosyntactic marking, such as prepositional marking, and is treated instead as an Undergoer, morphosyntactically speaking.
6. In English, for instance, the assignment of Object function to the Recipient argument, that is the neutralization between the grammatical properties of Undergoer and the Recipient, occurs basically in three-place predication frames. There seems to be, however, another condition for this neutralization to take place that is intimately related to the notion of animacy. In cases where the Recipient is inanimate and the Undergoer animate, such neutralized behaviour appears to be very unlikely. Compare, for example: I sent the children to the museum and I sent the museum the children(?). In Quechua, neutralization of Undergoer/Recipient argument also occurs in three-place predication frames but a question that remains to be answered is if this neutralization is also sensitive to ‘animacy’. The data I have analysed so far does not provide with evidence to corroborate this.
7. Another restriction, pointed out for other varieties of Quechua such as Imbabura (Cole, 1985) and Ancashino (Cerron-Palomino, 1987), that may be important to consider in order to distinguish adjectives from nouns, and their semantic-syntactic repercussions, is that generally a single noun can be used as a modifier within a Referential Subact, whereas multiple adjectives can behave as modifiers of a nominal head. Cases in which two nouns modify a head noun should be better understood as a compound modifier, as suggested by Cerron-Palomino (1987:300). Beck (2002)) suggests that these compound nominal modifiers appear to be uncommon or otherwise highly lexicalised in Quechua.
8. In FDG, we shall refer to languages as nucleus-marking rather than head-marking, as a distinction is made between nucleus-dependent relations and head-modifiers relations. In this section, we are dealing basically with the former type of relations.
9. It is also possible to mark only the argument bearing the function of Subject on the nucleus, whereas the one bearing the function of Object gets overtly realized on an independent form, instead of being cross-referenced on the verb, carrying dependency marking,
10. According to Weber (1989:421), the evidential suffix –mi (drct) can be interpreted as both an evidential, to indicate firsthand information, or validational marker, to indicate commitment to the truth of the proposition. At the Representational Level, this evidential fulfils the role of an operator at the layer of the Propositional Content and is represented as ‘drct’, following the convention suggested by Weber.
11. For more details on the contexts where this phenomenon occurs, see Weber (1989:463)