Pernilla Hallonsten Halling

Adverbs as modifiers of predicates

Manner is a semantic class label among many others used to classify words on various levels. More general classes such as objects, actions, and properties have traditionally been used as a semantic basis for the part of speech categories noun, verb, and adjective. Such a semantic division can easily be proven inadequate for part of speech classifications, since the classes are not restricted to one part of speech each (e.g. action words can be nouns, as in the case of the English destruction). According to Croft (1991, 2001, 2003), although the three semantic classes do not suffice for discerning part of speech categories, they are still needed in order to do so. Another dimension must nonetheless be added, namely the functions that the semantic classes are used in. Croft argues that every language has expressions for three major pragmatic or propositional act functions, a term originally introduced by Searle (1969: 23–24) (also called communicative or discourse functions). These three functions are: reference, predication, and modification.
The act of reference identifies a referent and establishes a cognitive file for that referent, thereby allowing for future referring expressions coreferential with the first referring expression. The act of predication ascribes something to the referent...The act of modification (of referents) functions to enrich a referent’s identity by an additional feature of the referent, denoted by the modifier. (Croft 2001: 66)
The three major semantic classes can be used in any of the three major propositional act functions. In certain combinations, prototypical parts of speech can be identified. Accordingly, the prototypical noun is found in reference to an object, that of verb in predication of an action, and that of adjective in modification by a property (2001: 89). According to Croft, this is supported by universal-typological markedness patterns. The term markedness here refers to how much structural coding, or how many morphemes, are used in the encoding of an expression: “the marked member [of a category] is encoded by at least as many morphemes as the unmarked member” (2001: 90). Thus, an unmarked item cannot be encoded by fewer morphemes than a marked member, but a marked item is always encoded by the same amount of morphemes or more than the unmarked one. When objects words are used for reference, they are unmarked. When action words are

used as predicates, they are unmarked. And when property words are used as modifiers, they are unmarked. Any other combination results in markedness, as illustrated for English in table 2.2.

Table 2.2. Encoding of semantic classes and discourse functions (Croft 2001: 88)

Reference Modification Predication

Objects

unmarked nouns
genitive, adjectivilizations, predicate PPs on nouns nominals,
copulas