Pernilla Hallonsten Halling

Reference Modification Predication within
reference predication

Objects
unmarked nouns

DIMENSION

Properties


Actions

AGE COLOR

PHYSICAL PROPERTY HUMAN PROPENSITY VALUE
NOISE CARE SPEED TIME ASPECT

unmarked adjectives

unmarked adverbs

unmarked verbs

Simple adverbs, as attested in the languages in table 9.2, instantiate
unmarked
ad-

verbs. Such adverbs consist of property words that, when used as modifiers within predicating expressions, do not have any structural coding. If modification is an acces- sory function not only to reference, but also to predication, adverbs fill a region that would otherwise be empty, or filled by other constructions. However, it would be strange for a certain combination of semantic class and pragmatic function not to have any unmarked instance, when all other combinations do. For property modification, it would mean that the unmarked region would be located only in the upper left corner in which unmarked

adjectives are found in table 9.3. If this argument is pushed further, one interpretation is that this would go against the universal of unmarked combinations of semantic class and discourse function. At the very least, the skewed unmarkedness of property words only in modification within referring expressions would have to be explained. As unmarked property-denoting adverbs are consistently attested, they appear to have a natural place in the lower right corner of the domain of modification. The fact that unmarked adjectives appear to be much more common than unmarked adverbs is not a problem, but rather an interesting tendency that points to the characteristics of these two types of modifiers, which will be discussed in detail in section 9.4.1.
Table 9.4 presents an alternative version of table 9.3, based on languages with general modifiers (as attested in a third of the sample languages, see chapter 6).

Table 9.4. Encoding of semantic classes and discourse functions; expanded version of Croft (2001: 88)

Reference Modification Predication within
reference predication

Objects
unmarked nouns

DIMENSION

AGE COLOR

PHYSICAL PROPERTY

unmarked

Properties
Actions


HUMAN PROPENSITY VALUE
NOISE CARE SPEED TIME ASPECT

general modifiers

unmarked verbs

In table 9.4, unmarked general modifiers cover the regions of unmarked adjec-

tives and unmarked adverbs combined. In order to represent languages that have
both general modifiers and adjectives and/or adverbs, the table would have to be adjusted to make room for both. For a full account of parts of speech, a further possibility would be to include intensifiers, as modifiers of other modifiers, or within modifying expressions. In tables 9.3 and 9.4, the specification of semantic types among property words in the left-most column illustrates the continuum from prototypically adjectival properties

(dimension, age, etc) to prototypically adverbial properties (speed). At the adverbial extreme of the continuum, property words tend to shift towards meanings that describe events, resulting in types that cannot be defined as properties, such as time and aspect. Here, we re-encounter the issue of adverb as a lexical versus functional category. While it is highly important to distinguish between these two types of categories from some perspectives (such as within the generative tradition, e.g. Chomsky 1981), certain uses of adverbs show that lexical and functional meaning cannot always be kept apart. Alterna- tively, certain adverbs can be argued to constitute a domain in which a shift from lexical to functional meaning is quite likely to take place (the reasons for this will be discussed below in section 9.4.1).
Based on the results of this dissertation, adverbs evidently have a place as a part of speech in the typological approach that defines parts of speech as prototypical categories. Instead of questioning the actual part of speech status of adverbs, other questions are warranted. Why are simple adverbs not that common cross-linguistically? And more specifically, when simple adverbs are found in a language, why do the members of this category tend to be so few? In sum, why are adverbs so remarkably different from other parts of speech? We turn to these questions in the next section.

1.