Actions action nominals, participles, converbs complements, relative clauses
infinitives, gerunds
unmarked verbs
Attributive adjectives, predicative adjectives, and adverbs are here defined according to their functions, as illustrated by the column headings in table 4.1. Attributive adjectives belong to the semantic class of property words that modify nouns within referring ex- pressions. Predicative adjectives are property words that are predicated. Adverbs are property words that modify verbs within predicating expressions. The combination of property concepts and each of these three functions will henceforth be referred to as
attr, pred, and adv, respectively. In the typological study, the three functions will
be used as comparative concepts in the sense of Haspelmath (2010). Based on the three functions, labels such as adjective and adverb, in all lower case, are also used, referring to the comparative concepts defined in (4.1). These definitions are partly based on the functions from the expansion of Croft (2001: 88) and partly on Haspelmath (2010: 670).
1 Table 4.1 might suggest that unmarked adjectives and predicate adjectives are not related, which they clearly are. For a fully expanded version that clarifies this, where also the horizontal row of property words has been added to, see tables 9.3 and 9.4 in chapter 9.
(4.1) (a) an attributive adjective is a lexeme that denotes a property and that can be used to modify a noun within a referring expression
a predicative adjective is a lexeme that denotes a property and that can be used in predication
an adverb is a lexeme that denotes a property and that can be used to modify a verb in a predicating expression
As mentioned in the introduction in chapter 1, when language-specific lexical category terms are used, these are capitalized (e.g. ‘English has a class of Adjectives’) as ac- cording to convention following, e.g., Comrie (1976: 10). I use the term simple adverbs (and sometimes adjectives) to refer to monomorphemic adverbs (or adjectives). When I examine attr, pred, and adv, the purpose is to compare their encoding in different languages. Encoding refers to the structural form attested in a function, for instance, whether property notions in adv are found as lexemes derived from, e.g., adjectives, as simple lexemes, or as complex constructions, involving anything from a serialized verb to a prepositional phrase. The aim of this study is to examine the variation in encoding of attr, pred, and adv. For instance, a specific language may have the same encoding for two or all of the functions, or the encoding may be different for each of the functions. One single function within one single language may also have several different encodings. The primary goal is nonetheless to compare the three functions across languages, in order to see if any typological patterns of encoding can be found, and to what extent such patterns vary. If a language has exactly the same encoding in more than one function, I will call this an encoding overlap. In language examples, I highlight property items in boldface.
Let us first consider what combinations of encoding overlap there are for attr, pred, and adv in English. In (4.2), a typical property in attr, pred, and adv is illustrated. Note that the function (i.e. attr, pred, or adv) that an individual example instantiates is indicated in the right margin of the specific example. This is a procedure that will be followed throughout the remainder of the thesis.
(4.2) English [attr pred] [adv] lexeme
The sad person over there is Sheila. attr
Sheila is sad. pred
Sheila smiled sadly. adv
These examples instantiate an overlap of attr and pred on the level of the lexeme, since these two functions are identically encoded (sad ). adv, on the other hand, requires the ending -ly (sad-ly ). The overlap is indicated with square brackets in the example heading, but can also be illustrated as in figure 4.1.
Figure 4.1. The English lexeme overlap of attr and pred
It may be objected that it is far from clear whether the -ly ending is derivational (cf., e.g., Haspelmath 2002: 60). Happily could be regarded as the same lexeme as sad. Such an analysis could also be used for the purpose of the present study. However, I will assume that examples such as sad and sadly instantiate different lexemes, and base my analysis on this assumption. It is nonetheless important to note that an inflectional analysis of -ly, which implies that sad and sadly are different forms of the same lexeme, could also be applied, rendering somewhat different results. The treatment of sad and sadly as different lexemes also highlights the need for another more narrow level of analysis to capture the fact that English has the same root in all examples in (4.2).
(4.3) English [attr pred adv] root
The sad person over there is Sheila. attr
Sheila is sad. pred
Sheila smiled sad-ly. adv
A root is here defined as a form that cannot be divided into any morphologically simpler
form. On the root level, the overlap in (4.3) includes attr, illustrated in figure 4.2.
pred, and adv. This is
Figure 4.2. The English root overlap of attr, pred, and adv
The encoding in examples (4.2) and (4.3) is the most common pattern for attr, pred, and adv in English. But there are also other instances, such as the examples with fast in (4.4). Here, it is not possible to attach the -ly ending, and all three functions of the property word are encoded in the same way on the level of the lexeme as well as that of the root.
(4.4) English [attr pred adv] lexeme/root minor
The fast runner over there is Sheila. attr
Sheila is fast. pred
Sheila runs fast. adv
The examples with fast show an overlap of all three functions on both the lexeme and root levels, as illustrated in figure 4.3.
Figure 4.3. The English lexeme/root overlap of attr, pred, and adv
Yet other minor examples do not even occur in all three functions, as in the case of alone
in (4.5), found only in pred and adv (cf. *The alone person over there is Sheila).
(4.5) English [pred adv] lexeme/root minor
Sheila is alone. pred
Sheila runs alone. adv
The example with alone shows a case in which the lexeme in question cannot be encoded in attr. Not being able to be encoded in a function is clearly different from being encoded in a way separate from the other two functions. This overlap is illustrated in figure 4.4.
attr
Figure 4.4. The English lexeme/root overlap of pred and adv
As noted in the example, alone constitutes a minor pattern and since it is not a typical property word, it is of limited interest here. The example of fast in (4.4) also instantiates a minor pattern, since the great majority of English Adverbs in adv are formed with the
Do'stlaringiz bilan baham: |