Logic and Lexicon pp 72-111 | Cite as
Vagueness and Ambiguity
Chapter
Abstract
In Chapter 1, I introduced vagueness and ambiguity together, and so far I have not attempted to distinguish them. There is no doubt that vagueness and ambiguity are related phenomena; the fact that both unproblematically allow precisifications is their most important common feature, which distinguishes them from phenomena like presupposition failures. But there can also be no doubt that “vagueness” and “ambiguity” refer to different things; and it seems that there are nuances and distinctions beyond this fundamental dichotomy within the realm of indefiniteness.1 In this Chapter, I will go into the distinction between vagueness and ambiguity and their further classification, especially of ambiguity phenomena. This will result in a systematic clarification of some concepts that are often discussed in lexicology: “ambiguity”, “polysemy”, and “homonymy”. At the same time, this Chapter ought to provide an impression of the wealth of word meaning phenomena that a thorough analysis of indefiniteness and precisification reveals.
Keywords
Borderline Case Definite Description Common Noun Lexical Ambiguity Lexical Semantic
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
Preview
Unable to display preview. Download preview PDF.
References
- 1.In the following, the term “indefiniteness”, if used without modification, refers to vagueness and ambiguity — the two kinds of indefiniteness that obviously allow symmetric precisifications to “true” and “false”. [As discussed in the previous Chapters, there are phenomena of a different kind (presuppositions, sortal restrictions, metaphor) that could be justifiably described as indefiniteness phenomena, but are not subject to the precisification principle (cf. also the discussion of presuppositions in 5.4). The terminology for this second kind of indefiniteness will be explicitly distinguished from indefiniteness in the “narrow sense” (vagueness and ambiguity). In the original, the term präzisierungsfähige Unbestimmtheit, which may be rendered as “indefiniteness with the capacity for precisification”, was used as the superordinate notion to vagueness and ambiguity, in order to distinguish them from phenomena of the second kind (cf. the note in 3.2, p. 59).]Google Scholar
- 2.The quotation originates from Erdmann (1910) [Ger.: ein Ausdruck, “der gradweise Abstufungen zuläßt und durch unmerkliche Übergänge in sein Gegenteil überführt werden kann”]. Erdmann himself does not use the term “vagueness” but rather “ambiguity” [Ger.: Mehrdeutigkeit] throughout his work.Google Scholar
- 3.The special assumptions that Lakoff makes may be disputed (cf. Catlin/ Catlin 1972); the four possible “readings” that he distinguishes are not actually readings, but rather specifications of alternative states of the world; and he announces his test as a method of distinguishing ambiguity from vagueness, but in fact uses it to distinguish ambiguity from generality. Nevertheless, the test offers a relatively certain means of identifying ambiguity.Google Scholar
- 4.The problem is discussed extensively in Zwicky/Sadock (1975) and Kempson (1977). This problem is crucial to theories of presuppositions due to the problem of “double negation”.Google Scholar
- 5.Assumption (5) is similar to a suggestion made by Kindt (1983). There, the disjointness of the denotations of different precisifications is proposed as a criterion of ambiguity. 6 The phenomenon is analyzed extensively in Bierwisch (1979). Bierwisch discusses additional examples, some of which are more complicated, such as the German predicate Schrift, which is ambiguous between at least four readings (in English: handwriting, type font, document, publication).Google Scholar
- 7.For psycholinguistic evidence cf. Rosch (1976), (1979). * [MP: A perspective for an appropriate extension of the framework of lexical semantics is suggested by the work of Manfred Bierwisch on “Two-LevelSemantics”, which stresses the existence of an autonomous level of semantic representation. Hans Kamp, in his work on Discourse Representation Theory, also argues for a representational level which cannot be eliminated in favor of a direct model-theoretic interpretation, since it makes its own contribution to the semantics of sentences and texts (Kamp 1981).] 8 For a comprehensive survey of traditional attempts at definitions of homonymy and polysemy, the reader is referred to Ullman (1957); more recent suggestions for a precise treatment of these concepts are discussed in Lyons (1977).Google Scholar
- 9.And there are further problems: in cases where the original relationships are not entirely clear, should we leave the decision open? How do we deal with ambiguities that arise from loan translations?Google Scholar
- 10.Modal expressions (necessarily, possibly, and above all the modal verbs) are also to be included in this group. For a discussion of their “domainspecific” ambiguity, see Kratzer (1979), (1981).Google Scholar
- 11.In (Partee 1984), Barbara Partee discusses this phenomenon in a different context. But she arrives at similar conclusions, which would be stated in their application on the (j 1)/(j2) example as follows: if the suppressed argument in (j2) is bound by an existential quantifier, then the quantifier must have wide scope (in (j1), it must have narrow scope).Google Scholar
- 12.This distinction is related to the conventional syntactic distinction between “optional” and “obligatory” complements. From a semantic point of view, the direct object of report is obligatory (if necessary, it must be taken from context), but not the indirect object. But there are verbs with an “indexical” character (like come and bring) whose semantically obligatory complements are usually taken from context and not realized (see Fillmore 1966). Thus caution is advised in drawing parallels between syntax and semantics.Google Scholar
- 13.In the classification of adjectives in the last chapter, I left out this less prototypical but important class. The terms “relational” and “relative” must not be confused: “relational” refers to the number of arguments, while “relative” refers to precisification behavior, as shown in 3.1.Google Scholar
- 14.Naturally, the semantics of noun modifiers is more complicated than I can do justice to in two sentences.Google Scholar
- 15.Plural definite descriptions seem to belong to the P-type group: as long as a sufficiently large common core is present, the collection of objects referred to need not be completely precisified.Google Scholar
- 16.Game, Wittgenstein’s example of family resemblance (Wittgenstein 1953), is generally used safely, although there is not one binding defining criterion for its use (see p. 107 below). — The problem of sort-specific multiplicity of use is discussed in Wunderlich (1983) using the German prefix verbs as an example for which this problem is especially prominent. Wunderlich takes a position similar to Quine’s.Google Scholar
- 17.To a certain extent, Quine concedes this himself: “There is an air of syllepsis about ‘The chair and the questions were hard’.” (Quine 1960, p. 130).Google Scholar
- 18.For example, Rolf (1980) argues that vagueness of individuation is due to the “fuzziness” of reality and thus should be eliminated as a semantic phenomenon.Google Scholar
- 19.I cannot quite accept Putnam’s assumption that stereotypes are irrelevant to truth conditions, even in the case of natural kind concepts. They do play a role in truth-conditional semantics, albeit an ambivalent one.Google Scholar
- 20.In the ideal case — in fact we must take evolutionary processes, mutations, cross-breeding, and the like into account; furthermore, indefiniteness may result from certain secondary dimensions, such as the beginning of individual existence in the case of kind concepts, or the degree of purity in the case of mass nouns (cf. Quine 1960, p. 126). Also, the potential influence of prescientific stereotypes on truth behavior — as mentioned above — must be considered. However, these restraints are unimportant in the context of this argument.Google Scholar
- 21.This is why the scientific definitions for natural kind concepts that were just discussed should not be seen as regulations but as “real definitions”.Google Scholar
- 22.[Footnote deleted.]Google Scholar
- 23.In fact, Erdmann hardly manages to obey his own self-restriction: his little book (which is a collection of six essays) is a rich source of interesting vagueness and ambiguity phenomena in all lexical and conceptual domains.Google Scholar
Copyright information
© Springer Science+Business Media Dordrecht 1995