17.1.2 Unambiguous Representations

The domain of semantics, like all the other domains we have studied, is subject to ambiguity. Specifically, single linguistic inputs can legitimately have different meaning representations assigned to them based on the circumstances in which they occur.

Consider the following example from the BERP corpus:

(17.2) I wanna eat someplace that's close to ICSI

Given the allowable argument structures for the verb eat, this sentence can either mean that the speaker wants to eat at some nearby location, or under a Godzilla as speaker interpretation, the speaker may want to devour some nearby location. The answer generated by the system for this request will depend on which interpretation is chosen as the correct one.

Since ambiguities such as this abound in all genres of all languages, some means of determining that certain interpretations are preferable (or alternatively less preferable) than others is needed. The various linguistic phenomena that give rise to such ambiguities, and the techniques that can be employed to deal with them, will be discussed in detail in the next four chapters.

Our concern in this chapter, however, is with the status of our meaning representations with respect to ambiguity, and not with the means by which we might arrive at correct interpretations. Since we reason about, and act upon, the semantic content of linguistic inputs, the final representation of an input's meaning should be free from any ambiguity. Therefore, regardless of any ambiguity in the raw input, it is critical that a meaning representation language support representations that have a single unambiguous interpretation $ ^{1} $.

A concept closely related to ambiguity is vagueness. Like ambiguity, vague-

ness can make it difficult to determine what to do with a particular input based on its meaning representation. Vagueness, however, does not give rise to multiple representations.

Consider the following request as an example:

(17.3) I want to eat Italian food.

While the use of the phrase Italian food may provide enough information for a restaurant advisor to provide reasonable recommendations, it is nevertheless quite vague as to what the user really wants to eat. Therefore, a vague representation of the meaning of this phrase may be appropriate for some purposes, while a more specific representation may be needed for other purposes. It will, therefore, be advantageous for a meaning representation language to support representations that maintain a certain level of vagueness. Note that it is not always easy to distinguish ambiguity from vagueness. Zwicky and Sadock (1975) provide a useful set of tests that can be used as diagnostics.