Human language

인간의 언어

Ambiquity, uncertainty, vagueness

Ambiquity¹

• Phonological ambiguity arises when there is more than one way to compose a set of sounds into words. For example, “ice cream” and “I scream” sound more or less the same.
• Syntactic ambiguity arises when a sentence can have two or more different meanings because of its structure. For example, the sentence “John ate the cookies on the couch” could mean either that there were some cookies on a couch and John ate them, or that John ate some cookies while sitting on a couch.
• Anaphoric ambiguity arises when a phrase or word refers to something previously mentioned, but there is more than one possibility. For example, in the sentence “Margaret invited Susan for a visit, and she gave her a good lunch,” the pronoun she may refer to Margaret or Susan.
• Term-level semantic ambiguity is also known as lexical ambiguity and arises when a term (single-word or compound) can have more than one meaning. For example, similarly to the aforementioned Tripoli, the term Kashmir may refer to the song by the band Led Zeppelin or to the geographical region in India and Pakistan.
• Sentence-level semantic ambiguity arises when even after the syntax and the meanings of the individual words in a sentence have been resolved, the sentence can still be interpreted in more than one way. For example, the sentence “John and Jane are married” can mean either that John and Jane are married to each other or that they are both married but to different people.

Uncertainty¹

The phenomenon in which a statement’s truth cannot be determined due to complete or partial lack of required knowledge, e.g. “It’s probably raining right now in Stockholm.”

Uncertainty in context of semantic modeling:

• Explicit uncertainty is when the statement contains keywords like probably, might, perhaps, apparently, etc., that clearly communicate the lack of absolute certainty.
• Implicit uncertainty is when it is expressed in a certain way, yet we have reasons to still doubt its truth. For example, if we know that some knowledge source contains a substantial percentage of false statements, then we are naturally suspicious for every statement that comes from it.

Vagueness¹

• A predicate has degree or quantitative vagueness if the existence of borderline cases stems from the (apparent) lack of precise boundaries between application and nonapplication of the predicate along some dimension.
• A predicate has combinatory or qualitative vagueness if there is a variety of conditions, all of which have something to do with the application of the predicate, yet it is not possible to make any sharp discrimination between those combinations that are sufficient and/or necessary for application and those that are not.

Footnotes

1. Chapter 3, Semantic modeling for data ↩ ↩² ↩³