NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED347788
Record Type: RIE
Publication Date: 1989
Pages: 11
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
Has Abstractness Been Resolved?
Al-Omoush, Ahmad
Dirasat, v16 n11 p68-77 1989
A discussion focusing on the abstractness of analysis in phonology, debated since the 1960s, describes the issue, reviews the literature on the subject, cites specific natural language examples, and examines the extent to which the issue has been resolved. An underlying representation is said to be abstract if it is different from the derived one, but as the underlying representation deviates from the corresponding phonetic representation, rules required to cover the deviating phonetic forms increase. Postulating abstract representations is pointless unless they are well-motivated and have a simplifying effect on grammar. Phonological theory must impose some constraints on underlying representations so unmotivated analyses can be excluded. However, some proposed constraints prohibit certain analyses that are well-motivated on internal grounds. It is argued that the problem of abstractness is unresolved without sufficient external evidence that touches on the validity of internal evidence. Examples are presented that provide evidence for abstract analysis, including: a study of a Polish vocalic phonology; an analysis of vowel harmony in Manchu, a Uralic-Altaic language; and vowel harmony in Yawelmani. The last is a case in which the need for resorting to abstract analysis arises because significant linguistic generalizations can be achieved only by such means. (MSE)
Publication Type: Journal Articles
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers: Natural Language