Difference between revisions of "One-affix-one-rule hypothesis"
Haspelmath (talk | contribs) |
Wohlgemuth (talk | contribs) m |
| (One intermediate revision by one other user not shown) | |
(No difference)
| |
Latest revision as of 16:13, 8 July 2009
The One-affix-one-rule hypothesis is a hypothesis which says that a word formation rule specifies a unique phonological operation which is performed on the base. It specifies a unique syntactic label and subcategorization frame, as well as a unique semantic reading.
Example
in English the suffix -er is used to form agent nouns (worker, player) as well as comparatives (quicker, happier). Given the differences in syntactic category and meaning, the one-affix-one-rule hypothesis entails that there are two -er affixation rules. This example seems rather trivial, but things become less trivial if one knows that -er is also used to form instrument nouns (hanger, glider) since the circumstances under which agent nouns and instrument nouns are formed are identical. The one-affix-one-rule hypothesis says that we have two rules here, but this claim does not explain the similarities between both rules of -er nominalization.
Links
Utrecht Lexicon of Linguistics
References
- Scalise, S. 1984. Generative Morphology, Foris, Dordrecht.
