Premium
Ambiguous Labeling and Full Interpretation
Author(s) -
Mizuguchi Manabu
Publication year - 2019
Publication title -
studia linguistica
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.187
H-Index - 28
eISSN - 1467-9582
pISSN - 0039-3193
DOI - 10.1111/stul.12109
Subject(s) - ambiguity , merge (version control) , computer science , interpretation (philosophy) , computation , syntax , set (abstract data type) , natural language processing , artificial intelligence , programming language , information retrieval
Syntactic objects constructed by Merge are label‐less and are identified or labeled by a labeling algorithm called Label for interpretation at the interfaces. It has been argued that the set of the form {{ XP }, { YP }} (or simply, XP ‐ YP ) incurs labeling ambiguity and cannot be labeled unless [i] either XP or YP moves out or [ii] X and Y agree. This paper considers labeling of XP ‐ YP and argues that labeling ambiguity can be tolerated, causing no labeling failure. I claim that XP ‐ YP can be labeled either 〈X〉 or 〈Y〉, which is a natural consequence of minimal computation. The well‐formedness of labeling is attributable to the CI interface (or conditions imposed by the CI system), with syntax not caring about the outcome of labeling. I show that ambiguous labeling is empirically supported and removes unwanted stipulations on labeling.