We love clitics.

General remarks

The status of clitics (syntactic word vs. bound morpheme) is problematic. Most Romance clitics have the properties of bound affixes but for primarily orthographic considerations, the Delph-In Romance grammars realize clitics in syntax (exceptions: post-verbal clitics in Spanish and French).

Issues to be addressed:





Miller and Sag (1997) established that French pronominal clitics are bound affixes (pre-verbal subject clitics are possibly less bound). This approach is not uniformly adopted in the LKB grammar, in order to reflect conventional French orthography. Pre-verbal clitics are analyzed as lexemes and inserted in syntax by "compounding" rules (which must apply lower than ordinary syntactic rules). This is made possible by the following hierarchy under sign:

gram-cat := sign.
lexical := sign.
  word := lexical & gram-cat.
  lexeme := lexical.
phrasal := gram-cat.
  phrase := phrasal.
compound := phrasal & lexical.

Post-verbal clitics are realized morphologically using inflectional lexical rules.

Although vowel elision is correctly analyzed, the apostrophe is currently thrown out (not generated). The hyphen is treated as a normal character. Neither one of these approaches is adequate.

Clitic climbing is analyzed using Abeillé, Godard, and Sag's argument raising analysis: temporal and causative auxiliaries systematically inherit all complements of the participle/infinitive (which is not allowed to project a VP). Currently untreated: downstairs cliticization with causative faire and "clitic trapping".

Currently no analysis for any cases of clitic doubling.

Modern Greek

RomClitics (last edited 2011-10-08 21:12:09 by localhost)

(The DELPH-IN infrastructure is hosted at the University of Oslo)