Cliticization in Old Spanish and Wackernagel’s Law

Clitics and functional categories: Comments on Agr-based approaches to cliticization in Old Spanish

“…OS clitics could be separated from the verb with which they were thematically associated, for example, by a (subject or object) NP or an adverb (e.g., mds 'more'): It has been suggested that this phenomenon, standardly referred to as interpolation in historical studies of Spanish (and Romance), is not problematic to the Agr-analysis of clitics since "Agr is not prevented from raising to a higher head (for stylistic reasons) once licensing of a projection in the domain of Comp has been accomplished" (Mallen 1999: 12, though he fails to elaborate on this point). 11 First, note that all of the above examples involve embedded environments (most frequently in subordinate clauses with que; see, e.g., Ramsden 1963: 136).…”

“…Tobler-Mussafia's Law; e.g., Rivero 1997, Fontana 1997, Mallen 1999, among others; cf. Martins 1995 for Portuguese) have proposed minimalist feature-checking analyses of the placement of pronominal eli tics.…”

“…Fontana assumes that clitics could be: (a) adjoined to the left of IP; (b) substituted in the Spec, IP; or (c) adjoined either to the left or to the right of an XP in the Spec, IP position (p. 233) in the syntax. In another study, Mallen ( 1999) relates the raising of Agr to Camp to the clitic-verb order, and the absence of such raising to the verb-clitic order. The present paper, without attempting to investigate the OS clitic placement possibilities in any comprehensive or exhaustive manner, examines the alleged descriptive and/or explanatory adequacy of the Agr-based (i.e., syntactic-checking) approach to OS cliticization, focusing on one such proposal, namely, the Agr-raising analysis proposed by Mallen ( 1999).…”

“…Suggesting that the preverbal (proclitic) vs. postverbal (enclitic) placement of OS pronominal clitics was conditioned by the syntactic feature-checking mechanism, 1 Mallen ( 1999) proposes that the contrast between (1 a), proclisis, and (l b), enclisis, can be attributed to the difference in Agr-movement options: 2 (l) a. El te mostrara una albuhera de que ... (EE §316, l84b7) (= M22a) (he to-you will-show a pond of which) b. Eel respondiol (EE §316 184b22) (= M22b) (and he answered-him)…”

“…'-adjunction (which seems to be ad-hoc), if adopted, would make a grammar unnecessarily powerful, and consequently would weaken its explanatory potentials. The only remaining option would be the generation of multiple specifier positions, which would, on the other hand, be in conflict with the assumption that the OS CP was underdeveloped (Mallen 1999; cf. Rivero's (1997) articulated C-system).…”