<i>SE</i>-verbs, <i>SE</i>-forms or <i>SE</i>-constructions? <i>SE</i> and its transitional stages between morphology and syntax
The article treats a problem that has, until now, received only a minor focus in the literature on Romance <i>SE</i> (but see Fagan 1992): the question being wether or not the formation with <i>SE</i> plus verb should be analyzed as a lexical <i>SE</i>-verb generated via a word-formational device, as an inflectional verbal <i>SE</i>-form or as a syntactic <i>SE</i>-construction, or if none of these traditional analyses is preferred. A closer look at the characteristics of <i>SE</i> and the <i>SE</i>+verb constructions shows that <i>SE</i> (<i>+</i>verb) behaves differently and therefore must be categorized differently depending on the function the whole <i>SE</i>+verb construction has. Some <i>SE</i>+verb formations are more derivation-like (antiagentive, reflexive, reciprocal), some more inflectional (medio-passive), and some behave more like syntactic constructions (indefinite). This variable synchronic behaviour can be explained by looking at the history of the Romance <i>SE-</i>construction(s): one can detect a process of grammaticalization (see Mutz 2005; Russi 2008) reaching from a phase when <i>SE</i> was a referential object pronoun (reflexive) to a phase in which it has developed into a mere (inflectional) marker indicating a diathetic process (medio-passive), by passing through other different functional “stages” with their own structural characteristics. The different structural and distributional behaviour of <i>SE</i> depending on its function, thus speaks in favour of the Autonomy of Morphology.