Satellite-framed patterns in Romance languages: A corpus-based study

The present analysis is grounded in the belief that linguists, when describing a language, should aim for a full and comprehensive coverage. Talmy’s (1985, 2000) influential two-way typology, verb-framed vs. satellite-framed patterns, represents the preferred option here for the encoding of motion events cross-linguistically, but does not cover other peripheral uses that a language may show. This paper provides evidence for the growing assumption that languages may in fact show both encoding options (Beavers, 2008; Beavers et al., 2010; Filipovic, 2007; Iacobini and Masini, 2006, 2007; Fortis, 2010, Croft et al., 2010, inter alia). The analysis of a large corpus sample of satellite-framed constructions shows that in Spanish this pattern is not only available but indeed is preferred under some circumstances. Previous assertions that Romance languages have poor lexical manner inventories and lack resultatives can help explain low productivity, but they do not argue against the existence of a satellite-framed encoding choice per se. By analysing naturally occurring constructions in their contexts, I will outline the pragmatic conditions that compensate for lexical and aspectual limitations. When the resultative element (change of location) is a default inference, it can be lexicalized.