See / Witness and the ‘setting-subject construction’

Perception verbs prototypically occur with a grammatical subject NP referring to a person. However, see and witness also license an inanimate grammatical subject, more precisely a spatial or temporal setting, in a “setting-subject construction” (Langacker 1991, 2008). The present study addresses this kind of variation, and demonstrates how the two alternate constructions reveal shifts from an egocentric perspective to an anthropocentric perspective. It sets out to accomplish three main goals: first, to establish whether each construction aligns perfectly with one particular perspective; second, to identify the semantic and syntactic characteristics of setting-subject constructions and explain how an inanimate subject NP can be favored over a human subject NP; third, to determine what can motivate speakers’ choices between the two alternate constructions licensed by see and witness. To achieve this, a qualitative, corpus-based analysis is carried out, which helps to understand to what extent the grammatical coding embodies a specific way of viewing the scene. First, the cognitive theoretical concepts (e.g., the Extended Animacy Hierarchy (Croft, 2003), egocentric and canonical viewing arrangements, cognitive schemas and models) that are helpful for the proper characterization of the two structures are presented, as well as the methodology employed to collect data for the present study. I then focus on prototypical, human subject NP constructions which reveal either an egocentric or an anthropocentric point of view of the scene. Finally, setting-subject constructions are addressed: not only are the characteristics of such structures highlighted but also the parameters and factors that contribute to their occurrence are identified. The study shows that such constructions convey the conceptualizer’s assessment of a situation, as the viewing relationship is construed subjectively. A setting-subject construction thus reveals a perspective that indirectly turns out to be more anthropocentric than ‘setting-centric’, as the inanimate locative subject, ranking at the bottom of the Animacy hierarchy, winds up alluding to any possible human being, including the speaker, the addressee and the Other.