Focus projection need not be based on pitch accents: evidence from Georgian

Abstract

Based on experimental evidence, this paper shows that the prosodic realization of focus in Georgian is consistent with focus projection/percolation – the phenomenon by which prosodic prominence on a sub-constituent signals focus on a larger constituent – and can be accounted for with a Default Prosody-style analysis. In focus projection in English, object-focus utterances are realized in the same way as VP- or broad-focus utterances, because in all three cases the object carries the nuclear pitch accent. In contrast, in subject-focus utterances, the subject carries the nuclear pitch accent, which is incompatible with broad- or VP-focus interpretation; focus projection does not arise. Unlike English, Georgian, a ‘phrase language’, relies not on pitch accents but on boundary tones and phrasing in focus marking (Skopeteas & Féry 2010; 2016). This language type has not been explicitly addressed from the perspective of focus projection. Nevertheless, the results reported here demonstrate that the prosodic realization of subject- and object-focus in Georgian, expressed with boundary tones and duration of the stressed syllable, fits with the focus-projection pattern. This paper shows that the Georgian data can be accounted for with the Default Prosody approaches to focus projection, but not approaches employing the formal mechanism of F-projection. Accordingly, the Georgian facts provide a novel argument in favor of the Default Prosody approaches.