The incremental processing of focus, givenness and prosodic prominence

Abstract

This study on German investigates the real-time comprehension of items in First Occurrence Focus (focused and new), Second Occurrence Focus (focused and given), Quasi Second Occurrence Focus (derogatory expressions that are referentially given and lexically new) and Background (non-focused and given), which are marked by different levels of prosodic prominence. While previous electrophysiological research tested mismatches between prosody and information structure, the present study assessed contextually licensed, appropriate prosodic realizations of stimuli. Our EEG experiment revealed distinct topographic profiles for information structure and prosody. As to prosody, we found a biphasic pattern over anterior brain regions for (secondarily prominent) phrase accents (marking Second Occurrence Focus) and deaccentuation (marking Background) but not for pitch accents (marking First Occurrence Focus), indicating an inverse relation between processing effort and the level of perceived prominence. The event-related potentials for Quasi Second Occurrence Focus items resembled First Occurrence Focus items although the former were deaccented. As to information structural contrasts, First Occurrence Focus engendered a pronounced negativity over posterior sites relative to Second Occurrence Focus and Background. Quasi Second Occurrence Focus showed an intermediate negativity. These differences can probably be accounted for by (lexically) new rather than focused information. In general, the data indicate that both prosodic cues and information structural categories influence the incremental processing of spoken language and that pitch accents and newness fulfill independent prominence-lending functions.