Prosodic prominence in Hungarian and German: same, same, but different

Katalin Mády & Ádám Szalontai (RIL HAS)

 

9. February 2017  

Former studies on prosodic prominence were based on the assumption that the presence or even existence of stress or emphasis was observable and even measurable in general. Acoustic investigations in the first half of the 20th century took into account that stress can be realised by several acoustic cues and might vary across languages (e.g. Fónagy 1958). With the growing importance of intonation research, the terms pitch accent being relevant for sentence-level prominence and word-level stress referring the potentially accentable syllable of the word emerged. Investigations of English and other Germanic languages showed that while stress is expressed by higher intensity, longer duration and segmental strengthening (such as aspiration in syllable onsets), accent is aditionally marked by specific pitch movements.

However, recent research has shown that this picture is far from being universal. First, the investigation of word-level stress usually focussed on target words bearing a sentence-level pitch accent, whereas the realisation of stress in deaccented words has received much less attention. Second, not all languages behave like English and other Germanic languages. According to the typology of Jun (2014), there are three patterns used by languages to express prosodic emphasis: (1) head marking, (2) head-edge marking and (3) edge marking. Languages in the 3rd category such as Korean do not show any sign of having a word-level stress, instead, prominence is linked with a certain phrasal position, i.e. the left edge of an accentual phrase, a lower-level prosodic phrase.

In our talk, we present two production experiments in which Hungarian and German are compared, based on the CoPaSul intonation stylisation tool (Reichel 2016). We show that while German speakers mark both sentence- and word-level prominence consistently by various acoustic cues, Hungarian speakers use these only for sentence-level prominence marking. At the same time, Hungarian speakers use boundary strength to mark stronger prominence, while Germans do not. Thus we argue that German is a head-marking language, while Hungarian is a head-edge marking one.