the focus prosody of chichewa and the stress-focus constraint: a response to samek-lodovici (2005)

samek-lodovici (2005) contributes to a well-established tradition of work on the prosody-focus interface, which proposes that, cross-linguistically, there is a correlation between culminative prosodic prominence and focus. chichewa focus prosody is problematic for the stress-focus correlation, because words with in-situ focus do not bear sentential stress, according to kanerva’s (1990) description. to account for chichewa, samek-lodovici proposes that, in essence, chichewa does not have culminative focus prosody because it does not have culminative sentential prosody. kanerva (1990:139–140) suggests that chichewa does, though, have sentence-level stress: the penult of the final word of the intonation phrase (ip) is described as having ‘considerable lengthening’ making it ‘noticeably longer’ than any ip-internal stressed syllable whether focused or not. however, he does not provide phonetic figures to support this description. surprisingly, no thorough follow-up phonetic study of chichewa has systematically investigated this issue. in this paper we present the results of an experiment conducted in malawi involving several non-linguist native speakers of chichewa, which set out to test kanerva’s (1990) and samek-lodovici’s (2005) conflicting claims about the relative length of penult vowels in focused vs. ip-final position. we show that kanerva’s claim is correct: the penult vowel of an ip is significantly longer than other penult vowels in the ip, including those of words in focus. however, our study contradicts kanerva (1990), as we find that focused constituents do not attract even phrasal stress. we suggest that the focus prosody reported is actually emphasis prosody, and that there is no obligatory focus prosody in chichewa.