The puzzle of misleading focus clitic position and prosody

leads us to expect that the focused argument of a focus-related morpheme should be made prominent either phonologically, by having the same focus prosody as other focus constructions; or morphologically, by adjacency of the focusing morpheme and its argument. For example, in English, sentential accent marks all types of focus, including focus on the italicized argument of ‘also ’ in (1c): (1) (a) Where are you going to eat dinner on Friday? We are going to an Italian restaurant for dinner on Friday. (b) We are going to an Italian restaurant, not a Thai restaurant. (c) We are also going to an Italian restaurant on Saturday night. However, analogous focus particles in Chitumbuka, a Bantu language spoken in Malawi, do not conform to this proposal, as the position of the particle and/or prosody do not always highlight the focused argument. The association-with-focus verbal enclitic,-so ‘also; again ’ illustrates the problem. It attaches only to verbs and is followed by a phonological phrase boundary. Notice, the verb is not always the argument of this clitic even though it is always the host, as shown in (2b). Further, boundary narrowing is consistently triggered by the clitic, not by its argument – the constituent in focus. This leads to ambiguity about what is in focus, as shown in (4b), where the verb, the verb phrase or the object could be the argument of-so. (2) (a) (n-khu-limilíra ma-púuno). I-TAM-weed 6- tomatoes ‘I am weeding tomatoes.’ (b) (Ku-limiliráa-so) (ngóomá)? You/TAM-weed-also maize ‘Are you also weeding the maize?