This paper describes Bantu imperatival and prohibitival speech acts. The study is set against the background of the formal instability of directives and grammaticalization theory. On the basis of a sample of 100 languages, we conclude that imperatival strategies are limited to imperatives, subjunctives, and indicatives while prohibitival strategies range from negative subjunctives and negative auxiliary constructions through constructions with prohibitive markers and negative infinitives to negative indicatives and negative imperatives. Politeness is shown to play an important role in the development of new strategies, which often have a more polite character and which become neutral themselves over time. We argue that it may even partly explain why prohibitival strategies exhibit more variation than imperatival ones. However, it is also clear that new directive strategies need not be more polite and that politeness is just one of the possible factors contributing to the difference between imperatival and prohibitival strategies.

No abstract

This paper is the first in-depth study of the main human impersonal pronouns in Afrikaans: jy 'you', ('n) mens '(a) human' and hulle 'they'. It adopts a double questionnaire approach, consisting of an acceptability judgment task for one group of participants and a completion task for another group. On the theoretical side, we test the different dimensions proposed in two of the most recent semantic maps of human impersonal pronouns. The first map features vague, inferred and specific existential uses, which vary in the kind/degree of (un)knownness. The second one distinguishes existential contexts that only allow a plural interpretation from existential contexts that are neutral with respect to number. The results of our questionnaires indicate not only that the dimensions of number and (un)knownness involve gradual instead of categorical distinctions but also that they interact with one another, with decreasing acceptability and usage of hulle along both of them. More generally, the completion task data suggest that human impersonal pronouns are not the preferred strategy for impersonalization in existential contexts anyway. On the descriptive side, we show that Afrikaans has a division of labor between ('n) mens and jy on the one hand and hulle on the other. The former are restricted to universal-internal uses, the latter to universal-external, speech act verb and existential ones. The data also reveal that speakers may consider the less grammaticalized form 'n mens more acceptable but that they tend to employ more grammaticalized mens. It thus attests to the usefulness of combining the two types of questionnaire. Keywords impersonal, pronoun, Afrikaans, questionnaire, semantic map 1. are a collective that has the right and/or power to install speed cameras. The police and the government are two likely candidates. This semi-impersonal existential use is typically called 'corporate' (a term that goes back to Pesetsky 1996, p. 39). (2) a. Eng In Greece, they drive quite unpredictably. b. Eng They have installed new speed cameras here. The three uses that Siewierska and Papastathi (2011) regard as genuinely impersonal are illustrated in (1b) and (3). They differ in the type or level of (un)knownness, as argued by Cabredo Hofherr (2006) among others. In the 'vague' use in (1b), the set of human participants cannot be identified by the interlocutors but there is said to be at least one specific person who committed the known act of stealing the car. In (3a), the speaker gathers from the situation, i.e. the smell in the room, that the essentially unknown event of consuming pizza there must have occurred and that an unidentifiable (group of) eater(s) must have existed. This use is labeled 'inferred'. In the 'specific' use in (3b), the event takes place at a particular place and time and the interlocutors may thus have certain expectations about who is performing it. Despite the situational potential for identification, the (set of) individual(s) knocking on the door is not explicitly named, however. (3) a. Eng They ha...

No abstract

No abstract