An analysis of naturally occurring data reveals that the FALL–RISE intonational contour makes a context-independent contribution to utterance interpretation: the conveyance of speaker UNCERTAINTY. Two necessary conditions are proposed for the felicitous use of fall–rise: (a) some item referenced in an utterance must be perceived as linked by a scalar relationship with its context; and (b) speaker uncertainty with respect to the scale or scalar value evoked must be plausible. The conveyance of uncertainty constitutes a pragmatic contribution to utterance interpretation, which we analyse in Gricean terms as a case of conventional implicature. Other effects, such as irony or politeness, may be conveyed via this implicature.

The much-debated syntactic status of voici ‘here is’ and voilà ‘(t)here is’ in French relates principally to the theoretical status of syntactic categories and of grammatical subject. The first issue can be properly addressed only if one accepts a minimal theory of constituent structure. The surface distribution of voici/voilà then shows that they are two subjectless finite verbs with a single tense and mood: present indicative. Not surprisingly, they have lost some of these exceptional characteristics in certain dialects of French: they have acquired a personal conjugation in Québec French, and new tenses in Walloon and Franco-Provençal. These facts are relevant to syntactic theories which require all basic sentences to have grammatical subjects at some level of analysis, since no independent parameter is available to account for the systematic absence of surface subjects before a small set of verbs in a non-‘PRO-drop’ language like French.

It is well known that, in languages of the world, the passive expression is related to constructions such as reflexives and reciprocals. This paper pays special attention to the less familiar correlations of passives to spontaneous, potential, honorific, and plural formations; and it is argued that presently available characterizations of the passive are unable to offer explanations for these correlations. A satisfactory characterization must include a framework in which one can determine the nature of ‘pseudo-passives’, ‘impersonal passives’ etc. To achieve these goals, a prototype approach is adopted, and the passive prototype is defined. It is also shown that current controversies over whether passives should be analysed as promotional or demotional phenomena find a natural answer when a wider range of passive-related phenomena is examined.