Volume 28, Issue 1

One construction that has traditionally been neglected in the typological study of clause-linkage is that built on ‘let alone’ (e.g., the baby can’t even talk, let alone walk ). The present study explores this construction in a convenience sample of 47 languages. There are languages in which ‘let alone’ clauses appear not only with a clause-linking device, but also with an optional standard negative marker. Moreover, there are languages in which standard negation is forbidden in the ‘let alone’ clause. Here it is shown that optional standard negation may be semantically empty or may have an expressive-evaluative layer of semantic interpretation. On the other hand, standard negation tends to be forbidden in ‘let alone’ clauses appearing with semantically monofunctional clause-linking devices. The paper further investigates whether the analysis advanced for ‘let alone’ clauses can be generalized to other semantically negative adverbial clauses: ‘without V-ing’, ‘instead of V-ing’, and ‘before’ clauses. It is demonstrated that in these latter adverbial clauses, standard negation may be forbidden or optional. Interestingly, unlike the situation with ‘let alone’ clauses, there are languages in which standard negation may be obligatory in the ‘without V-ing’, ‘instead of V-ing’, and ‘before’ clause. In this scenario, the adverbial relations are compositionally encoded by a standard negative marker together with a general marker.

This investigation is a large-scale comparative corpus study of the oppositive contrast domain (also called “semantic opposition”) based on parallel texts. Oppositive contrast is established as a fuzzy region of the similarity space of contrast (‘but’), a domain also characterized by the occurrence of selectives (“topic markers”) and of initial non-predicative phrases in VSO/VOS-languages. Major findings are that many languages have special oppositive contrast markers and that there is a continuum between oppositive contrast markers and selectives, although truly intermediate markers are rare. The gradualness between oppositive and counterexpectative contrast is explained by semantic fuzziness and by emphasis, with strong emphasis being dependent on scales. Contrast is a rhetorical discourse relation and strong oppositive contrast can be used as a persuasive strategy aiming at establishing new common ground stepwise. The fuzziness of oppositive contrast has major theoretical and methodological implications. The encoding of the domain neither follows strict universals nor is it maximally diverse (diversity is strongly constrained). Due to its syntactic properties, oppositive contrast cannot be conceived of merely as a preestablished extralinguistic semantic domain. Furthermore, contrast exhibits a high degree of language-internal variability. General trends are reflected both by stable and by emergent grammar.

This paper explores the expression of mistaken beliefs – as in for example the boy mistakenly believes that the turtle is dead (whereas in fact it is alive) – in the Indigenous languages of Australia. It is shown that some mode of expressing this meaning is attested in around 40% of the languages in a selection of 149 language varieties. In over 90% of the languages showing some mode of expressing the target meaning, it is – or can be – achieved through grammatical morphemes or constructions more or less dedicated to the expression of mistaken beliefs. These include particles, enclitics and various types of complement construction involving verbs of thinking, that frequently also convey meanings of saying, doing and hearing, rarely that specify the thought as mistaken. In just four or five languages, however, the meaning is attested only as a pragmatic implicature of a general statement of belief. To the extent possible given the limitations of the sources, the paper examines the range of meanings and uses of the morphemes/constructions expressing mistaken beliefs.

Languages display global preferences for transitive, intransitive or underspecified roots in their verbal lexicon and correspondingly for the use of processes for deriving related concepts with different valency (cf. Nichols, Johanna, David A. Peterson & Jonathan Barnes. 2004. Transitivizing and detransitivizing languages. Linguistic Typology 8. 149–211). This classification is particularly relevant when applied to psych verbs, since variable linking is a widely recognized feature of this domain. This paper reports on the results of a larger typological study, involving 26 languages from 15 language families, aimed at investigating directionality in the psych alternation. In our data, most languages show preferences for one of the alternation strategies (augmented, reduced, undirected) which is then pervasive in their psych inventory, while the alternative patterns are marginally represented. Furthermore, the Indo-European languages of Europe stand out in being detransitivizing in the psych domain whereas transitivizing and underspecified languages do not show areal patterns. Moreover, we found a significant impact of alignment type on the occurrence of alternation strategies to the effect that reducing strategies are significantly less frequent in languages with ergative traits compared to languages without ergative traits. Our data also showed a positive effect of alignment in the undirected strategies meaning that undirected pairs are significantly more frequent in languages with ergative traits.