Ahead of Publication / Just Accepted

In this paper, we introduce the notion of positional differential argument marking, a pattern in which the use of argument-coding devices, such as flags or indexes, is contingent upon the linear position of arguments in the clause. Compared to factors such as animacy and referentiality, the impact of word order on differential argument marking is understudied. In order to fill this gap, we compiled and annotated a dataset of 93 PDAM patterns, identified in a genealogically and areally diverse convenience sample of 70 languages. Most patterns involve a competition between a default, or zero, form and a non-zero form, but word order was found to affect argument indexing and argument flagging in a non-identical way. Non-zero indexing is more likely in the case of preverbal arguments. Non-zero flagging is favored in verb-edge orders and also when the argument is used in a non-default position. We hypothesize that this distinction reflects avoiding higher costs in processing referents in the discourse in the case of indexing, but more efficient comprehension of the clause in the case of flagging.

We test the dependency relations between gender and number posited by Greenberg’s Universal 36 by focusing on patterns of gender and number agreement in the noun phrase from an evolutionary perspective. To do so, we use data from Grambank, the largest existing database of morphosyntactic structures in the world’s languages. Based on data from 1,608 languages worldwide, we use a Reverse Jump Markov Chain Monte Carlo method to investigate the order of emergence of gender and number marking on adjectives, demonstratives and (for a smaller dataset) articles. Globally, our findings support Greenberg’s idea that gender marking hinges on number marking. In addition, they show that both adjectives and demonstratives play a special role in the development of noun-phrase internal agreement, in that under different evolutionary scenarios the occurrence of gender and number agreement on adjectives or demonstratives may favor the spreading of agreement to other target types. We compare these results with family-specific patterns of language change and further discuss their relevance to the general understanding of nominal morphosyntax.

This study investigates cross-linguistically the position of speech verbs and the occurrence of silent pauses relative to direct speech reports in narrative oral texts. We hypothesize that the position of speech verbs depends on the basic word order of verbs and their complements (VO vs. OV), and that the occurrence of speech verbs and pauses exhibit complementarity based on their shared function of marking the onset or end of reported speech. We use language documentation corpora from an areally balanced sample of 12 languages. We show that in most VO languages speech verbs exclusively precede speech reports, while in OV languages they may precede or follow them, or intervene, as middle verbs. Speakers of VO languages very often pause after speech reports, i.e. the position where there is no speech verb, while speakers of OV languages often pause before the speech report, and almost never between a speech report and a following ending verb. We conclude that basic word order substantially influences how speech reports are construed and prosodically phrased, in addition to asymmetries resulting from constraints on sequential ordering that are specific to complex units in general and to reported speech in particular.