Journal of Historical Linguistics - Volume 8, Issue 1, 2018

This special issue of JHL reconstructs the diachrony of a number of innovations in the coding of argument structure, particularly in the domain of verbal indexation, in four Amazonian language families (Chapacuran, Sáliban, Tukanoan and Tupi). It is one result of an international workshop on “Diachronic Morphosyntax in South American Languages” held in Lyon (France) in 2015, with financial support from the Collegium de Lyon (Institute for Advanced Study) and the LabEx ASLAN of the Université de Lyon. The goal was to encourage methodologically innovative (and more rigorous) historical studies of morphosyntactic patterns in languages or language families of South America. The five papers that comprise this collection all demonstrate the viability of syntactic reconstruction, even in languages with little or no written history.

The reported speech construction found in the Chapacuran language family of South America has undergone a number of changes in the individual languages, such that its uses extend beyond that of merely reporting speech. In many languages, it is used to express the inner states of the reported speaker, and in some cases it is used to express imperfectivity and causation. This paper argues that the future construction in Moré is a further development of the reported speech construction, one that has been reanalyzed as a basic main clause type. The morphosyntactic properties of the source construction explains the divergent inflectional forms, the loss of object indexation, and the innovation of an object case marker in the future construction. This paper provides new insights into the diachronic pathways that can lead to innovative future constructions as well as the origins of a tense-based split in case marking in Moré.

Piaroa, a member of the Sáliban language family, is spoken on both sides of the Colombian-Venezuelan border. Based on unpublished fieldwork data for Mako and Piaroa and published Piaroa and Sáliba data, this article focuses on the Piaroa subject marking system and its origins. I show that the subject prefixes and inner suffixes used in future tense were inherited from Proto-Sáliban and must therefore have preceded the rise of the right-margin subject markers ‑sæ, -hæ and ‑Ø. Based on comparative Mako data, I propose that these markers are old copular suffixes that entered the verbal domain through a nominal predication construction whose use expanded to encode habitual aspect. This research not only constitutes an important contribution to the description of Piaroa but also expands, within a Diachronic Construction Grammar approach, our understanding of complex systems of person marking, the origins of multiple exponence, and the role of multiple source constructions in paradigm creation.

This article describes the evolution of past/perfective subject-verb agreement morphology in the Tukanoan family, reconstructing relevant aspects of Proto-Tukanoan verbal morphology and delineating the subsequent diachronic development of verbal subject agreement morphology in the Eastern branch of the family. We argue that suffixes that cumulatively expone past/perfective and person, number, and gender (png) subject agreement resulted from the fusion of post-verbal demonstratives/pronouns expressing png information with suffixes expressing past/perfective tam information. We propose that different png agreement categories developed at successive stages in the diversification of the family, with third person masculine singular subject agreement emerging before other png categories, followed by animate plural agreement, then finally by the development of third person feminine agreement. The result in Eastern Tukanoan was a cross-linguistically unusual agreement system that contrasts four agreement categories: (i) first and second person singular and third person inanimate (singular and plural); (ii) third person animate masculine singular; (iii) third person animate feminine singular; and (iv) third animate plural.

This article proposes that the divergent pattern of verb argument marking found in object focus clauses in the Tuparian branch of the Tupian family comes from the reanalysis of an object nominalization in a cleft construction. Based on the distribution of free and bound person markers, the major alignment pattern can be characterized as nominative-absolutive in simple clauses, with free pronouns expressing the nominative, whereas bound person markers express the absolutive. However, object focus clauses show a distinct alignment pattern: the ergative, and not the absolutive, is indexed by the bound markers on the verb. We present arguments for identifying the object nominalization as the source of this grammar in the object focus clause, showing also how this reanalysis resulted in the nominalizer morpheme and the person markers gaining new functions.

Subject marking in the Western Tukanoan language Ecuadorian Siona is part of a complex system of portmanteau morphology that also marks tense and clause type. This system shows a remarkable number of regularities that hint that it might be possible to tease apart these functions. Synchronically, it is problematic to posit distinct markers for each of the three relevant linguistic categories, but diachronically, it is likely that these categories were expressed by distinct markers. This article reconstructs the pathway of the formal merger of these three linguistic categories, comparing the expression of Ecuadorian Siona’s system to the expression of these categories in other Tukanoan languages.