Amazonia and the Typology of Tone Systems

The typology of tone and word-prosodic systems in general has been of longstanding interest to theorists and field workers alike. Unfortunately, much of the early typologizing dates from a time when we did not have adequate understanding of the full range of tonal phenomena across languages, e.g. African and Asian tone systems figured centrally in such typologizing, e.g. Pike’s (1948) famous distinction between register vs. contour tone languages. Even though some of the Mexican tone languages were beginning to be understood, they too were subordinated to the relatively well-studied and larger languages of Africa and Asia. Recognizing obvious areal distributions, we now know that tone is, with the exception of Australia, widespread throughout the globe, e.g. in the lower Pacific and in the Amazon. The purpose of this paper is to address the question of how the tonal systems of South America fit into the overall scheme of tone system typology: How they are the same vs. how they are different? While it may come at first as a disappointment that the vast majority of Amazonian systems are limited to two tone heights, the range of phenomena is considerable. First, there is the question of how the two heights should be analyzed, particularly in terms of underlying specification. As in the case of two-height systems elsewhere, the Amazonian cases neatly divide into those which have an “equipollent” two-tone opposition between /H/ and /L/ vs. those which have a “privative” one-tone system, either /H/ vs. Ø or /L/ vs. Ø (lack of tone). Still others have a combined equipollent+privative system: /H/ vs. /L/ vs. Ø. While none of these constituents a heretofore unknown “type”, as an outsider to the continent, what strikes me is the special status of /L/ tone in so many Amazonian tone systems: A number of languages treat /L/ as the “marked” tone which often repels other /L/ tones (some do not permit two /L/ tones in sequence), and which intersects in interesting ways with laryngeal features, the focus of this volume. Drawing from a database of approx. 600 tone systems, of which (so far) 53 are from South America, I will attempt to situate the Amazonian Hs and Ls within the spectrum of tone systems found throughout the world, both with respect to their phonology and to their linguistic function (lexical vs. grammatical). While I will argue that the key issue with two-height systems is to determine which tone or tones are “phonologically active”, motivating the above distinction between privative and equipollent systems, I end with discussion concerning the interaction of tones with syllable structure, segmental phonology, and of course, laryngeal features.