Iconicity as the motivation for the signification and locality of deictic grammatical tones in Tal

Abstract

We present novel evidence for iconicity in core morphophonological grammar by documenting, describing, and analysing two patterns of tonal alternation in Tal (West Chadic, Nigeria). When a non-proximal deixis modifies a noun in Tal, every tone of the modified noun is lowered. When the nominal modifier is a proximal deixis, the final tone of the modified noun is raised. The tone lowering and raising are considered the effects of non-proximal and proximal linkers, which have the tone features [–Upper, –Raised] and [+Raised] as their respective exponents. The realisation and maximal extension of the non-proximal tone features are considered effects of morpheme-specific featural correspondence constraints. Similarly, the exponent of the proximal linker docking on the final TBU is due to the relative ranking of the proximal-specific correspondence constraints. The association of the tone features [–Upper, –Raised] and [+Raised] with non-proximal and proximal linkers, respectively, is in line with crosslinguistic patterns of magnitude iconicity. Given that the local and long-distance realisations of the proximal and non-proximal featural affixes respectively are perceptually similar to deictic gestures, the locality of the featural affixation is considered a novel pattern of iconicity. To motivate this pattern of iconicity, we extend the notion of perceptual motivation in linguistic theory to include the crossmodal depiction of sensory imagery. Consequently, Tal presents evidence for iconicity as a motivation for morphophonological grammar.

The anatomy of iconicity: cumulative structural analogies underlie objective and subjective measures of iconicity

“…For instance, here we focus on cross-linguistic attestation because we aim to predict guesses and ratings of ideophone meanings from a bunch of languages by nonspeakers of those languages. One consequence is that we chose not to focus on form or meaning features that applied to only some of the languages in our sample, even if we were aware of their relevance -as for instance tone in West-African languages (Westermann, 1927;Akinbo & Bulkaam, 2024) and subtle vowel alternations in Korean (Kim, 1977;Kwon, 2018). If instead one wanted to explain or predict first language acquisition in, say, Japanese, then it would make sense to derive cumulative iconicity from a more detailed and partially language-specific coding scheme (e.g., one based on insights from Hamano, 1998); the prediction would be that this would correlate better with, for instance, ratings by native speakers.…”