First #TidyTuesday in a very long time. This one made in a record (short) time.
D&D monsters seem to have longer names the larger they are. More is more #iconicity?
🔗: https://github.com/borstell/tidytuesday/tree/main/2025/2025-05-27
New! A multi-methods toolkit for ideophone research, w/ Bonnie McLean
On the blog: https://ideophone.org/new-methods-for-ideophone-research/
We present evidence for iconicity as the motivation for two patterns of morphophonological alternation in Nigerian Pidgin, also known as Naijá. To express an ‘unconventional positive’ in all varieties of Naijá, some nouns with the tone melodies H-L and L-H surface with the tone melodies L-H and H-L, respectively. In addition to unconventional positive, the Wafi variety of Naijá also expresses ‘unconventional negative.’ In this case, the first syllable of bisyllabic nouns (analogously, the first foot of four-syllable nouns) is transposed with the second syllable. However, in onsetful monosyllabic nouns, the initial consonant is deleted to express an unconventional negative. We account for the metatheses and truncation using transderivational faithfulness constraints and other independent constraints. Expressing the notion of unconventionality by changing the prosaic linear order of phonological elements in a word is a kind of form-meaning resemblance. Similarly, the association of (unconventional) negative with a reduction in a string of segments is another pattern of iconicity. This suggests that crossmodal depiction of sensory imagery, in addition to articulatory and auditory factors, can motivate morphophonological patterns. The morphophonological metatheses and truncation in Naijá contribute to the typology of morphophonological alternations with iconicity as their motivation. Most importantly, these patterns run counter to the claim that pidgins and creoles have simplified morphophonology.
New preprint! Triangulating iconicity, with Stella Punselie & Bonnie McLean (Uppsala) https://doi.org/10.31234/osf.io/z9a4e
We combine three ways of looking at the iconicity of 239 #ideophones from 5 languages: (i) coding structural correspondences between form and meaning; (ii) collecting subjective iconicity ratings, and (iii) relating these to performance in guessing tasks. Triangulation contributes to an empirically grounded, theoretically informed, experimentally robust understanding of #iconicity.
Having fun with logical types
Four years since this paper came out. We did it as a side project so the epigraph is doubly apposite
We investigated the utility of vector-based distributional semantics for lexical norm imputation — and cautioned against uncritical use. Today I see people using LLMs to do this and our caution still holds: "Norm imputation can distort rating scales and can amplify rating artefacts"
+ open data & code https://doi.org/10.1017/langcog.2019.49
📢 New paper! 📄
As had been shown with lexical signs previously, signers perceive also their own manual alphabet as generally more iconic than a foreign one.
However, signers still agree that some handshapes are more iconic relative to their corresponding written character, regardless of language. For example, 🤞🏻 is rated as more iconic when mapped to <x> (Swedish SL) than to <r> (American SL).
This project presents a case of a tonal alternations and phonological locality that are motivated by iconicity in Tal, an endangered an understudied. The language data that form the basis of the discussion are presented here. Hosted on the Open Science Framework
I was watching the Oscar nominated for best animated short film "Ninety-Five Senses" and thinking about how this production would benefit from the linguistics/cognitive sciences research on ideophony and iconicity...
it is possible to watch it online at: https://www.docplus.com/details/ninetyfive-senses/FwnPa7Bz/
Annika Schiefner
Iraide Ibarretxe-Antuñano
Calle Börstell
Mark Dingemanse
Samuel Akinbo
gabi