The article describes an experiment in semiotic creativity and code alignment without explicit feedback. Six bird songs are presented to subjects (grouped in teams) that are asked for each song to invent a name for the relative bird. Then, in a game setting, each team presents the bird names to a different team that has to guess the songs for which the names had been originally invented. If the guess is correct both teams are rewarded with a positive score, else both are given a negative one. A tournament is organised, so that each team is coupled with all the remaining ones. Col- lected results indicate a high degree of success in name/song matching. Strategies for name in- vention are investigated by taking into account for each song spectral data and character distri- bution in the relative names. Two main strategies emerge in the obtained corpus, involving both motivatedness and arbitrariness. One relies on onomatopoeia, and exploits audible affordances in the songs in order to map them onto phonological patterns. The other deals with triggering plau- sible figurative models that are socially shared in a common encyclopedia. As the game is coop- erative, a Gricean principle is supposed to be at work in the game context. Given the latter, these two strategies are proposed to account for the largely successful code alignment among coupled teams, even without feedback. A tentative model to integrate the two paths is then introduced and discussed.
Naming Birds from their Sounds
Andrea Valle
2024-01-01
Abstract
The article describes an experiment in semiotic creativity and code alignment without explicit feedback. Six bird songs are presented to subjects (grouped in teams) that are asked for each song to invent a name for the relative bird. Then, in a game setting, each team presents the bird names to a different team that has to guess the songs for which the names had been originally invented. If the guess is correct both teams are rewarded with a positive score, else both are given a negative one. A tournament is organised, so that each team is coupled with all the remaining ones. Col- lected results indicate a high degree of success in name/song matching. Strategies for name in- vention are investigated by taking into account for each song spectral data and character distri- bution in the relative names. Two main strategies emerge in the obtained corpus, involving both motivatedness and arbitrariness. One relies on onomatopoeia, and exploits audible affordances in the songs in order to map them onto phonological patterns. The other deals with triggering plau- sible figurative models that are socially shared in a common encyclopedia. As the game is coop- erative, a Gricean principle is supposed to be at work in the game context. Given the latter, these two strategies are proposed to account for the largely successful code alignment among coupled teams, even without feedback. A tentative model to integrate the two paths is then introduced and discussed.| File | Dimensione | Formato | |
|---|---|---|---|
|
qsem 2024 Valle bozza_annotated.pdf
Accesso aperto
Tipo di file:
PREPRINT (PRIMA BOZZA)
Dimensione
697.23 kB
Formato
Adobe PDF
|
697.23 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.



