Ironic twists of sentence meaning can be signaled by forward move of prosodic stress.

Prosodic stresses are known to affect the meaning of utterances, but exactly how they do this is not known in many cases. We focus on the mechanisms underlying the meaning effects of ironic prosody (e.g., teasing or blaming through an ironic twist), which is frequently used in both personal and mass-media communication. To investigate ironic twists, we created 30 sentences that can be interpreted both ironically and nonironically, depending on the context. In Experiment 1, 14 of these sentences were identified as being most reliably understood in the two conditions. In Experiment 2, we recorded the 14 sentences spoken in both a literal and an ironic condition by 14 speakers, and the resulting 392 recorded sentences were acoustically analyzed. In Experiment 3, 20 listeners marked the acoustically prominent words, thus identifying perceived prosodic stresses. In Experiment 4, 53 participants rated how ironic they perceived the 392 recorded sentences to be. The combined analysis of irony ratings, acoustic features, and various prosodic stress characteristics revealed that ironic meaning is primarily signaled by a stress shift from the end of a sentence to an earlier position. This change in position might function as a "warning" cue for listeners to consider potential alternative meanings of the sentence. Thus, beyond giving individual words a stronger contrastive or emphatic role, the distribution of prosodic stresses can also prime opposite meanings for identical sentences, supporting the view that the dynamic aspect of prosody conveys important cues in human communication. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Modelling individual and cross-cultural variation in the mapping of emotions to speech prosody.

The existence of a mapping between emotions and speech prosody is commonly assumed. We propose a Bayesian modelling framework to analyse this mapping. Our models are fitted to a large collection of intended emotional prosody, yielding more than 3,000 minutes of recordings. Our descriptive study reveals that the mapping within corpora is relatively constant, whereas the mapping varies across corpora. To account for this heterogeneity, we fit a series of increasingly complex models. Model comparison reveals that models taking into account mapping differences across countries, languages, sexes and individuals outperform models that only assume a global mapping. Further analysis shows that differences across individuals, cultures and sexes contribute more to the model prediction than a shared global mapping. Our models, which can be explored in an online interactive visualization, offer a description of the mapping between acoustic features and emotions in prosody.