Volume 73, Issue 3-4

Background/Aim: Encoding intended meanings in the type and strength of prosodic boundaries and strategies for communicating these meanings in ambient noise use similar prosodic cues. We analyze how increasing the level of ambient noise affects the realization of Slovak prosodic boundaries. Methods: Five native speakers of Slovak read sentences, manipulating the boundary type (weak, rise, fall) and the location of pre-boundary pitch accent. Ambient noise of several levels was administered via headphones. Acoustic and articulatory data (electromagnetometry) were collected. Results: Under normal condition, boundary strength is signaled with longer pre-boundary rhymes, more frequent pauses, greater crossboundary f0 resets and jaw displacement. The strength of falls is realized in crossboundary features (pauses, f0 reset), and rises in pre-boundary features (rhyme duration, f0 range). Pitch-accented rhymes are strengthened in all features, but f0 range. In noise, the increase in boundary strength is weak, and falls strengthen more than rises. F0 targets for falls and rises are adjusted in addition to noiseinduced global f0 scaling and lengthening. Conclusion: Hyper-articulation of prosodic boundaries in ambient noise is not robust and uniform; rather, durational, f0 and jaw displacement features co-create complex prosodic patterns in a complementary and synergetic manner based on affordances in normal speech.

Russian and German have pr eviously been described as ‘truncating‘, or cutting off target frequencies of the phrase-final pitch trajectories when the time available for voicing is compromised. However, supporting evidence is rare and limited to only a few pitch categories. This paper reports a production study conducted to document pitch adjustments to linguistic materials, in which the amount of voicing available for the realization of a pitch pattern varies from relatively long to extremely short. Productions of nuclear H+L*, H* and L*+H pitch accents followed by a low boundary tone were investigated in the two languages. The results of the study show that speakers of both ‘truncating languages' do not utilize truncation exclusively when accommodating to different segmental environments. On the contrary, they employ several strategies - among them is truncation but also compression and temporal re-alignment - to produce the target pitch categories under increasing time pressure. Given that speakers can systematically apply all three adjustment strategies to produce some pitch patterns (H* L% in German and Russian) while not using truncation in others (H+L* L% particularly in Russian), we question the effectiveness of the typological classification of these two languages as ‘truncating'. Moreover, the phonetic detail of truncation varies considerably, both across and within the two languages, indicating that truncation cannot be easily modeled as a unified phenomenon. The results further suggest that the phrase-final pitch adjustments are sensitive to the phonological composition of the tonal string and the status of a particular tonal event (associated vs. boundary tone), and do not apply to falling vs. rising pitch contours across the board, as previously put forward for German. Implications for the intonational phonology and prosodic typology are addressed in the discussion.

This stud y was ai med at analyzing empirical duration data for Polish spoken at different tempos using an updated version of the Coupled Oscillator Model of speech timing and rhythm variability (O'Dell and Nieminen, 1999, 2009). We use Bayesian inference on parameters relating to speech rate to investigate how tempo affects timing in Polish. The model parameters found are then compared with parameters obtained for equivalent material in Finnish to shed light on which of the effects represent general speech rate mechanisms and which are specific to Polish. We discuss the model and its predictions in the context of current perspectives on speech timing.

In this study, local and global prosodic cues for information structure are examined in the elicited production of six Bulgarian sentences. The sentences were produced in response to different questions, devised to prompt different focus realizations (broad focus and non-contrastive and contrastive narrow focus). Results show that speakers consistently differentiate broad and narrow focus by means of both local and global acoustic cues, by producing different pitch accent types on the nuclear syllable and reducing the ‘phonetic strength' of the default pre-nuclear accent in the narrow focus condition. Thus, the difference between the acoustic properties of the nuclear and the pre-nuclear accented syllables is smaller in the broad focus condition and greater in the narrow focus condition. Contrastive and non-contrastive narrow-focus accents are differentiated by local cues, i.e., by longer duration when the focus is early in the sentence and by global cues, i.e., by enhancing the tonal contrast between the nuclear prominence of CW2 and the pre-nuclear prominence of CW1 when the focus is late in the sentence.

This study examines prosody in read productions of two published narratives by 15 Russian speakers. Two distinct sources of variation in acoustic-prosodic expression are considered: structural and referent-based. Structural effects refer to the particular linearization of words in a sentence or phrase. Referent-based effects relate to the semantic and pragmatic characteristics of the discourse referent of a word, and to grammatical roles that are partially dependent on referent characteristics. Here, we examine referent animacy and the related grammatical function of subjecthood, and the relative accessibility or information status of a word. We document patterns of prosodic augmentation and prosodic reduction due to structural and referent-based factors, as evident from change in values of acoustic-prosodic measures mean intensity, duration and f0 range. Prosodic augmentation due to structural effects is observed for words positioned ex-situ, independent of their semantic, grammatical or pragmatic features. Prosodic augmentation due to referent-based effects is observed for words that are grammatical subjects with animate referents. Prosodic expression is further affected by referent information status. Discourse-given and discourse-new information show greater prosodic augmentation than inferable information. A closer look at individual speakers' production styles reveals that structural and referent-based variations occur in combination and interact.

Words like yeah, okay and (al)right are fairly unspecific in their lexical semantics, and not least for this reason there is a general tendency for them to occur with highly varied and expressive prosodic patterns across languages. Here we examine in depth the prosodic forms that express eight pragmatic functions of the Czech discourse marker jasně, including resignation, reassurance, surprise, indifference or impatience. Using a collection of 172 tokens from a corpus of scripted dialogues by 30 native speakers, we performed acoustic analyses, applied classification algorithms and solicited judgments from native listeners in a perceptual experiment. There appeared to be multi-parametric differences between jasně realizations in terms of their F0, timing and intensity patterns, which gave rise to generally consistent form-function mappings. For example, resignation seems to be realized with a falling intonation contour, relatively slow tempo, long wordinitial consonant and a short word-final vowel. Although the most significant prosodic parameters used for clustering analysis involved segment durations, all pragmatic functions were expressed by patterns of multiple features.

Two calling melodies of Polish were investigated, the routine call, used to call someone for an everyday reason, and the urgent call, which conveys disapproval of the addressee's actions. A Discourse Completion Task was used to elicit the two melodies from Polish speakers using twelve names from one to four syllables long; there were three names per syllable count, and speakers produced three tokens of each name with each melody. The results, based on eleven speakers, show that the routine calling melody consists of a low F0 stretch followed by a rise-fall-rise; the urgent calling melody, on the other hand, is a simple rise-fall. Systematic differences were found in the scaling and alignment of tonal targets: the routine call showed late alignment of the accentual pitch peak, and in most instances lower scaling of targets. The accented vowel was also affected, being overall louder in the urgent call. Based on the data and comparisons with other Polish melodies, we analyze the routine call as LH* !H-H% and the urgent call as H* L-L%. We discuss the results and our analysis in light of recent findings on calling melodies in other languages, and explore their repercussions for intonational phonology and the modeling of intonation.