Interaction between Phrasal Structure and Vowel Tenseness in German: An Acoustic and Articulatory Study

Phrase-final lengthening affects the segments preceding a prosodic boundary. This prosodic variation is generally assumed to be independent of the phonemic identity. We refer to this as the ‘uniform lengthening hypothesis’ (ULH). However, in German, lax vowels do not undergo lengthening for word stress or shortening for increased speech rate, indicating that temporal properties might interact with phonemic identity. We test the ULH by comparing the effect of the boundary on acoustic and kinematic measures for tense and lax vowels and several coda consonants. We further examine if the boundary effect decreases with distance from the boundary. Ten native speakers of German were recorded by means of electromagnetic articulography (EMA) while reading sentences that contained six minimal pairs varying in vowel tenseness and boundary type. In line with the ULH, the results show that the acoustic durations of lax vowels are lengthened phrase-finally, similarly to tense vowels. We find that acoustic lengthening is stronger the closer the segments are to the boundary. Articulatory parameters of the closing movements toward the post-vocalic consonants are affected by both phrasal position and identity of the preceding vowel. The results are discussed with regard to the interaction between prosodic structure and vowel tenseness.

Prosodic realization of tonal target and F0 peak alignment in Mandarin neutral tone

Neutral tone in Mandarin is generally believed to lack tonal identity and exhibit more variability in its phonetic realization. We examined the tonal target of neutral tone in a prosodic word consisting of a full syllable (S) and one, two, or three neutral-tone syllables. In the experiment, the test words, presented in isolation and embedded in a carrier sentence, were read in two intonation patterns: declarative and interrogative. The results showed: (1) the tonal target of neutral tone is L(ow) at the end of the intonation phrase in declarative intonation and M(id) in question intonation; (2) its phonetic realization is influenced by intonation patterns, the tone of S and the number of neutral-tone syllables in the prosodic word; (3) the influence of the tone of S is more robust in shorter sequences than in longer ones with three neutral-tone syllables; (4) placement of the F0 peak in T2 (LH) and the neutral tone immediately following T3 (L) is susceptible to the number of neutral-tone syllables. It seems clear from our study that while the tonal target of neutral tone is related to prosodic structure, its actual F0 scaling is sensitive to prosodic manipulations such as intonation patterns and prosodic word length. In addition, tonelessness of neutral tone allows for more freedom in the alignment of the F0 peak, whose temporal coordination with its segmental host is, nevertheless, subject to both phonological and phonetic constraints.

Origins of vocal-entangled gesture

Humans move their upper limbs for communicative purposes during speaking. They gesture. Such movements interact on multiple levels with speaking. In connection to what is said, gestures meaningfully shape with varying means of representation. Yet, gestures also have non-representational aspects; they quasi-rhythmically pulse with prosodic structure in speech. In explaining how modern human gesturing practices emerge in phylogeny or ontogeny, it is undisputed that gestures proliferated because they provide particularly effective means to refer to absent or distal state of affairs. It suggested that displaced or deictic reference is gestures' most basic proper function. The upshot is that the non-representational pulsing quality of gesture is completely ignored as something a) that requires an explanation or b) something that can elucidate how gesture practices emerged from more basic beginnings shared with other animals. However, recent research provides evidence for direct biomechanical interaction between pulsing manual movements and respiratory-vocal activity. We argue that this physical link is enacted during infant vocal-motor babbling - way before infants learn to represent manually. Further, we argue that gesture-vocal biomechanics directly relates to the cross-species phenomenon of locomotor-respiratory(-vocal) coupling. Given that gesture-speech biomechanics has its roots in locomotor-respiratory coupling, it can be related to bipedalism and respiratory complexification, i.e., an adaptation for the faculty of speech. We conclude that the physical origins of vocal-entangled gesture run much deeper and unfolded more gradually than currently assumed. The entanglement of sound and movement arose out of natural physical coalitions between vocal, respiratory, and limb systems that are forced to interact. We thus invert current argumentation of how gesture and vocalization must have evolved and rethink what is foundational of human gesture. This perspective underlines that a more comprehensive investigation of the physical basis of bodily communication can yield new sources of semiotic significance in human and non-human animals.

Glottal stops do not constrain lexical access as do oral stops

This study explores processing characteristics of a glottal stop in Maltese which occurs both as a phoneme and as an epenthetic stop for vowel-initial words. Experiment 1 shows that its hyperarticulation is not necessarily mapped onto an underlying form, although listeners may interpret it as underlying at a later processing stage. Experiment 2 shows that listeners’ experience with a particular speaker’s use of a glottal stop exclusively as a phoneme does not modulate competition patterns accordingly. Not only are vowel-initial words activated by [ʔ]-initial forms, but /ʔ/-initial words are also activated by vowel-initial forms, suggesting that lexical access is not constrained by an initial acoustic mismatch that involves a glottal stop. Experiment 3 reveals that the observed pattern is not generalizable to an oral stop /t/. We propose that glottal stops have a special status in lexical processing: it is prosodic in nature to be licensed by the prosodic structure.

Additive particles, prosodic structure and focus sensitivity in Hungarian

The main aim of this article is to investigate the prosody-information structure interface in the analysis of the Hungarian additive particle is ‘also, too’. We present a prosodic study of narratives, collected through guided elicitation, and provide a prosodic basis for a focus-based analysis of is. Standard formal semantic approaches to the interpretation of additive particles regard additive particles as focus sensitive, hence the associate of the particle is focal and the focus interpretation (in terms of alternatives) is a significant part in its semantics. This view is considered crosslinguistically valid, although the discussion mostly concerns English. In Hungarian, the focus sensitivity of the additive particle is not directly transparent and needs more elaboration. In the relevant literature, the issue of focus marking with respect to the additive particle is has been insufficiently studied or merely stipulated. In this article, we argue for the importance of a more elaborate study of the prosody-information structure interface in the analysis of Hungarian additive particles. Accordingly, we provide data and its analysis to support our core argument and claims. Our study contributes to the overall understanding and analysis of is and to the general claims about focus marking and focus types in Hungarian. We aim to complement the standard semantic analyses by providing a prosodic analysis supporting the focus-sensitive analysis of is instead of merely stipulating an association with focus. On a more general level, we show that the various readings of additive particles can be explained by taking the prosodic patterns of the relevant constructions into account.

Introduction

Modular design of grammar: Linguistics on the edge presents the cutting edge of research on linguistic modules and interfaces in Lexical-Functional Grammar (LFG). LFG has a highly modular design that models the linguistic system as a set of discreet submodules that include, among others, constituent structure, functional structure, argument structure, semantic structure, and prosodic structure, with each module having its coherent properties and being related to each other by correspondence functions. The contributions in this volume represent the broad range and interconnection of theoretical, formal, and descriptive considerations that continues to be the hallmark of LFG.

The Technique of English Word Syllable Division in Speech Synthesis Based on Neural Network

With the increasing maturity of speech synthesis technology, on the one hand, it has been more and more widely used in people’s lives; on the other hand, it also brings more and more convenience to people. The requirements for speech synthesis systems are getting higher and higher. Therefore, advanced technology is used to improve and update the accent recognition system. This paper mainly introduces the word stress annotation technology combined with neural network speech synthesis technology. In Chinese speech synthesis, prosodic structure prediction has a great influence on naturalness. The purpose of this paper is to accurately predict the prosodic structure, which has become an important problem to be solved in speech synthesis. Experimental data show that the average error of samples in the network training process is lel/85, and the minimum value of the training error after 500 steps is 0.00013127, so the final sample average error is lel = 85 ∗ 0.0013127 = 0.112 < 0.5, and use the deep neural network (DNN) to train different parameters to obtain the conversion model, and then synthesize these conversion models, and finally achieve the effect of improving the synthesized sound quality.