Prosodic Correlates Research Articles

Cue words and the topic structure of spoken discourse: The case of Swedish men ‘but’

The Swedish cue word men ‘but’ can mark the boundary between both different topic units as well as topic-internal units in spontaneous speech. The goal of this study is to see if these two functions of men can be distinguished on the basis of their local prosodic correlates and co-occurring lexical items. Men-tokens in spontaneous narrations were labelled as to their function, first using text-only data. The ‘strong’ tokens (categorized identically by all labellers) were subsequently seen to be clearly differentiated into two classes on the basis of related prosodic parameters and co-occurring lexical items. This distinction was, however, not found for the corresponding ‘weak’ tokens which were subsequently relabelled using both text and speech nor for the data-base as a whole. A test using a neural network trained using strong tokens was seen to be able to correctly categorize 90% of the strong men-tokens as to their associated boundary-type (topic-shift vs. topic-internal). The results show that cue words along with their prosodic correlates and co-occurring lexical items constitute a constellation of important information for understanding how segmentation of spoken discourse is produced and understood.

Prosodic and lexical indications of discourse structure in human-machine interactions

From a discourse perspective, utterances may vary in at least two important respects: (i) they can occupy a different hierarchical position in a larger-scale information unit and (ii) they can represent different types of speech acts. Spoken language systems will improve if they adequately take into account both discourse segmentation and utterance purpose. An important question then is how such discourse-structural features can be detected. Analyses of monologues and human-human dialogues have shown that a good indicator of these factors is prosody, defined as the set of suprasegmental speech features. This paper explores whether speakers also use prosody to highlight discourse structure in a particular type of human-machine interaction, viz., information query in a travel-planning domain. More specifically, it investigates if speakers signal (i) the start of a new topic by marking the initial utterance of a discourse segment, and (ii) whether an utterance is a normal request for information or part of a correction sub-dialogue. The study reveals that in human-machine interactions, both discourse segmentation and utterance purpose can have particular prosodic correlates, although speakers also mark this information through choice of wording. Therefore, it is useful to explore in the future the possibilities of incorporating prosody in spoken language systems as a cue to discourse structure.

Duration and fundamental frequency correlates of phrase boundaries in productions by children and adults

Two experiments examined how children and adults produce acoustic correlates for phrase boundaries during speech. Adult and children (ages 5 and 7) were asked to describe groupings of colored blocks, first in a relatively spontaneous manner and next under more structured conditions designed to elicit grouping information concerning "which blocks go together." In the spontaneous condition, adults gave elaborate descriptions of block positions and color, whereas children produced short descriptions such as "pink, green, white;" "pink, green, and white;" and "pink and green and white." In the structured condition, all subjects produced utterances corresponding to three syntactic bracketings of the phrase "pink and green and white:" [pink and (green and white)]; [(pink and green) and white]; and [pink and green and white]. Acoustic analyses indicated that adults reliably control both duration and fundamental frequency (F0) to signal phrase boundaries, whereas children of both age groups demonstrate little evidence of either type of information being used. We interpret these finding as suggesting that children as old as age 7 do not produce prosodic cues for this type of standing ambiguity in their everyday speech. Possible reasons for this lack of phrase boundary prosodic correlates are examined.

Linguistic aspects of speech synthesis.

The conversion of text to speech is seen as an analysis of the input text to obtain a common underlying linguistic description, followed by a synthesis of the output speech waveform from this fundamental specification. Hence, the comprehensive linguistic structure serving as the substrate for an utterance must be discovered by analysis from the text. The pronunciation of individual words in unrestricted text is determined by morphological analysis or letter-to-sound conversion, followed by specification of the word-level stress contour. In addition, many text character strings, such as titles, numbers, and acronyms, are abbreviations for normal words, which must be derived. To further refine these pronunciations and to discover the prosodic structure of the utterance, word part of speech must be computed, followed by a phrase-level parsing. From this structure the prosodic structure of the utterance can be determined, which is needed in order to specify the durational framework and fundamental frequency contour of the utterance. In discourse contexts, several factors such as the specification of new and old information, contrast, and pronominal reference can be used to further modify the prosodic specification. When the prosodic correlates have been computed and the segmental sequence is assembled, a complete input suitable for speech synthesis has been determined. Lastly, multilingual systems utilizing rule frameworks are mentioned, and future directions are characterized.

Attitudinal judgments in intonation: Some evidence for a theory

In an attempt to resolve the seeming contradiction between lack of direct correspondence between intonation and emotive meaning on the one hand, and a beyond chance recognition of emotions as judged on the basis of prosodic cues on the other, the author examines results of the relevant psycholinguistic studies. The survey includes evidence related to such pertinent issues as recognition accuracy in judgments involving vocal material, prosodic correlates of affect conceived of as dimensional structuring, and the relative contribution of vocal and visual channels in affective communication. The discussion leads to the postulation of an intermediary level of intonational description, following the fact that intonation patterns cannot express fully specified emotive meanings in a reliable way, but rather serve as devices indicating the activation dimension of affects. The proposed level of description, considered here to be the only “semantic” property of intonation contours, would refer to the potential of individual contours to reflect given degrees of activation and be thus related to both the formal properties of intonation and the actual emotions portrayed by speakers.