Séminaire de John Esling
Emeritus Professor of Linguistics, University of Victoria
4 séminaires en Novembre 2014
► Lecture 1 – John Esling I : Voice and Voice Quality
(Description of VQ, definitions, anatomy, physiology, states of the larynx, phonation types, the Laryngeal Articulator Model)
This lecture presents voice quality as a component of accent, following Abercrombie (1967). The Laryngeal Articulator Model is described in detail as the foundation of a revised theory of voice quality. As an introduction to the model, the presentation defines key terminology for all of the lectures, outlines the anatomical and physiological bases for the model with reference to an extensive database, including an outline of states of the larynx and phonation types. This general overview provides a unified account of laryngeal structure and function, bringing together a vast compendium of observations that help explain laryngeal phenomena that have often been portrayed in inconsistent, confusing, and incorrect ways in the existing literature on speech physiology and in describing the sounds of the world’s languages. Prior to the methodology described in this account, the structure and functions of the laryngeal articulator – and hence of voice quality – were poorly understood, partly because the articulator could not be visualized in the production of the full range of sounds used in the languages of the world. A revised graphic depiction of the vocal tract is introduced as a basis for explaining each voice quality category.
Voice Quality Classification
(Laryngeal and supralaryngeal voice quality categories, canonical phonetic categories)
The first lecture also presents the traditional classification of voice quality based on Laver (1980), reinterpreting the categories within the Laryngeal Articulator Model. Laryngeal voice quality categories, beginning with phonatory state, are presented first, then oral (supralaryngeal) categories. States of the larynx, revised from the earlier notion of ‘states of the glottis,’ defines all possible configurations of the isthmus of the airway that are necessary to characterize the sound qualities generated in the larynx. These states can now be shown to parallel the description of phonation types, which have a long history in descriptive phonetics as the core of the concept of voice quality since they are the first source of sound modulation in the vocal tract. All voice qualities from the taxonomy are illustrated canonically (‘cardinal’ reference points defined by auditory theory). States of the larynx are illustrated with still images, and the corresponding phonation types are illustrated with phonetically modelled video clips to show the movement of laryngeal structures and the sound quality that is generated. Drawings (artistic representations) will accompany the photographic images where possible. All of the speech sounds in the voice quality taxonomy and in an expanded set of IPA consonants and vowels are presented in an iOS app – iPA Phonetics – which will be shown and explained.
► Lecture 2 – John Esling II : Infant Acquisition of Speech and Voice Quality
(Infant speech, ‘learning’ the laryngeal vocal tract, audio exemplars)
This lecture describes how infants systematically employ voice quality throughout the first year of life through active exploration of the laryngeal articulator. It is shown that in the first several months of life, all infants use the same voice qualities, regardless of their ambient languages, because their laryngeal anatomy and physiology predisposes them to do so. Infants make extensive use of laryngeal constriction at the segmental and suprasegmental levels in the first months of life, and use this constricted setting as a basis for exploring unconstricted voice qualities and the segments produced in the oral vocal tract. In other words, before they begin to employ the sounds that are used in their ambient languages, infants begin to master the laryngeal vocal tract. Data from a variety of ambient language environments will be presented to illustrate the central point that speech begins in the larynx and pharynx. Elucidating ontological phonetic development allows us to clarify the distinction between ‘voice’ (e.g. the voicing of a cry), ‘long-term voice quality’ (where long-term settings in accent originate in the individual), and ‘articulatory settings’ (how the mechanics of articulation relate to the long-term, dynamic, and segmental strands of accent defined by Abercrombie).
Data will be presented which characterize the phonetic repertory of infants during the stages of their first year. Data are auditory (with acoustic correlates). The stages of the end of the first year will be shown to reflect the gradual acquisition of oral sounds. The inherent coarticulatory nature of this process will be explained and illustrated with examples. The developmental transition into the second year of life is an opportunity to examine the proliferation of the speech sound repertory. It is still an open question whether the laryngeal sounds that are employed earliest reappear in early words among the infants whose ambient languages contain those sounds in their phonologies or whether the early sound types disappear until a later stage in the acquisition process.
► Lecture 3 – John Esling III : Phonetic Modelling of Voice Quality
(Review of the analysis of oral voice quality and of laryngeal voice quality, images, drawings, computer analysis, computer modelling)
This lecture reviews the methods that have been used to analyze oral (supralaryngeal) voice qualities and the methods used to analyze laryngeal voice qualities, particularly the methods used in generating the Laryngeal Articulator Model (laryngoscopy, videofluoroscopy, ultrasound). These methods are the basis for the book’s unique contributions to an understanding of laryngeal phonetics as the foundation for a theory of voice quality. The focus is on the instrumentation, its history and limitations, and techniques used in laryngeal observation. Multimedia modelling of the larynx within the vocal tract is also presented. 3-D models of the laryngeal articulator, both in pictures and in videos, give an enhanced view of how the states of the structures in the larynx operate in sequence to produce the various types of phonation. Some phonation types are glottal, while others recruit the participation of structures throughout the course of the epilaryngeal tube, generating what are termed complex phonation types.
(The place of the laryngeal articulator in phonological analysis, the place of voice quality in phonological analysis, the Phonological Potentials Model)
The Laryngeal Articulator Model has significant implications for linguistic theory. In this lecture, the beginnings of a laryngeal phonetic approach to phonological theory will be outlined. The lecture addresses how phonological theory can be enhanced by consideration of the implications of the LAM. For the purposes of continuity with past phonological research, the lecture will contain a brief consideration of several previous theoretical treatments and demonstrates their failure to posit a direct connectivity between the laryngeal and supralaryngeal systems (very much in parallel with the source-filter dichotomization of the larynx and supralaryngeal vocal tract). It is then proposed that introduction of such connectivity provides unification and deeper phonetic grounding of numerous facets of phonological systems : theoretical consequences of this relate to the interpretation of the ‘guttural’ class, the link between vowel quality and the laryngeal/pharyngeal state embodying voice quality, the status of so-called sphincteric/strident/epiglottalized phonatory modes principally associated with Khoisan-type languages, and their relation to other related modes found in other languages, and the relationship among tone, vocal register, and ATR phenomena. The theory formalizes these interpretations in a model which explicitly renders the interactions binding the components of the laryngeal articulator with other articulatory subsystems of the vocal tract. The model thus represents an unprecedented attempt to formalize the essential intersection between phonation, tone, vowel quality, and global vocal tract states that underlie the concept of voice quality. The model provides a basis for the analysis of both diachronic and synchronic phonological processes ; several illustrative cases are presented to demonstrate this. With a view to future empirical investigation, the lecture also delineates the hypotheses relating to the proposed interactions and emphasizes the cross-domain nature of these interactions, i.e. spanning biomechanical, aerodynamic, acoustico-perceptual, somatosensory and somatomotoric dimensions.
► Lecture 4 – John Esling IV : Linguistic Exemplars
(Language examples, laryngeal and supralaryngeal voice quality categories, images)
This lecture shows how the voice qualities in a wide range of languages are related to the Laryngeal Articulator Model in creating linguistic contrast at the segmental and suprasegmental levels. All voice quality categories from the taxonomy (laryngeal and supralaryngeal) are illustrated with reference to examples from natural speech and song, providing links to or search terms for examples of voice quality categories. The linguistic systems that illustrate the Laryngeal Articulator Model per se come from a wide variety of language families : Nuuchahnulth (Wakashan), Nlaka’pamuxcín (Salish), Tigrinya (Semitic), Iraqi Arabic (Semitic), Somali (Cushitic), Amis (Austronesian), Yi (Tibeto-Burman), Bai (Tibeto-Burman), Tibetan (Tibeto-Burman), Sui (Kam-Daic), Thai (Daic), Pame (Oto-Manguean), Cantonese (Sinitic), Chinese (Sinitic), Danish (Germanic), English (Germanic), Korean (Altaic), Bor Dinka (Nilotic), Chong (Mon-Khmer), Akan (Niger-Congo, Kwa), and Kabiye (Niger-Congo, Gur). These examples illustrate the action of the laryngeal articulator in terms of a series of valves that can operate independently of one another but that can be combined in different ways to produce the range of laryngeal sounds observed in languages. The presentation highlights how terms that are often used to describe laryngeal contrasts in languages (e.g. ‘tense’ vs. ‘lax’, ‘slack’ vs. ‘stiff,’ +/-ATR, and so on) relate to laryngeal structures and functions, including how such distinctions are manifested differently in languages, depending on the valves that are activated to create the distinctions in one language compared to another. Throughout the lecture, still images, laryngoscopic videos, ultrasound images, and accompanying acoustic spectra are used to illustrate the action of the laryngeal articulator in various languages.
Voice Quality in Linguistic Theory
(Conclusion, voice quality as the driver of sound change)
The ramifications of the Laryngeal Articular Model for the phonetic description of voice quality are summarized, as are the implications for the description of speech sounds in general. Voice production per se and long-term quality are reconciled as phonetic concepts in the explanation of how infants acquire the speech production capacity, beginning in the laryngeal articulator and gradually expanding in both vocal tract geography and in the length of time over which speech sounds are sustained and combined. Coarticulation, i.e. the presence of secondary or tertiary etc. features, is shown to be endemic – an inherent and necessary property of how human speech is acquired. Sound change is therefore explainable as a function of the ontogenetic speech acquisition process. That is, the way that human infants develop speech inherently contains the necessary elements for the operation of the processes that are at work in phonetic change. Furthermore, infants are posited to be the engine of that change. We postulate that phonetic change, when it arises, occurs as the infant and the child develop perceptual and productive phonetic capacity, together with their adult caregivers. Laryngeal articulation evolves naturally into all three strands of accent – segmental articulations, intonational overlay, and long-term voice quality.
The full inventory of voice quality descriptors, to which these lectures make reference, are in the University of Victoria app ‘iPA Phonetics’ on the Apple Store :
Christopher Coey, John H. Esling & Scott R. Moisik. 2014. University of Victoria,
iPA Phonetics, Version 1.0 . Department of Linguistics, University of Victoria.