| |
|
| |
|
Title:
|
The Role of Aerodynamics in Phonology
|
Organizers:
Presenters:
|
John Ohala University of California, Berkeley
'Accommodation to the aerodynamic voicing constraint and its phonological relevance'
Ryan K. Shosted University of Illinois at Urbana-Champaign
'An EMA-aerodynamic approach to the velic opening hypothesis: evidence from Hindi vowel pairs'
Didier Demolin GIPSA-Lab, Université Stendhal Grenoble III, Grenoble
'The influence of aerodynamic constraints on the shape and the dynamics of phonological systems'
Maria-Josep Solé Universitat Autònoma de Barcelona,
Ronald Sprouse University of California, Berkeley
'Voice-initiating gestures in Spanish: prenasalization'
|
Summary Description:
|
While articulation and more recently perception have essentially contributed to our under-standing of phonological patterns and processes, aerodynamics is still under-represented in phonetic and phonological research.
In this workshop we would like to draw the audience's attention to the valuable contribution of aerodynamics to explaining phonological phenomena. As some recent studies have shown, aerodynamic factors play a key role in explaining both synchronic and diachronic phonological processes in which for example voicing, nasalization, trilling, affrication, implosivization are involved (cf. work by Demolin, Ohala, Shosted, Solé). Aerodynamics also appears to be indispensable to understand cross-linguistic gaps in phonemic inventories, such as the avoidance of voiced sibilants or voiced velar stops (Maddieson, Ohala). Furthermore, phonological processes which have traditionally been considered outside the scope of aerodynamics, such as strengthening (e.g., epenthetic stops, postnasal occlusivization), weakening (e.g., final fricative weakening), or certain types of assimilation may be better understood in terms of contextual/positional variation in the aerodynamic conditions.
In this workshop we will not only review phonological phenomena which have been explained by aerodynamic factors but we will also point to selected areas in phonology which could greatly benefit from this approach. For example, relatively little has been done to apply aerodynamic evidence to prosodic phonology although its contribution maybe essential in understanding, for example, the shaping of boundaries. Furthermore, some phonological features, despite being rooted in the aerodynamics, have not been explored from this perspective, specifically with respect to their universal character and/or language-specific variation. Finally, the aerodynamics of several sound types and especially their occurrence in clusters requires further research in order to better understand phonotactics and its preferred patterns in the world's languages.
The presentations in the workshop will address some interesting and novel phonological questions, such as different synchronic and diachronic outcomes of phonological nasal vowels related to differences in velopharyngeal opening –using a synchronized EMA-Aerodynamic system– (Shosted); the use of articulatory adjustments –such as nasal leakage– to achieve pressure and flow values for initiating voicing in languages that rely on occurrence of vocal fold vibration for the voiced-voiceless distinction (Solé and Sprouse); the connection between the feature [ATR] and voicing in obstruents (Ohala); and the role of aerodynamic factors in various sound patterns (Demolin). The workshop will also provide a forum for phoneticians and phonologists to discuss aerodynamic modelling and its use in phonological research. Finally, an overview of the latest technical developments relevant for aerodynamic research will be given with the focus on linking aerodynamics and articulatory data.
|
|
|
|
|
Title:
Organizers:
Presenters:
|
Sonya Bird University of Victoria, Canada
'Phonetic fieldwork in the Pacific Northwest'
Marc Brunelle University of Ottawa
'Perception in the field'
Didier Demolin GIPSA-Lab, Université Stendhal Grenoble III, Grenoble
'Aerodynamic techniques for phonetic fieldwork'
Jerold A. Edmondson University of Texas at Arlington,
Celine Chang National Tsing Hua University, Taiwan,
Feng-fan Hsieh National Tsing Hua University, Taiwan,
Jennifer Huang Academia Sinica, Taiwan
'Laryngoscopic fieldwork: a guide'
|
Summary Description:
|
Whatever the instrumental or areal focus, phonetic fieldwork is based on the analysis of observations or recordings obtained in the field. However, the practical and pragmatic aspects of phonetic research methodology are often discussed only in passing. As a result, the `best practices' of phonetic field methods remain largely unexplored, passed on only in the form of oral tradition from teacher to student. While several excellent general treatments of the techniques and methods of linguistic fieldwork exist, there have been fewer treatments of the methodologies and practices specific to phonetic fieldwork. The first purpose of this Special Session is to bring together researchers with expertise in phonetic fieldwork to share and discuss experiences and techniques, and to communicate them to the broader community of phonetic scientists.
A second motivation for this session is that there are reasons for all phoneticians, regardless of geographic, phylogenetic, or instrumental focus, to think about research methodology more carefully than we may have in the past. This is due in no small part to the fact that, as we continue to ask more and different questions (about more and different languages), we must continue to explore new ways of addressing them. As the focus of the field shifts from the examination of relatively homogeneous societies to more disparate and more heterogeneous ones, the list of variables which must be taken into account -- which in the past have included biological characteristics such as age and sex along with attributed characteristics such as gender and ethnicity -- must now be expanded to include traits that are harder to observe with the naked eye, such as social identification, tastes, and preferences along with individual identity. With new technologies for storing and processing information appearing with increasing rapidity, sources of data multiply and data sets will continue to grow.
How should field phoneticians prepare to meet these challenges, and what methodological best practices are recommended? Our panel of expert field phoneticians will discuss research methodology using illustrations from a diverse array of geographic and linguistic areas. The papers in this Special Session will emphasise three areas of research methodology:
Developing research questions. Often, field phonetic research begins when phoneticians find themselves in sites or situations that seem as if they might lend themselves to collecting some potentially interesting data. In such cases, the researcher may not have had the opportunity to explicitly formulated a research question in advance. What are some techniques for conducting methodologically sound experiments and data collection 'on the fly'?
Phonetic field methods. How do phoneticians assemble materials for analysis? With technological advances, it has become increasingly easier to transport the methods and equipment of modern laboratory phonetic data collection and analysis to the field. This has opened up the possibility to greatly expand the range of experimental data available for linguistic consideration. At the same time, these innovations are enabling researchers without access to permanently installed phonetics laboratory facilities to engage in high-quality phonetic research themselves. Each of our invited speakers has been tasked with describing and illustrating a range of field methodologies, such as transcription techniques, elicitation strategies, and the instrumental measurement of articulatory and aerodynamic data. Contributed papers will also consider the important issue in phonetic fieldwork concerns the ethical aspects of human subject research in the field. How are the rights of the phonetic consultants protected and what pitfalls might phonetic fieldworkers encounter that differ from fieldworkers who are primarily interested in higher order linguistic information?
Analytical methods. After the recordings are made or instrumental measurements taken, how should phonetic data be analysed? While the specific approach is again highly dependent on the data in question, there are some general best practices that apply to a broad range of situations. Participants will address a variety of analytic heuristics for quantitative study, such as lists of questions to ask about the data to ensure that nothing has been carelessly omitted, and will discuss the uses and abuses of counting and calculating in research that might run afoul of the assumptions of proper statistical analysis.
This Special Session will feature talks by four prominent field phoneticians, followed by a roundtable discussion. In addition to representing a wide variety of typological and experimental expertise, all of the participants have personal experience investigating the phonetic properties of lesser-studied languages in non-laboratory conditions using a range of instrumental and experimental techniques.
|
|
|
|
|
Title:
|
Phonetic Teaching and Learning: Recent Trends, New Direction
|
Organizers:
Presenters:
|
David Deterding University of Brunei Darussalam
‘English language teaching and the lingua franca core in East Asia’
Jose A. Mompeán University of Murcia, Spain,
Michael Ashby University College London,
Helen Fraser University of New England, Australia
‘Phonetics teaching and learning: an overview of recent trends and directions’
Joanna Smith Unitec Institute of Technology, New Zealand
‘The Youtube revolution: engagement, perception and identity’
Magdalena Wrembel Adam Mickiewicz University, Poland
‘Cross-modal reinforcements in phonetics teaching and learning: an overview of innovative trends
in pronunciation pedagogy’
|
Summary Description:
|
Theme. The special session will aim to bring together phoneticians interested in phonetics pedagogy (contents, methods, tools and resources, assessment, etc.), including the application of new technological developments to the teaching and learning of phonetics (IT-based teaching, distance/online education, etc.) and the intersection of phonetic science with different teaching and learning contexts (phonetics theory teaching, language teaching, etc.).
Form. The Phonetics teaching and learning special session (1hr 20m) will consist of an invited paper surveying the field, two papers describing new developments drawn from those submitted to the congress on pedagogical topics, and a discussion forum. From past experience, the total number of acceptable papers submitted to the congress on pedagogical themes is likely to be at least 20–30. The special session will not replace the oral and poster sessions which those papers would in any case require, but rather serve as a focal point for all pedagogically-oriented work within the Congress. Organisers of the session can provide reviewing expertise to assist the local organisers, and aid in the integration and timetabling of pedagogically-related papers and posters across the whole event.
Background. The proposal builds upon the growing interest in pedagogical issues in phonetic science. In this respect, previous ICPhS congresses have commonly incorporated phonetics pedagogy themes and sessions. To consider only the two most recent, at ICPhS XV in 2003, there was a Symposium “Tools for teaching phonetics” (4 oral presentations), a “Pedagogy” poster session (8 posters) and a “Pedagogy” oral session with 6 presentations. Under various aspects of Second Language Acquisition, there were 22 oral presentations and 25 posters. At ICPhS XVI in 2007 there was a Special Session “Second Language Acquisition and Exemplar Theory” with 7 papers, three oral sessions (9 papers in total) on aspects of “Foreign language acquisition”, including one session on “new methodology” plus a number of posters.
|
|
|
|
|
Title:
|
Shapes and Tones - Towards a More Holistic Perspective in Intonation Research
|
Organizers:
Presenters:
|
Jonathan Barnes Boston University,
Alejna Brugos Boston University,
Nanette Veilleux Simmons College,
Stefanie Shattuck-Hufnagel M.I.T.
'Voiceless intervals and perceptual completion in F0 contours:
Evidence from scaling perception in
American English'
Hyesun Cho Seoul National University,
Edward Flemming M.I.T.
'The phonetic specification of contour tones: the rising tone in Mandarin'
Ingo Feldhausen Collaborative Research Center 538 ‘Multilingualism’ Hamburg and University of Hamburg,
Andrea Pešková Collaborative Research Center 538 ‘Multilingualism’ Hamburg,
Elena Kireva University of Hamburg,
Christoph Gabriel Collaborative Research Center 538 ‘Multilingualism’ Hamburg and University of Hamburg
'Categorical perception of Porteño nuclear accents'
Oliver Niebuhr University of Kiel,
Mariapaola D’Imperio CNRS, Aix-en-Provence, France,
Barbara Gili Fivela University of Lecce, Italy,
Francesco Cangemi CNRS, Aix-en-Provence, France
'Are there ‘shapers’ and ‘aligners’? Individual differences in signalling pitch accent category'
|
Summary Description:
|
It is probably not much of an overstatement to say that the historic question of whether intonation is to be analyzed in terms of tones or configurations was resolved in current theoretical approaches by projecting tones and configurations onto different levels of representation. At the level of phonology the intonation contour is decomposable into a number of separate high(er) or low(er) tonal targets, whose nature is both static and acoustic, and which are hence more or less strictly identifiable with local f0 peaks and valleys (“elbows”). The target tones are associated with different types of suprasegmental units that represent pivotal points in the lexical and phrasal (incl. metrical) structure of the speaker’s utterance. Furthermore, each type of suprasegmental unit has its (language-)specific inventory of tonal contrasts. So, expressing different intonational meanings is a matter of discrete changes in the order and/or the f0 levels of the high and low target tones. The concrete implementation and concatenation of the phonological target tones in utterances yields complex intonational configurations at the phonetic level. In these configurations the alignments and scalings of the targets as well as the (linearly) interpolated slopes between them vary gradually due to contextual, dialectal, idiosyncratic, and paralinguistic factors.
The outlined autosegmental-metrical (AM) concept has guided intonation research since more than 30 years now and inspired a whole generation of speech scientists to study the production and perception of f0 contours within and across languages. The resulting impressive “accumulation of evidence” substantially advanced our understanding of the patterns and functions of intonation in spoken communication. More recently, however, the accumulation of evidence also yielded findings that raised questions about fundamental claims of the AM phonology. For instance, studies on various languages undermined the concept of local acoustic target tones. As for production studies, for example, work on Korean indicated that the alignment specifications of the tones involved in a rise are not completely independent of each other. Similarly, as regards pitch-accent productions in Italian and German, speakers show individual strategies that can be projected onto a continuum from “aligners” to “shapers”. For the latter two languages it was also found that the slopes concatenating tonal targets of both prenuclear and nuclear pitch accents have different shapes (i.e. concave or convex) depending on the expressed focus or sentence-mode categories. Investigations into the perception of Italian, English, and Frisian varieties showed that f0 peaks and plateaux are associated with different pragmatic meanings, and phrase-final rises with different shapes proved to fulfill different discourse-related functions in German. Addressing a different issue than the status of local targets, studies on German and Italian showed that intonational meaning is not conveyed by f0 alone, but that other cues (e.g., fricative spectra, vowel intensities and durations) play a role as well.
All these findings cross the lines drawn by the AM framework between phonology (i.e. local tonal targets) and phonetics (i.e. configurations, interpolations) as well as between intonation (i.e. f0) and segments or other prosodies. In this way, the findings actuated controversial and fruitful discussions in the field, which culminated for the first time in the special session on “Shape, Scaling, and Alignment of f0 Events” held at the 5th International Conference of Speech Prosody, Chicago, USA (organized and chaired by Jonathan Barnes, Nanette Veilleux, Alejna Brugos and Stefanie Shattuck-Hufnagel). This special session presented on the one hand new evidence of production and perception experiments on the relevance of f0 slopes, shapes and plateaux. On the other hand, the session offered possible ways to handle this evidence on the basis of the AM approach. We think that this debate, which was started in the special session of the previous Speech Prosody conference, and the body of evidence that stimulated the discussion and that has again grown since then, have the potential to lead to a new major turning point in intonational theory and modeling. At the same time, we think that this turning point will not be a turning back in the sense of a complete rejection of the notion of tonal targets in favor of configurations. Rather, it will be a turn towards a more global and more balanced perspective that goes beyond mere local f0 values and integrates, for example, both tones and configurations, the listener and the traditional segmental level. Facilitating joint analyses of melodic and segmental aspects would also be a great benefit for the whole field of phonetic research. An additional benefit lies in the fact that the more detailed investigation of intonational patterns requires a stronger focus on intonational functions as a tertium comparationis in comparative studies, which has already led to the discovery of new form-function relationships. The awareness that local f0 targets are not sufficient is constantly growing, and common efforts are needed to define future research directions.
This special session presented on the one hand new evidence of production and perception experiments on the relevance of f0 slopes, shapes and plateaux. On the other hand, the session offered possible ways to handle this evidence on the basis of the AM approach. We think that this debate, which was started in the special session of the previous Speech Prosody conference, and the body of evidence that stimulated the discussion and that has again grown since then, have the potential to lead to a new major turning point in intonational theory and modeling. At the same time, we think that this turning point will not be a turning back in the sense of a complete rejection of the notion of tonal targets in favor of configurations. Rather, it will be a turn towards a more global and more balanced perspective that goes beyond mere local f0 values and integrates, for example, both tones and configurations, the listener and the traditional segmental level. Facilitating joint analyses of melodic and segmental aspects would also be a great benefit for the whole field of phonetic research. An additional benefit lies in the fact that the more detailed investigation of intonational patterns requires a stronger focus on intonational functions as a tertium comparationis in comparative studies, which has already led to the discovery of new form-function relationships. The awareness that local f0 targets are not sufficient is constantly growing, and common efforts are needed to define future research directions.
|
|
|
|
|
Title:
|
Ultrasound Studies of Speech Production
|
Organizers:
Presenters:
|
Scott Moisik University of Victoria, Canada,
John H. Esling University of Victoria, Canada,
Sonya Bird University of Victoria, Canada,
Hua Lin University of Victoria, Canada
'Evaluating laryngeal ultrasound to study larynx state and height'
Ya Li University of Victoria, Canada,
Sonya Bird University of Victoria, Canada
'Ultrasound study of getural timing in Mandarin vowel-nasal production'
Yu Chen University of Victoria, Canada,
Hua Lin University of Victoria, Canada
'Analysing tongue shape and movement in vowel production, using SS ANOVA in ultrasound imaging'
Thomas Magnuson University of Victoria, Canada,
Chris Coey University of Victoria, Canada
'Synchronizing video, ultrasound, and audio with a water baloon'
|
Summary Description:
|
Ultrasound has been increasingly used in speech research in recent years. At our Speech Research Lab, we have begun to pursue this flexible research technology. To date our work has focused on simplifying methods for synchronization and analysis, innovating new methods for data capture, and expanding the scope of research from lingual ultrasound to laryngeal ultrasound and from gestural coarticulation to gestural timing.
The purpose of this special session is to share with other ultrasound researchers the insights we have gained to this stage of the lab’s development and to provide a forum at this year’s ICPhS where phoneticians from all sub-disciplines can come together and discuss the challenges and possibilities of ultrasound.
The following is a description of our current projects to be presented at the special session:
1. Synchronizing video, ultrasound, and audio with a water balloon (Thomas Magnuson and Chris Coey)
The problem of synchronization makes using ultrasound for phonetics research difficult, particularly with methods that incorporate video data of external reference points such as Palatron (Mielke et al., 2005). The addition of the third stream of data represents a significant challenge. This paper proposes using a water balloon for a common reference event for these tri-modal data. ‘Snapping’ the knot on the balloon creates a visible deformation in the video and ultrasound data at the same time as it produces a transient acoustic signature: it acts as an ultrasound clacker-board. This paper demonstrates this technique at 30 and 60 f.p.s. and discusses some of the technological barriers to audio/video synchronization at higher frame rates.
2. Evaluating laryngeal ultrasound for studying larynx state and height (Scott Moisik, John Esling, Sonya Bird, and Hua Lin)
We use ultrasound to examine the larynx and demonstrate what structures can be imaged using ultrasound. We claim that it is possible to capture some changes in laryngeal state, such as ventricular incursion during glottal stop production. Then we present a method for studying vertical laryngeal kinematics using optical flow analysis. To demonstrate the robustness of this approach, we provide two methods of validation. Independent video data of a control object in motion is analyzed with our optical flow algorithm to test its accuracy. We then present a pilot study of phonetic productions of Mandarin tones to demonstrate how the technique can be used to quantify larynx height during speech production. To further validate the quantitative results, we perform laryngoscopy simultaneously with the laryngeal ultrasound and compare the visual impression of larynx height from the laryngoscopy with quantitative measurements of larynx height. We conclude that laryngeal ultrasound is a viable technique to measure larynx height which has many advantages over other techniques, such as ease, non-invasiveness, and subject applicability.
3. Analysing tongue shape and movement in vowel production, using SS ANOVA in ultrasound imaging (Yu Chen and Hua Lin)
Using the General Smoothing Splines (Gu, 2011) and related software packages for the R environment, we developed a code for simultaneously analyzing tongue shapes for three or more phonemes with SS ANOVA. We applied this technique to the analysis of ultrasound images of three Mandarin vowels (/a/, /i/ and /u/) produced by two native speakers. Our results show that there is a tendency to move the tongue less in producing /i/ than /a/, which in turn had less tongue movement than /u/. Also, for /a/, the tongue moved downwards without evidence of any dorsal rising. For /i/, the anterior part of the dorsum raised upwards while the radix retracted. For /u/, the posterior part of the dorsum raised and the apex descended relative to the tongue’s neutral position. The results also show that the articulatory spaces of these vowels were roughly aligned with their respective formant patterns, except for one speaker, whose results for /i/ and /u/ do not support the theory that the higher the tongue, the higher the F1.
4. Ultrasound study of gestural timing in Mandarin vowel-nasal production (Ya Li and Sonya Bird)
The present study uses ultrasound in combination with acoustic analysis to examine how gestures are temporally organized in Mandarin vowel-nasal production. Specifically, it examines how a vowel gesture or the oral closure gesture of a nasal is timed in relation to acoustic syllable duration. The results show that gestural timing is relevant to both tongue and syllable positions; for example, an anterior gesture tends to occur closer to syllable peripheries than an open one and towards which periphery the anterior gesture occurs is determined by its position in a syllable. While most timing patterns such as the one in the above example can be explained by the biomechanically based model of gestural organization, namely the Jaw Cycle Hypothesis (Redford, 1999), others seem to be perceptually linked.
|
|
|
|
|
Title:
|
Prosodic Focus: Cross-Linguistic Distribution and Historical Origin
|
Organizers:
Presenters:
|
Bei Wang, Ling Wang, and Tursun Qadir Minzu University of China, Beijing, China
'Prosodic realization of focus in six languages/dialects in China'
Wing Li Wu University College London,
Lisa Chung University College London
'Post-Focus Compression in English-Cantonese Bilingual Speakers'
Canan Ipek University of Southern California
'Phonetic Realization of Focus with no On-Focus Pitch Range Expansion in Turkish'
Yi Xu University College London
'Post-focus compression: Cross-linguistic distribution and historical origin'
|
Summary Description:
|
Prosodic focus is widely recognized as one of the essential components of speech prosody, and it has been a major topic of prosody research for several decades. Much of the existing research on focus has been mainly concerned with the prosody of the focused component, paying little attention to components outside of focus. There is increasing evidence, however, that the prosody of components after focus is just as important. In many languages, the pitch range and amplitude of all post-focus components are reduced, a phenomenon referred to as post-focus compression (PFC), which is also previously known as post-nuclear tail or post-nuclear deaccenting. Evidence of PFC has been found in many languages, including Dutch, Greek, English, Swedish, Arabic, Mandarin, Finnish, Korean, Japanese, Turkish, Uyghur, Hindi and Persian, etc. However, there is recent evidence that PFC is not universal, because it is absent in many other languages, including Yucatec Maya, Chichewa, Chitumbuka, Durban Zulu, Hausa, Yi, Deang, Buli, Northern Sotho, Wolo, Taiwanese and Cantonese.
The differential distribution of PFC among the world languages may raise interesting questions. First, are the distribution patterns related to language specific features such as lexical tone, lexical stress or availability of morpho-syntactic markers of focus? There is initial evidence that at least some language specific features are unlikely to be the cause of PFC. For example, Taiwanese and Cantonese are both Chinese languages, just like Mandarin, and all Chinese languages are tonal. Thus lexical tone is not likely to be a determining factor for PFC. Also, morpho-syntactic means of marking focus are found in both PFC and non-PFC languages, which seems to reduce the likelihood of syntax or semantics as determiners of PFC. While complete exclusion of language specific features as determining factors will await further research, some alternative possibilities could also be considered. That is, is it possible that PFC in all the involved languages has a common origin? And, if yes, how could PFC get into so many different languages, through language contact or direct inheritance? These possibilities are highly hypothetical, but their potential implications are apparently profound.
The goal of this Special Session is to stimulate more research in this new and exciting area with four presentations that will address the following questions:
1. What is the distribution of PFC among the languages in China? Is there a tendency that languages or dialects there with a northern origin are more likely to have PFC? A positive answer to this question may support the suggestion that, contrary to the current consensus view, so-called Chinese languages actually have very different historical origins.
2. What is the distribution of PFC among the world languages? Does the distribution pattern show close links between PFC and language specific factors, thus supporting the independent genesis hypothesis? Or does it rather show that presence or absence of PFC in a language has little to do with language specific features, thus supporting the common origin hypothesis?
3. How easy is it for PFC to transmit from one language to another through contact in the form of bilingualism and second language acquisition? Ease of transmission would favor horizontal spread as a major mechanism of PFC distribution, whereas difficulty in cross-linguistic transmission of PFC may favor vertical descent as the only mechanism behind the distribution of PFC.
4. Methodologically, what are efficient ways of determining the presence/absence of PFC in a language? The development of effective methodology should make it easy to conduct large-scale survey of cross-linguistic distribution of PFC.
|
|
|
|
|
Title:
Organizers:
Presenters:
|
Part I
John Ohala University of California, Berkeley
'Christian Gottlieb Kratzenstein: Pioneer in speech synthesis'
Jürgen Trouvain Saarland University, Saarbrücken,
Fabian Brackhane Institut für deutsche Sprache (IDS), Mannheim, Germany
'Wolfgang von Kempelen’s ’Speaking Machine’ as an instrument for demonstration and research'
Hartmut Pfitzinger Christian-Albrechts-University, Kiel,
Oliver Niebuhr Christian-Albrechts-University, Kiel
'Historical development of phonetic vowel systems - the last 40 years'
Part II
Mária Gósy Hungarian Academy of Sciences, Budapest
'From stomatoscopy to BEA: the history of Hungarian experimental phonetics'
Rüdiger Hoffmann University of Technology, Dresden,
Dierter Mehnert University of Technology, Dresden,
Rolf Dietzel University of Technology, Dresden
'Measuring the accuracy of historic phonetic instruments'
Michael Ashby University College London
'Film from a phonetics laboratory of the 1920s'
|
Summary Description:
|
During the last decade, a growing interest in the history of phonetics can be observed not only among phoneticians but also in the community of engineers who are working in the field of acoustics and speech technology. Examples of this tendency include: the guide to the history of phonetics in the US, distributed at ICPhS San Francisco, 1999; the workshop “Un siècle de Phonétique expérimentale” in Grenoble, 2005; a continuous series of special sessions on the history of acoustics at the annual German acoustics conferences (DAGA), since 2006.
Rüdiger Hoffmann (Dresden) is engaged in the history of experimental phonetics and speech technology because it is hosting a historical acoustic-phonetic collection (HAPS). This collection includes the material heritage of the well-known pioneers of experimental phonetics, Franz Wethlo (Berlin) and Giulio Panconcelli-Calzia (Hamburg). Therefore the collection demonstrates the development of this field with a high degree of completeness. An overview of the collection was presented at the last ICPhS in Saarbrücken 2007, complemented by a small exhibition. Dresden University prepares a printed catalogue of the HAPS collection. For this purpose, Dieter Mehnert investigated and reconstructed the function and the way of application of numerous historical phonetic devices. Specifically, a selected number of measuring devices which were important for basic research with the speech signal (different resonators and transducers) was investigated with respect to their measuring accuracy. Results of this study will be presented.
In Budapest the permanent exhibition on "The history of experimental phonetics in Hungary" shows the long tradition of phonetics in this country of which Mária Gósy will report. Hungarian scholars were often among the first to do high quality research or to adopt the most advanced methods of their time. The first objective method in the investigation of the articulation of speech sounds called 'stomatoscopie' was used about 10 years earlier than Rousselot's famous work. More than 100 years ago various tools were developed for the measurement of air flow, lip movements (by means of 'phonoscope'), the amount of energy necessary during articulation. The 'voicing indicator' measured differences in voice onset time in stop consonants between Hungarian and German. Speech sound articulation by palatography, labiography and the X-ray technique followed some years later. The acoustic measurements of speech began at the end of the 19th century with the analysis of vowels. Oscillograms and the kymograph were used soon after that. At the beginning of the 20th century research on the artificial recognition of vowels started.
Michael Ashby from University College London presents a film from a phonetics laboratory of the 1920s. Silent 35mm movie footage has been discovered at UCL showing in detail the operation of a kymograph and of a sensitive flame. The experimenter who appears in the clips is almost certainly Stephen Jones, superintendent of the phonetics lab from its foundation in 1914. This paper documents the discovery and restoration of the film and analyses its content in relation to published work of the early department.
The Austro-Hungarian polymath Wolfgang von Kempelen (1734-1804) is without doubt one of the most influential thinkers of the production and acoustics of speech in the 19th and the beginning of the 20th century. His speaking machine can be regarded as the first working mechanical speech synthesis system in the world (an informal demo of a replica has been given at ICPhS in Saarbrücken). The presentation from Jürgen Trouvain (Saarbrücken) summarises the Kempelen research of the last years with a special focus on the relevance of phonetic research today.
A predecessor of Kempelen with regard to modelling vowels was Christian Gottlieb Kratzen¬stein (1723-1795), who won the prize of the St. Petersburg Academy of Sciences answering the question "Qualis sit natura et character sonorum litterarum vocalium a, e, i, o, u tam insigniter inter se diversorum". John Ohala (Berkeley) will present a review of his efforts which must constitute the first published account of mechanical synthesis (1781).
Hartmut Pfitzinger (Kiel) will review the historical development of phonetic vowel systems – from 1617 to 2005. In times where the vowel system of the International Phonetic Alphabet (IPA) is the de facto standard in phonetic sciences it is important to learn more about predecessors and alternative considerations of how vowel systems could be organised. This concerns matters of transcriptions, ideas on parameters of vowel production as well as the acoustics of vowel sounds and the relation between acoustics and perceived vowel quality. The concepts of such renowned historical personalities as Robinson, Wallis, Reyher, Hellwag, Chladni, Du Bois-Reymond, Willis, Wheatstone, Brücke, Helmholtz, Bell, Winteler, Sievers, Trautmann, Techmer, Vietor, Jespersen, Hermann, Lloyd, Scripture, Forchhammer, and Jones, will be reviewed and critically discussed.
With these six contributions phonetic research in a time span of four centuries will be covered. These historical considerations allow a re-view and a potential re-thinking of key concepts in the phonetic sciences which are relevant for inspection of phonetic data, the description of vowel structures and the development of speech technology.
|
|
|
|
