Invited speakers at SST-2002.
Opening session:
How and why languages change:
Associate
Professor Kate Burridge
"In
Forme of Speche is Chaunge"
Human processing
of spoken languages theme:
Professor
Willem Levelt
Multiple
lexical access in speech production
Professional applications
of speech science theme:
Professor
Max Coltheart
Modular modelling of language-processing and
its applications to the understanding, assessment and treatment of disorders
of language
Machine Processing
of spoken language theme:
Richard
Cox
Multimedia
Textbooks and Beyond: Speech Processing for People with Disabilities
Kate
Burridge
 |
Kate
Burridge completed her undergraduate training in Linguistics and German
at the University of Western Australia. This was followed by three years
postgraduate study at the University of London. Kate completed her PhD
in 1983 on syntactic change in medieval Dutch. She also taught at the
Polytechnic of Central London before joining the Department of Linguistics
at la Trobe University in 1984, where she is now an Associate Professor.
Her main areas of research are: grammatical change in Germanic languages;
the Pennsylvania German spoken by Anabaptist communities in Canada; the
notion of linguistic taboo; the structure and history of English. Kate
is a member of the Australian Academy of Humanities. She is also
a regular presenter of language segments on ABC Radio. |
Kate’s
published books include: “A Localized
Study of Pennsylvania German Dialect in Waterloo County, Ontario” (1989, The Pennsylvania Folklore Society of
Ontario); “Euphemism and Dysphemism: Language Used as Shield and Weapon” (1991,
Oxford Uni Press) — co-authored with Keith Allan; “Diachronic Studies on the Languages of the Anabaptists” (1992,
Universitätsverlag Brockmeyer) — co-edited with Werner Enninger
“Syntactic Change in Germanic” (1993, John
Benjamins); “Canada — Australia: Towards
a Centenary of Partnership” (1997,
Carlton Uni Press) — co-edited with Lois Foster & Gerry Turcotte; “English in Australia and New Zealand — An
Introduction to its Structure, History and Use” (1998, Oxford Uni Press) —
co-authored with Jean Mulder; “Introducing
English Grammar” (2001, Edward Arnold) — co-authored with Kersti Börjars.
"In
Forme of Speche is Chaunge"
Kate Burridge,
La Trobe University
"Ye know ek that in forme of speche is chaunge / Withinne a thousand yeer,
and wordes tho / That hadden pris, now wonder nyce and straunge / Us thinketh
hem, and yet thei spake hem so" [Geoffrey Chaucer, c. 1385, Troilus and Criseyde,
11, 22]
"Copious without order, and energetick without rules [...] perplexity to be
disentangled, and confusion to be regulated" - this is how Samuel Johnson
described the English language in the preface to his dictionary of 1755. Like
others of his time, Johnson couldn't help but contrast English with the classical
languages Latin and Greek. These were written varieties, of course, which
meant they lacked the natural flux and variance of any normal spoken language.
Against these ossified paragons of linguistic virtue, it is not surprising
that English appeared unregulated and unrefined.
Since Johnson's time, a standard language has stabilised. English has now
acquired its army and navy (to adapt a description attributed to Max Weinreich);
in other words, the arsenal of prescriptive texts like dictionaries, grammars
and style guides that give standard languages like English so much of their
clout. Once again we are confronted with neat lists and fine-spun paradigms
that present a regularity and homogeneity that is a total fiction. Even the
labels we use like "standard" and "non-standard English" imply that we are
dealing with a clear-cut distinction. Such labels obscure a reality of variation
and fuzziness that is an inherent part of any language system.
Variability is what provides the vehicle
for change. The infinite variation in everyday speech - the so-called slipshod
pronunciations, mistakes in grammar, coinages, new-fangled meanings and so
on - these are the basis for real change. The majority of them will drop by
the wayside, but some will catch on, be used more and more and will eventually
form part of the repertoire of Standard English in the future. Most of the
time we fail to see the stirrings going on around us. Changes in language
are as gradual and imperceptible as the changes in the growth of a plant or
tree. But the clues to where our language is heading are everywhere.
All aspects of the system - sounds,
words, grammar - are constantly on the move. Most of the changes result from
a complex network of different influences. One of the most important of these
is repetition or ritualisation and part of this talk will explore the many
different ways change is shaped by the frequency with which we use words and
phrases. The talk will also focus on the way in which language evolves. Often
changes introduce complexity and anomaly elsewhere. Typically they take a
long time to complete - in fact many never come to completion. Competing changes
can interfere. Other changes simply peter out. They might even reverse themselves.
Frequently they leave behind relic forms. These become the eccentricities
of the language - the forms, functions and meanings that are out of alignment
with the usual patterns of behaviour in the language. Modern English is full
of rubble, and hidden in all this debris it is possible to uncover a partial
history of the language.
Willem
J.M. Levelt (Pim)
Presentation Title: "Multiple lexical access in speech
production"
 |
Pim
Levelt studied psychology
at Leyden University (including six months experimental work under Albert
Michotte's direction at Louvain University), completing his Ph.D. at Leyden
University (cum laude), 1965. During his career, has held positions at
Harvard University, University of Illinois, Groningen University, Louvain
University, Princeton, Nijmegen University with Max Planck Society involved
in Psycholinguistics and psychophysics. Pim has been involved with many
journals including The Linguistic Review, Language and Cognitive Processes,
Mind and Language, Journal of Semantics (-1986), International Journal
of Sign Language, Spatial Cognition and Computation and as editor of the
Springer Series in Language and Communication. (1978-1990) as well as
Language Sektion Editor of The new cognitive neurosciences, M.S. Gazzaniga
(ed.). MIT Press, 2000. Currently Pim is the President of the Royal Netherlands
Academy of Sciences. |
Pim's
Academy memberships
-
Member
of the Royal Dutch Academy of Sciences
-
Member of the
Holland Society of Sciences
-
Member of the
Academia Europaea
-
Member of the
Deutsche Akademie der Naturforscher Leopoldina
-
Member of the
Royal Flemish Academy of Belgium for Sciences and Arts
-
Member of the
Bayerische Akademie der Wissenschaften
-
Foreign Associate
of the National Academy of Sciences (USA)
Pim has also
been the receiptent of many awards and honours
-
Radboud
Stichting award for junior scientists (1968)
-
Hendrik Muller
award for the Social and Behavioral Sciences of the Royal Dutch Academy
of Sciences (1993)
-
Heymans award
of the Dutch Psychology Association (NIP) (1996)
-
Honorary member
of the Linguistic Society of America (1998)
-
International
Fellow of the American Psychological Society Ridder in de Orde van de Nederlandse
Leeuw (1998)
-
Doctorate honoris
causa, Maastricht University (2000)
Published
books include:
- Levelt,
W.J.M. (1965). On binocular rivalry. Assen: Van Gorcum. Second printing: The
Hague: Mouton (1968).
- Flores
d' Arcais, G.B. & Levelt, W.J.M. (Eds.). (1970). Advances in psycholinguistics.
Amsterdam: North Holland.
- Levelt,
W.J.M. (1974). Formal grammars in linguistics and psycholinguistics. The Hague:
Mouton. Vol. I, An introduction to the theory of formal languages and automata;
Vol. II, Applications in linguistic theory; Vol III, Psycholinguistic applications.
- Dutch
edition: Formele grammatica's in linguistiek en taalpsychologie. Deventer:
Van Loghem Slaterus (1972).
- Levelt,
W.J.M. & Flores d' Arcais, G.B. (Eds.). (1978). Studies in the perception
of language. New York: Wiley.
- Sinclair,
A., Jarvella, R., & Levelt, W.J.M. (Eds.). (1978). The child's conception
of language. Heidelberg: Springer.
- Klein,
W., & Levelt, W.J.M. (Eds.). (1981). Crossing the boundaries in linguistics.
Studies presented to Manfred Bierwisch. Dordrecht: Reidel. Levelt, W.J.M.,
Mills, A., & Karmiloff-Smith, A. (1981). Child language research in ESF Countries.
An inventory. Strasbough: European Science Foundation.
- Levelt,
W.J.M. (1989). Speaking: From intention to articulation. Cambridge, MA: The
MIT Press.
- Levelt,
W.J.M. (Ed.). (1993). Lexical access in speech production. Cambridge: Blackwell.
- Levelt,
W.J.M. (Ed.). (1996). Advanced psycholinguistics. A Bressanone retrospective
for Giovanni B. Flores d'Arcais. Nijmegen:
Max Planck Institute.
MULTIPLE
LEXICAL ACCESS IN SPEECH PRODUCTION
Willem J.M. Levelt
Max Planck Institute for Psycholinguistics, Nijmegen, The Netherlands
ABSTRACT It is quite
normal for a fluent speaker to produce two to four words per second. The accessing
system is robust; lexical error rates are far below 1%. How is such fast, repeated
lexical access organized? I will first present a bird eye's view of the architecture
of single-word lexical access, as emerging from speech latency studies, such
as picture/word interference and implicit priming experiments. Preparing a content
word involves a system consisting of two major components: the first one, lexical
selection, is conceptually driven and produces a syntactically specified lexical
item as output. The second component, form encoding, retrieves the item's phonological
code, incrementally generates the item's syllabification in context and computes
the corresponding articulatory gestures. These gestures are executed during
articulation.
Starting from this
architecture, I will begin discussing the process of multiple access from the
analysis of simple two-content word utterances such as scooter and camel. I
will contrast three possible planning structures: level-by-level, incremental,
and unit-by-unit. Level-by-level: first do lexical selection for the whole utterance,
then form encoding for the whole utterance, then initiate articulation. Incremental:
do form encoding of word-1 while selecting word-2, or articulate word-1 while
doing the form encoding of word-2, etc. Unit-by-unit: First go through all encoding
steps for word-1, then through all encoding steps for word-2 (etc.). These procedures
make different predictions with respect to the fluency of the resulting utterance,
and with respect to the effect of word-2 related distracter words on speech
latency. The level-by-level planning procedure, suggested in the early days
of modern psycholinguistics, is quite untenable and will not be further discussed.
Although most data support the incremental structure of utterance planning (introduced
a quarter century ago by Dennis Fry), a small adaptation of the unit-by-unit
planning structure generates exactly the same latency and fluency predictions.
A different type of experimental data is necessary to distinguish these latter
two procedures. Eye tracking registration during description of multiple-object
pictures can do the job. The tracking data support the unit-by-unit rather than
the incremental type of planning. Further theoretical analysis shows that unit-by-unit
planning, different from incremental planning, circumvents buffering. However,
not in all cases. When the articulation of word-1 is relatively short as compared
to the preparation of word-2, articulatory buffering cannot be evaded if within-utterance
fluency is to be preserved. Recent counterintuitive data by Griffin support
this analysis. In conclusion, multiple word planning strategies are shaped by
fluency requirements, by a general aversion against (articulatory) buffering,
and by the relative durations of articulation and preparation of successive
words.
Max Coltheart
 |
Professor
Max Coltheart completed his PhD at the University of Sydney in 1968.
He was awarded a Doctorate of Science by the Macquarie University in 2001
in the field of Cognitive Neuropsychology. He is one of only two
scientists who are Fellows of both the Academy of the Social Sciences
and the Academy of Science in Australia. Professor Coltheart was
one of the founding figures in the development of the discipline of cognitive
neuropsychology, co-authoring the most significant volumes which launched
the discipline and founded the leading journal Cognitive Neuropsychology
in 1984. |
Professor
Coltheart was awarded an ARC Federation Fellowship at Macquarie University
in 2001. His programme of research and research training will initially focus
on (a) how people understand and produce language, and how these capabilities
can break down after brain damage; (b) the relationship between visual memory
and reading; and (c) the nature of the persistent delusions seen in certain
mental disorders, and what causes these delusions.
Modular
modelling of language-processing and its applications to the understanding,
assessment and treatment of disorders of language
Max
Coltheart
ARC Federation Fellow
Scientific Director,
Macquarie Centre for Cognitive Science (MACCS)
Psycholinguistic
investigations of reading, writing, speaking and understanding speech have
led to the development of a highly modular model of the human anguage-processing
system, at least with respect to language-processing at the single-word level.
For the most part, the most incisive evidence assisting the development of
this view of the normal language-processing system has come from data from
people in whom brain damage has affected some aspects of language-processing
whilst sparing others.
Such
acquired disorders of language-processing can be highly specific, and there
is a bewildering array of them; no two patients with impaired language processing
after brain damage are the same. Nevertheless, one can understand each patient's
disorder by investigating which modules of the language-processing system
have remained intact and which have been impaired by the brain damage. Test
batteries for assessing language impairments, such as PALPA (Kay, Lesser &
Coltheart, 1992) have thus been derived directly from the theoretical model
of language-processing. Model-based assessment of this kind allows the devising
of focussed treatment programs, and a number of studies have demonstrated
the success of such treatments in improving the language abilities of such
patients.
In
this talk I will describe how this model was developed over the past century
or so, particularly with reference to data from people with acquired disorders
of language; how it is used for assessment; and how treatment programs developed
from such model-based are devised, executed and evaluated
References:
Kay,
J., Lesser, R. and Coltheart, M. (1992) PALPA: Psycholinguistic Assessments
of Language Processing in Aphasia. Hove: Lawrence Erlbaum Associates. Spanish
translation by Cuetos, Kay, Lesser, Coltheart & Valle published 1995. Dutch
translation by Bastiaanse, Bosje, Kay, Lesser, Coltheart and Visch-Brink,
published 1995. Hebrew adaptation by Kay, Lesser, Coltheart, Gil & Edelman
published 1999.
Coltheart,
M. (2001). Assumptions and methods in cognitive neuropsychology. In Rapp,
B. (Ed.) Handbook of Cognitive Neuropsychology. New York: Psychology Press
Coltheart,
M. (2002). Cognitive Neuropsychology. In Wixted, J. (Ed.) Stevens' Handbook
of Experimental Psychology, Third Edition - Volume 4: Methodology. John Wiley
& Sons, pp 139-174.
Richard
Cox
 |
Richard
V. Cox received his Ph.D. in Electrical Engineering from Princeton University.
In 1979 he joined the Acoustics Research Department of Bell Laboratories.
He has conducted research in the areas of speech coding, digital signal
processing, analog voice privacy, audio coding, real-time implementations,
speech recognition, and speech enhancement. He is well known for his work
in speech coding standards. He collaborated on the low-delay CELP algorithm
that became ITU-T Recommendation G.728 in 1992. He managed the ITU effort
that resulted in the creation of ITU-T Rec. G.723.1 in 1995. In 1987 he
was promoted to Supervisor of the Digital Principles Research Group. In
1992 he was appointed Department Head of the Speech Coding Research Department
of AT&T Bell Labs. In 1996 he joined AT&T Labs as Division Manager of
the Speech Processing Software and Technology Research Department. In
1999 he was awarded the AT&T Science and Technology Medal. In August 2000
he was appointed Speech and Image Processing Services Research Vice-President.
In February 2002 as part of a reorganization, he was appointed Voice Enabled
Services Research Vice-President. |
Dr. Cox is a Fellow of the IEEE and is currently
President of the IEEE Signal Processing Society. He is also a member of the
Board of Directors of Recording for the Blind & Dyslexic, the only US provider
of textbooks and reference books for people with print disabilities. At RFB&D
he is presently helping to lead the effort to develop digital books combining
audio, text, images, and graphics for their consumers. These 'multimedia books'
will become generally available in September 2002 for RFB&D K-14 students
throughout the US.
Multimedia
Textbooks and Beyond: Speech Processing for People with Disabilities
Richard V. Cox
AT&T Labs-Research and Recording for the Blind & Dyslexic
In
1995, Recording for the Blind & Dyslexic began their Digital Audio Book Project.
RFB&D is the United States' only federally chartered provider of recorded
textbooks and reference materials for people who cannot read standard text
due to a visual or perceptual disability. This is an attempt to combine recorded
speech, text, graphics, and images on a CDROM that can be accessed using standard
technologies. The project was originally intended to provide people with print
disabilities an alternative that is functionally equivalent to conventional
books. Instead of an audio tape player and multiple cassettes, a PC and one
or two CDROMs can be used for each book. There are also specialized players
now available. In the future, it may be possible to use commercially available
MP3 players. The first prototype was demonstrated in 1996. In 1997 RFB&D took
the lead in a worldwide effort, known as DAISY, to define the format and the
software needed to author, manufacture, and access such books. The project
attracted wide attention because it sets the foundation for creating multimedia
textbooks. The possibilities for how books, magazines and journals can evolve
are greatly expanded. In the fall of 2001, RFB&D began pilot testing in selected
schools throughout the United States. Beginning with the 2002-2003 school
year, the digital books will be generally available to RFB&D's registered
members. In this talk, the project will be reviewed and the functionality
of the books will be demonstrated.
This
is one successful example of combining speech processing technologies to produce
an aid for a class of people with disabilities. People with communication
disabilities (hearing, speech, vision, dyslexia) represent a substantial economic
market segment. With the advent of more powerful speech technologies, many
more new products and service opportunities are possible. For every opportunity
there are multiple new technical challenges to overcome. The goal is to inspire
many others to take up these opportunities and challenges.
Conference
Secretariat
Bronwen Hewitt
Conference Management
The University of Melbourne
Victoria, Australia, 3010
Phone: +61-3-8344-6389
Facsimile: +61-3-8344-6122
bhewitt@unimelb.edu.au