Page 71 - Spring 2015

Comparison between languages
Direct comparisons of different languages or dialects is one of the largest research areas these days. Frequently, this re- search compares similar sounds, such as a class of vowels or consonants, across two or more languages. This work often demonstrates surprising differences between the phonetics of the two languages. These results keep expanding our view of the actual space of phonetic options available to human speakers.
new Methods in study of
Dialect Variation and Change
The recent development of large-scale corpora of recorded speech is enabling many new research methods. Indeed, many new questions can be addressed and explored regard- ing regional and social dialects and stylistic effects, such as in American English, and for tracing the evolution of pho- netic changes in the speech of individuals over their life and of communities over generations.
One recent study of American English vowels debunked the common implicit claim that it makes sense to speak of the “general American dialect” (Clopper et al., 2004). The authors showed that there are many region-specific ways of pronouncing English vowels but nothing that could be called a dialect-free general American dialect.
Gender Differences in speech
Another aspect of speech variation coming under increas- ing scrutiny is a difference in speech details traceable to the speaker’s gender or sexual orientation.
Phonetics and Phonology of little-studied and endangered languages
Research on phonetics and phonology has studied only a tiny fraction of the 7,000 or so languages spoken around the world. For some years now, some researchers from the Acoustical Society of America have been attempting to doc- ument and study as many of these languages as possible. In addition, they seek to at least describe the major outlines of the phonetic and phonological characteristics of these lan- guages.
One interesting study in a recent issue of JASA (Heering and van Heuven, 2012) explored how listeners are able to deter- mine the intended intonation of Dutch sentences where the intonation, for example, marks a question versus a statement when the speech is whispered. It turns out that Dutch lis-
teners can do this reasonably well and that speakers modify properties like the formant values of vowels to signal rising pitch to help listeners understand.
speech Production
Speech production involves very complex and rapid move- ments which place serious demands on the motor system from the cerebellum to the cortex to the peripheral muscu- lar system. Some members of the Speech Communication group of the ASA work on these issues.
Larynx and Glottis
The vocal folds are exquisitely controlled to change vocal pitch, loudness, and voice quality as well as to serve respira- tory functions for both speech and song. Furthermore, as humans mature and age, the behavior of the glottis changes. For example, male voices tend to rise in pitch relative to their voice when young. Technical research techniques such as electroglottography and high-speed laryngeal videoendos- copy as well as basic methods of acoustic analysis are used to account for the details of glottal motion during various kinds of speech.
Tongue Movements
The more gross movements of the vocal tract, that is, those by main articulators like the tongue, lips and velum, are also investigated in detail. These gestures are known to exhibit temporal overlap (often called coarticulation) and thus ex- hibit great complexity depending on contextual detail. They also exhibit subtle timing constraints. Research on these issues frequently involves neuroimaging techniques such as ultrasound and functional magnetic resonance imaging (fMRI). Even though one might think that these issues of motor control deal only tangentially with acoustics, given ASA audience interest in this research, this work is wel- comed at the ASA.
Another interesting recent paper (Ménard et al., 2013) com- pared articulatory and acoustic details of a group of con- genitally blind adult speakers of Canadian French with a control group of sighted speakers for the pronunciation of the full set of isolated Canadian French vowels. Using ultra- sound, video, and acoustic analysis, the authors found that the blind speakers exhibited somewhat smaller acoustic con- trasts between their vowels but large articulatory differences from the sighted speakers. For example, the blind speakers showed quite a bit less lip protrusion for the rounded vowels and correspondingly greater tongue movements.
| 69