Acoustics of American English Speech: A Dynamic Approach

Rating:
Summary: Bell Lab's Recreation of Visible Speech
Review: This is a group effort from a trio of investigators at Bell Labs in Murray Hill, New Jersey, an institution whose researchers were among the creators of modern acoustic phonetics. It is almost a research monograph disguised as a textbook, published by a distinguished scientific publisher. This leads to frustration of the specialist reader at times, when one looks for technical summaries of the material which may be lacking. The book could be considered a complete revision of the classic Bell Labs book by Potter, Kopp, and Green (Kopp) under the title Visible Speech, published in 1947 and including with broader material on speech acoustics and its applications a 225-page description of the sounds of American English from an acoustic point of view.

The book is a systematic description of the phonetics of American English based on acoustic data from one male informant, a native of Pittsburgh, with limited data from a woman speaker. This is a weakness shared by most of the literature on speech acoustics; because the fundamental frequency of women's speech is higher than that of men, women's speech is technically more difficult to analyze precisely in acoustic terms. The limitation to one main subject is far outweighed by the precious systematic data which the book contains for most aspects of American English phonetics. Of course, it also suggests further investigation of the ways in which humans speaking somewhat different versions of the same language understand one another, which could be said to constitute the other, perceptual, side of the subject matter of this work. This aspect, by the way, is well reviewed in the excellent chapters on perception in J.M.Picket's revised book The Acoustics of Speech Communication (Allyn and Bacon, 1999).

The Bell authors provide valuable introductory chapters on basic concepts of phonetics and phonology, and speech acoustics. There is a growing literature on the basic concepts of acoustic phonetics, such as the new editions of the well-established books by Ladefoged (1996) and Kent and Read (2001), which some readers may wish to consult. Then follows a chapter on the static properties of speech sounds; many treatments of acoustic phonetics never get past this elementary level. However, the book by Olive, Greenwood, and Coleman makes a precious contribution in the systematic following chapters on the treatment of English sounds in context. Many individual aspects of this topic are well treated in the research literature, but
this comprehensive investigation of the speech of a single speaker is unique. There is broad treatment not only of formant transitions between consonants and vowels, but also of consonant clusters and interactions, contextual variants of sounds such as [l] and [r](using standard notation for phonetic units), and a long final chapter on acoustic variability of sounds and dialectal variability. A precious feature of the index is that it not only includes a detailed phonetic and technical listing but also gives the location in the text of treatments of individual sound sequences, shortening many searches.

Apart from its merits as a research monograph, this work
also works well as a classroom text. The reviewer has used it in a class on applied English phonetics with extensive spectrographic lab experience, in which the students studied their own speech acoustically and compared it to the results described by Olive et al. A particularly valuable role of the book is for non-English speakers who are able to use the objective character of the acoustic analysis to bypass the obstacle of uncertain perception of English sounds and assist themselves in improving their mastery of English phonetics. Elementary speech acoustics on this level is not difficult, and this aspect of the work was very much appreciated by foreign students in my class.

This is a work of enormous value to a variety of students and specialists, particularly in phonetics, linguistics, and speech pathology, but also for engineers and the increasing ranks of those working on speech recognition. The reviewer finds that he has had to gradually increase the number of loaner copies of this work on his office shelf, because the Olive book is not in the local library, and his students want prolonged access to a work which describes English acoustic phonetics competently and(within its limits)comprehensively. This book is is not as well known as it deserves to be.

The reviewer has some complaints. In general, the articulatory discussions are not as accurate or precise as the acoustic ones; of course, that is not what the book is about. If one must mention published collections of papers on speech acoustics, the indispensable book Readings in Acoustic Phonetics edited by Lehiste (1967 with several reprints) which preserves publication format is more useful than the one cited. It is regrettable and puzzling that there is not a single first-author work of Gordon Peterson in the bibliography. These are minor details. This work deserves to be well known, and to be in all major university libraries and institutions in which linguistic or clinical phonetics has significance, as well as in scholars' libraries. It should be read carefully and repeatedly until its pages are dirty and scribbled on, like my copy. The first acoustic specification of sounds of a language known to the reviewer was the publication of formant frequencies of whispered vowels published by Samuel Reyher in 1679. The book by Olive and his colleagues reflects a long and proud past, and its subject matter seems at present to be enjoying a deserved revival of interest.