Phonetic Data Analysis: An Introduction to Fieldwork and Instrumental Techniques (review)

‘Old people can be hard to control.’

(Phonetic data analysis, p. 14)

Phonetic data analysis (PDA) is a course book for instrumental or field phonetics courses and, like Peter Ladefoged’s other major textbook A course in phonetics (2006), defines not only (or primarily) the content of the course but also a set of attitudes for doing successful phonetic research. This book is important because L, more than any other person in the twentieth century, defined the linguistics/phonetics curriculum and a large swath of the phonetics research agenda in linguistics.

L’s choices for the content of PDA were guided by his view of the aims of phonetic fieldwork. For him the chief aim of phonetic fieldwork is language description at a segmental phonetic level, with particular focus on unusual or rare sounds that challenge one’s view of what sounds are possible in language. This aim leads to a focus on discovering what exactly the speakers of a language do with their mouths, vocal folds, and lungs in order to pronounce the words of their language. As he discusses in this book, the instrumental tools that are necessary to achieve this aim are primarily based in acoustic speech analysis (spectrograms especially) with some supplemental physiological techniques for observing consonant place of articulation and consonant aerodynamics.

Thus, PDA is a practical book about how to discover patterns of consonant and vowel pronunciation using simple phonetic instruments that can be packed in a suitcase. This is the second of the widely imitated three-course phonetics curriculum that L helped to establish in the UCLA Department of Linguistics. The first is an introductory course, for which he wrote A course in phonetics, and the third is a course on ‘Phonetic theory’. I mention this curriculum because I think that it is useful, in considering the range and limitations of PDA, to keep in mind the pedagogical context in which it was developed—particularly that phonetic theory is not missing in the plan. But if your aims for crosslinguistic phonetic instrumental fieldwork are a little more oriented toward explanatory phonetics as grounding for phonological patterns, or toward cognitive theories of speech motor control or speech perception, you would write a different book. In this book L focuses on speech-sound description, introduces the tools of phonetic study with enough detail and clarity to get the student started, and vividly exemplifies his uniquely successful attitudes toward phonetic research.

Though two of the chapters in PDA are about ‘exotic’ phonetic instrumentation (palatography and aerodynamics), the tools of the trade illustrated in this book are mainly acoustic. Ch. 1 is a very detailed and helpful overview on how to conduct phonetic research. L discusses such seemingly mundane details as how often to take breaks, how to get a consultant back on task after some chitchat, and how much to pay speakers. He also emphasizes the importance of the word list in his style of phonetic research—with recollections of some of his favorite word lists. The discussion in this chapter on audio-recording techniques is also especially valuable.

The acoustic speech analysis chapters are then on ‘Pitch, loudness, and length’ (Ch. 4), ‘Characterizing vowels’ (Ch. 5), ‘Acoustic analysis of consonants’ (Ch. 6), and ‘Acoustic analysis of phonation types’ (Ch. 7). These are wonderful chapters full of tips and tricks for getting spectrograms and waveforms to give up their secrets. They are very detailed, providing an outline of his analysis strategy in particular cases rather than a cookbook of measurement techniques for every situation. One of the clearest messages of the chapters is pragmatism. For instance, the acoustic theory of speech production teaches that consonant place of articulation can be discerned from speech spectrograms—and clearly listeners get some consonant place information from the speech acoustic signal. L states though, ‘Acoustic phonetic analysis is not the best way to find out about different places of articulation. You can do much better with the simple...