Applying Phonetics. Murray J. Munro
the reverse. Nor is it true that printed texts are good models of how we should speak. It is true, however, that written style in many languages differs from spoken style. The formal usage in English academic textbooks is quite different from everyday spoken English, while English speech and writing are much more similar in other situations such as text messaging. For that reason, we cannot regard writing and speaking as different manifestations of the same thing. Sometimes they are very close; other times they are not.
1.3.1 branches of phonetics
Despite its concentration on sound, phonetics is a diverse field with a multidisciplinary reach—so much so that you will have to read this entire book to get a good picture of its scope. To begin, let's consider its three core branches, as shown in Figure 1.3.
Figure 1.3 Core branches of phonetics
Articulatory phonetics covers the anatomy and physiology of speech, with a focus on the structures we use to generate vocal sounds, such as the tongue, lips, nose, and larynx, and on the ways in which these anatomical components function together. An understanding of normal articulation is essential for helping people with speech disorders, who may need the assistance of a speech–language pathologist. A child who has undergone surgery for a cleft palate, for instance, may experience production difficulties that can be remedied by a speech professional. The domain of auditory phonetics is the structures and processes through which the human auditory system decodes the speech stream into meaningful messages. Numerous insights into child language acquisition have been gained because of perceptual research within this branch of the field. Finally, acoustic phonetics addresses the physical properties of the speech sounds themselves, often through analyses like the ones depicted in Figure 1.2. Thanks to acoustic‐phonetic research, we are able to synthesize the intelligible, natural‐sounding speech now available on computers, phones, and other household devices.
It will help you to remember the core areas of phonetics if you think of them as the three As: Articulatory, Auditory, and Acoustic.
Though the core branches provide us with one way of appreciating the nature of the field, another set of descriptors can be used to characterize phoneticians' approaches to their work. Historically, phonetics has relied extensively on the trained human ear, and even today, careful listening and skillful transcription are fundamental to IMPRESSIONISTIC phonetics. Some highly skilled phoneticians have played a key role in criminal cases by providing ear‐based analyses of threatening phone calls. However, thanks to the technological advances of the twentieth century, INSTRUMENTAL phonetics has taken a more prominent role than ever before. It involves the use of a variety of sophisticated tools for imaging the vocal tract during speech and for pinpointing important acoustic details. A third approach, the most recently developed, which we will refer to as AUTOMATIC SPEECH PROCESSING, uses artificial intelligence ( AI ) for a variety of purposes, including computer recognition of speech, forensic voice identification, and speech synthesis.
Finally, APPLIED phonetics, which is heavily emphasized throughout this book, is concerned with the ways in which our understanding of speech can be used to achieve practical ends. Strictly speaking, applied phonetics is not separate from any of the types we've mentioned already. In fact, it makes use of research findings from all three core branches and can be approached impressionistically, instrumentally, and through AI. One of the first applications of phonetics to come to mind for most people is language teaching. In fact, the speech sciences have been powerfully influenced by ideas about spoken language instruction, as we'll see in Chapter 11. However, there are many other sub‐branches of applied work, including forensic phonetics and clinical phonetics, as well as specialized areas of application relating to accessibility, automation, music, animation, stage and screen acting, and business.
Cat‐to‐Human: Feed Me!
Our family cat, Owen, happens to be extraordinarily vocal when he is hungry, needs to play, or spots a squirrel through the window. The left panel of Figure 1.4 shows an acoustic representation of one of his productions, which you can compare with the much more complex, human‐produced word on the right. While it's tempting to attribute specific meanings to the sounds he makes, the available evidence does not allow researchers to say what exactly cats “mean” when they vocalize. Nonetheless, a wealth of studies of domestic feline behavior have helped shed light on the problem. It is well established, for instance, that cats vocalize differently when they are around humans than they do in feral conditions. Perhaps the most famous study of feline communication is Mildred Moelk's “Vocalizing in the House‐Cat: A Phonetic and Functional Study,” published in 1944 in the American Journal of Psychology. Moelk identifies three kinds of production: mouth closed, mouth gradually closing, and mouth tensely open. The resulting sounds fall into 16 different patterns, for which she even provides transcriptions using the International Phonetic Alphabet. According to Moelk, the patterns correspond to (among other things) greetings, acknowledgments, demands, begging, and bewilderment.
Several perceptual experiments have shown that humans can accurately interpret cat vocalizations, at least to some extent. Schötz and van de Weijer (2014) collected audio recordings of cats in two contexts: at feeding time and while waiting at the veterinarian. They played the meows in random order to 30 human listeners tasked with guessing the context of each token. Not only did the listeners perform correctly at above chance levels, but those having experience with cats as pets scored higher than those without. So, it seems that humans can learn to partially interpret feline messages through experience. In another study by McComb, Taylor, Wilson, and Charlton (2009), humans judged solicitation purrs (recorded during food‐seeking) as more urgent and less pleasant than non‐solicitation purrs played at equal volume. The listeners' ability to make the distinction, which occurred with or without cat experience, was due to a particular high‐frequency component within the solicitation purrs that has a surprising parallel in human infant cries. In the authors' words, the less pleasant purrs “may be exploiting an inherent mammalian sensitivity to acoustic cues relevant in the context of nurturing offspring” (p. R507). If so, then when Owen howls from the kitchen, it's not simply because he has learned what gets our attention. Rather, his biology may be calling out to our own!
Figure 1.4 Waveforms and spectrographic representations of a cat's meow (top) and a human production of the word “mouse” (bottom)
for further thought, analysis, and discussion
1 Find an audio or video recording of a politician, entertainment figure, or sports personality responding extemporaneously (without preparation) to questions during an interview. Write out several of the speaker's utterances, including any hesitation forms like “um” or “ah” and any speech errors. Identify some of the ways in which the speaker's productions differ from written language.
2 With a classmate, discuss how many phones you think occur in each of the words below. Suggested answers are provided on the APSSEL website.a) mape) speechi) cheeseb) shotf) schoolj) tickc) fixg) waitedk) oftend) wheath) marchedl) epitome
3 Record yourself producing the words in item 2 using Praat, and examine the waveforms and spectrograms. (Consult the APSSEL website for basic instructions.) First, consider the appearance of the