Page 27 - Summer 2021

We all know that “the ear is too easily fooled” (Baken and Orlikoff, 2000).
Audio Recording
Rousselot expressed similar misgivings about IPA transcrip- tion some 114 years ago. In his obituary, Scripture (1925) wrote the following, “Coming to Paris in 1880, Rousselot was initiated into questions of Romance linguistics by Gaston Paris... Finding it impossible to detect the finer details of spoken sounds by ear, he complained to Gaston Paris, who said, ‘Only mechanical registration will give you accurate knowledge. Attempts have been made in Marey’s laboratory. Go and see!’” This exchange led Rousselot to invent the kymograph machine to record and account for speech sounds as accurately as possible (Figure 5).
Rousselot’s kymograph machine set the stage for the inven- tion of other audio recording devices. The kymograph drew graphs in the frequency and intensity domains on the basis of the articulatory effort exerted in producing specific sounds. The first major breakthrough came in 1935 with the invention of the magnetic reel-to-reel tape recorder. The next seminal moment came in 1963 when Philips introduced the portable cassette player. The adjec- tive “portable” must be taken under advisement because these early audio recorders were bulky and heavy. Ian Maddieson tells of his incredible phonetic field record- ing adventures in Africa. Their audio recorders and their accessories were so heavy that he and Peter Ladefoged hired porters to carry their equipment as they trudged from place to place. Ladefoged and Maddieson made good use of their equipment for language documentation. The University of California, Los Angeles (UCLA) Phonetic Lab Archive (1996; available at bit.ly/30fe3aJ) and their book The Sounds of the World’s Language (Ladefoged and Maddieson, 1996) are the fruit of their dedication to doc- umenting phonetic diversity. Ladefoged (1968) was first among linguists to have used this method for document- ing 61 of the 1,436 Niger-Congo languages spoken in West Africa. Increasingly, documentary linguists are making use of portable camera recorders to capture aspects of everyday life and festivals. Some have produced excellent documentaries about endangered languages using digital audiovisual devices (available at bit.ly/2NYdu2t).
Visualization Technologies
Linguists have used inventiveness and creativity in adapting medical devices for language documentation.
Ladefoged (2003) displays many “mom-and-pop” palato- graphs (devices to capture imprints of where articulators come into contact to produce sounds), airflow equipment, and a laryngoscope machine. Nowadays, researchers are using very sophisticated tools such as portable ultrasound and electroglottography (Whalen et al., 2011). Butcher (2013), Esling (2013), and Tabain (2013) provide excel- lent descriptions of various technologies used currently in language documentation and acoustic phonetic fieldwork.
Sound Spectrography
The sound spectrograph machine was invented at Bell Laboratories in the 1950s. Unfortunately, its prohibitive cost prevented its widespread usage. Fromkin (1985) indicates that in 1952, for the whole of England, there was only one sound spectrograph, which was owned by David Abercombie. When Fry (1955) needed to conduct his groundbreaking experiment on the acoustic correlates of lexical stress in English, he had to travel all the way from England to the Bell Laboratories in New Jersey to conduct his experiments. Fortunately, in 1995, Paul Boersma and David Weenink (2020) developed Praat, a free software platform for online speech analysis. It has become so pop- ular that Watt (2013) refers to it as “the industry standard for acoustic analysis of speech.” Ladefoged (2003) likens
Figure 5. Abbé Rousselot and his kymograph machine. Available at bit.ly/3acqZ5K, license by Creative Commons.
Summer 2021 • Acoustics Today 27