1. Bibliography

Bell, A. G. (1907) The Mechanism of Speech. New York: Funk and Wagnalls.

Kent, R. D., B. S. Atal, and J. L. Miller (eds.) (1991) Papers in Speech Communication: Speech Production. Acoustical Society of America.

Ohala, J. J. (1990) Respiratory activity in speech. In W. J. Hardcastle and A. Marchal (eds), Speech Production and Speech Modelling. Dordrecht: Kluwer. 25-53.

Ohala, J. J. (1993) The whole body plethysmograph in speech research. https://archive.org/details/MeasuringSpeech1993
 
Stetson, R. H. (1928) Motor Phonetics. Current (1988) edition, edited by J. A. S. Kelso and K. G. Munhall. Boston: College-Hill Press.


2. Background (Kent et al. 1991: 1)

'An early theoretical opposition in the study of respiratory function in speech was between the ideas of the two eminent phoneticians, Alexander G. Bell (1907) and Raymond H. Stetson (1928). Bell proposed that the respiratory system provided a continuous driving pressure to support speech production ... analogous to the bellows of a church organ. ... Stetson took the different view that the syllable, presumed to be the basic unit of speech, is associated with a pulse of air. He believed that "every syllable has its chest pulse delimited by the chest muscles (intercostals) or by the constriction (complete or partial) of the consonant, or both".'


3. Ohala (1990): literature review (sections 1 - 4.5.2) cites a long literature (e.g. Sweet 1911, Stetson 1928) claiming "that stress (or the prominence given to individual syllables) is implemented by greater respiratory effort. ... Ladefoged (1962b, 1967) and his colleagues found a momentary increase in the activity of the expiratory muscles (sampled via EMG needle electrodes) and in P_s [subglottal pressure] during and sometimes immediately before a stressed syllable".

Hypothesis: stress is implemented by expiratory muscle contractions.


4. Ohala (1990): Procedures

• Subjects: 2 M speakers of American English (one of whom was the author), one M Swedish speaker, one F Hindi speaker.
• Recordings and equipment: Oral and nasal airflow (from a face mask), lung volume (from a whole-body pressure plethysmograph) and its first derivative, speech waveforms.
• Data: i) n? tokens of the sentence "deem unilluminable real", with and without emphatic stress on the syllable [lum]; ii) tokens of the "deem {m/s/h/t^h}oon real"; Hindi nonsense words [ʔipi], [ʔibi], [ʔimi], [ʔip^hi], [ʔib^hi]; "Lure Bill near"; "You drew Will/pill near".
• Data analysis: informal, on visual examination of the recordings.
• 
  

5. Ohala (1990): Results ("observations")

• Relatively large, rapid decreases in lung volume during moments of high airflow, e.g. during aspiration, [h], and fricatives. "These presumably represent a passive collapse of the lungs due to the rapid air flow out of the lungs". For sonorant consonants, the rate of lung volume decrement is the same as that for surrounding vowels.
• Moments of less than normal lung volume decrement during periods of reduced airflow, e.g. closure portion of stops and affricates.
• Momentary greater-than-normal decreases in lung volume during emphatically stressed syllables.
• No obvious change in the rate of lung volume decrement in non-emphatic, normal conversational speech.
• A close and inverse relationship between lung volume decrement and volume of airflow exiting through the mouth and nose.

Ohala interprets these results as showing "that the primary function of the pulmonic system during speech is simply to produce P_s that is reasonably constant and above some minimal level. Short-term passive variations in lung volume decrement occur in reaction to variations in lung pressure which in turn are reactions to changing downstream resistance to airflow caused by oral articulations. Active short-term variations in lung volume decrement are probably limited to the production of variations in the loudness of speech". In short, Stetson's hypothesis disproved.