Dynamical Study of the Vowel Sounds 

 Part II 



By IRVING B. CRANDALL 



Synopsis: Comparative studies based on oscillographic records of the 

 principal characteristics of vowel, semi-vowel, and consonant sounds, 

 have contributed much to an understanding of the mechanism of speech. 

 Analyses of the frequency spectra of vowels show almost invariably two 

 principal resonance peaks which fact is suggestive of a double resonator to 

 produce them. 



The present paper is concerned with the mechanism of the double 

 resonator system and a mathematical treatment thereof. Based on the 

 volume, shape and coupling of the resonating chambers, some models of 

 cardboard, tube and plasticene were made, and with which some experi- 

 mental tests in the production of vowels were Carried out. The best 

 success was had with the sound a (father) while fair results were obtained 

 with the sound o, a and e. 



Introduction 



IN two earlier papers ^ a diagram has been given of the frequency 

 spectra of the vowel sounds, based on analyses of a large number of 

 accurate oscillographic records. In addition, there was given, in the 

 second of these papers, a comparative study of the principal character- 

 istics of vowel, semi-vowel and consonant sounds, and an account of 

 certain studies made by other investigators whose methods and results 

 have contributed to our understanding of the mechanism of speech. 



Among the more original of recent contributions are those of Sir 

 Richard Paget,- who has successfully employed multiple resonators to 

 simulate almost all the vowel and consonant sounds. In getting to- 

 gether the material for the second paper from the Bell Laboratories, 

 Paget's results for the vowel sounds were compared with ours only in a 

 general way, and not in so detailed a manner as was followed in the 

 discussion of consonant and semi-vowel sounds in that paper. It may 

 be permissible to return to a consideration of the vowel sounds in the 

 present paper, following Sir Richard Paget's idea of the double res- 

 onator as the instrument for vowel production. Indeed Sir Richard 

 has pointed out to us that, since our own data on the spectra of the 

 vowel sounds show almost invariably two principal resonance peaks, 

 there must be a double resonator to produce them, thus harmonizing 

 our results, at least for the male voices, with his ow^n. 



' I. B. Crandall and C. F. Sacia, "Dynamical Study of the Vowel Sounds," also 

 "The Sounds of Speech," Bell System Technical Journal, III, 1924, p. 232; 

 bid., IV, 1925, p. 586. 



^ Proc. Roy. Soc, A102, 1923, p. 752; ibid., A106, 1924, p. 150. 



100 



