REPORT OF EXECUTIVE COMMITTEE. II5 



facts like the following : American long vowels need not be diph- 

 thongized ; both short and long vowels may be diphthongized. The 

 short vowels are often different from the ones supposed to be present. 

 In the same word in similar phrases only a few seconds apart a 

 speaker may use two utterly different short vowels. The number 

 of typical vowels must be greatly increased beyond those recognized 

 by the dictionaries. The short vowels are often incorrectl)^ given 

 in the dictionary pronunciations. A vowel is not a constant thing, 

 but changes at every wave of its vibrations ; it is an activity and 

 not a dead product. The ear gets a general impression from the 

 whole series of waves and can not distinguish the actual sound at 

 any point. The ear is often misled in the rapidly changing short 

 vowels. Sounds have no definite limits, but fuse more or less grad- 

 ually into each other. The division of words into syllables and of 

 verse into feet on present principles is nonsense, which can be avoided 

 only by a new view (psychological and not typographical) of the 

 nature of S3^11ables and speech units. This new view (the centroid 

 theory) is in accord with the experience of writers of verse (the at- 

 tempt of modern writers on prosody to fit Greek and Latin notions 

 to English verse results in a pedantic scheme of spelled verse that 

 ignores the poet and the public, although it may please the printer); 

 the frequent presence of ' ' sonant h " in American English was 

 proved. Various individual differences were investigated. Melody 

 and intensity were found to vary in each vowel by different speakers. 



The vowel curves showed that ordinary views of resonance could 

 not be applied to speech ; the vocal cavities have a soft wall. The 

 laws of resonance for such cavities differ from those for cavities with 

 hard walls. The glottal lips (which are masses of flesh, and do not 

 in any way resemble cords or bands) emit series of puffs, and do not 

 vibrate in the ordinary sense. 



An investigation was conducted on the laws of resonators with 

 soft walls (of water, gelatine, flesh, etc.) and on the action of puffs 

 of varying sharpness. On the basis of the results a new vowel theory 

 was elaborated. This theory finds its confirmation in the fact that 

 all the vowels can be artificially produced by apparatus built accord- 

 ing to it ; and also in the fact that countless numbers of speech curves 

 become for the first time intelligible when interpreted according to 

 it. This theory takes account for the first time of the softness of the 

 walls of the vocal cavities and of the flesh character of the glottal 

 lips. It proposes the new notion that the muscular fibers in the 

 M. vocalis of the glottal lips contract differently for each vowel, and 



