246 CARNEGIE INSTITUTION 



the first of their kind ever obtained, other investigators having con- 

 fined themselves to single vowels sung into the apparatus (Hermann, 

 Pipping, Bevier) or to single spoken words (Pipping). 



Since the entire impression that passes by the voice from speaker 

 to hearer is contained in vibrations of the air at the mouth of the 

 speaker, the curves of speech can be assumed to contain solutions for 

 all the problems of vocal expression. Among these problems are the 

 following : 



1 . To determine the essential characteristics whereby one part of 

 speech is distinguished in general from another — for example, the 

 vowel a from the vowel c ; or, briefly, the acoustic nature of speech 

 sounds, particularly of vowels. Do the vowels consist of tones from 

 the vocal cavities that have fixed pitches (Helmholtz, Hermann) or 

 may they be of any pitch, provided they have fixed relations of pitch 

 among themselves (Lloyd) ? Are the resonance tones of vowels re- 

 lated to the tone from the vocal cords as overtones to a fundamental 

 (Helmholtz) ; or do they have no such relation (Hermann) ? If the 

 latter is true, what is the explanation of such an apparent physical 

 impossibility (Rayleigh) ? Are previous investigators right in look- 

 ing for the essentials of a vowel in the resonance tones alone ? Why 

 is the prevailing view of the nature and action of the vocal cavities 

 inadequate to explain the thoroughly established results? 



2. To investigate the rise and fall of the voice during speech — 

 briefly, the characteristics of the melody of speech. In song the 

 melody is apparently of a very simple nature ; in spite of its use for 

 emotional expression, the melody of song cannot convey all the 

 shades of feeling that are possible by the melody of speech. This 

 melody differs for every condition of mind, for every person, for 

 every dialect, for every language. What are its fundamental laws? 



3. To investigate the characteristics of rhythm in prose speech. 

 Both prose and verse have their elements arranged on more or less 

 regular systems in respect to stress, duration, melody, and other 

 elements. A complete study of these elements for verse would solve 

 the problem of the nature of verse. The theories hitherto proposed 

 were established on the basis of judgments by the ear alone, and 

 they certainly all quite miss the essentials of the case. They are 

 theories of verse as it appears in type rather than of verse as it comes 

 from the mouth of the poet. 



4. To establish the types of American vowels. There are not 10 

 or 12 typical vowels, as often supposed, but more nearly a hundred 

 of them, as distinct and as indistinct from each other as the races 



