256 BELL SYSTEM TECHNICAL JOURNAL 



vertical axis. The duration of the sounds in fractions of a second is 

 represented on the horizontal axis. It will be seen that the pitch 

 rises and falls as the various sounds are spoken. This representation 

 of the pitch variation is called the fundamental melodic stream. It is 

 the melody in the same sense as this term is used in music, although 

 it is evident that the pitch changes do not take place in musical 

 intervals as would be the case if the sentence were sung. 



To show the contrast, a graph was made when the sentence was 

 intoned on the musical intervals do, re, mi, fa, mi, re, do. An analysis 

 of the graph gave the result shown in Fig. 1. In the case of the sung 

 sentence the pitch changes are in definite intervals on the musical 

 scale, while for the spoken sentence the pitch varies irregularly, 

 depending upon the emphasis given. The pitch of the fricative and 

 stop consonants is ignored in the musical score, and since these 

 consonants form no part of the music, they are generally slid over, 

 making it difficult for a listener to understand the meaning of the 

 words. Some of our friends in the musical profession may object to 

 this statement of the situation, but I think it will be agreed that a 

 singer's principal aim is to produce beautiful vowel quality and to 

 manipulate the melodic stream so as to produce emotional effects. 

 To do this, it is necessary in singing to lengthen the vowels and to 

 shorten and give less emphasis to the stop and fricative consonants. 

 It is for this reason that it is more difficult to understand song than 

 speech. 



There are two secondary melodic streams of speech represented by 

 the second and third curves from the bottom of Fig. 2, which are due 

 to the resonances imposed upon the speech sound by the throat and 

 mouth cavities. The numbers on these curves give the number of 

 the harmonic which is reenforced. These two secondary melodic 

 streams are not sensed as changes in pitch, but rather as changes in 

 the vowel quality. Then there is a fourth stream, or, it would probably 

 be better to say, a fourth series of interrupted sounds which are very 

 high in pitch and are the sounds which enable us to identify the 

 fricative consonants. The secondary melodic streams produced while 

 speaking the same sentence are approximately the same for different 

 persons, even for a man and a woman, while the fundamental melodic 

 stream is usually quite different. This latter stream is not used in 

 identifying words, but it is used sometimes to give different meanings 

 to the same words. 



As one listens to this sentence he hears the variations in loudness 

 as well as in pitch. Loudness is related to the amplitudes and fre- 

 quencies of the components of the tone, but this relationship is very 



