118 BELL SYSTEM TECHNICAL JOURNAL 



pressing an idea, the kind and number of words used and hence the 

 time taken. 



The Hstener also adjusts himself to the conditions by an amount 

 which is determined somewhat by his interest in the matter under 

 discussion. He may strive to comprehend the transmitted ideas and 

 require few repetitions by the speaker or he may refrain from exerting 

 himself and so tend to evoke greater effort on the part of the talker. 

 At times he may pretend not to understand in order to get confirmation 

 of a statement or to gain time in replying to a question. 



In view of these factors and of the normal variations of different 

 talkers and listeners in all these respects, the portion of the questions 

 and statements of conversation which is correctly understood and the 

 time required to interchange certain ideas may vary widely for different 

 conversations even when they are carried on under a fixed set of local 

 conditions. If it were desired to determine a measure of the conversa- 

 tional satisfactoriness of these conditions, in addition to some quantita- 

 tive method for rating each conversation, there would be required, 

 therefore, observations on a large number of conversations between 

 different people in order to take account of the variables due to the 

 material of conversation, the people, and their abilities and desires to 

 accommodate themselves to the conditions. 



Telephone Conversations 



In telephone conversations, there are adjustments between talker 

 and listener as is the case in direct conversations, but there are certain 

 definite differences in this regard because of the interposition of the 

 telephone circuit between the participants. Here also, the speaker 

 tends to adjust his talking level to the loudness with which he hears his 

 own voice. In this case, however, he hears his own voice not only 

 through the air path, but also through the "sidetone" of the telephone 

 set, that is, through the electrical path from his own transmitter to his 

 own receiver. When this electrical path is more efficient than the 

 acoustic path, the sidetone will tend to control the talking level. It 

 has been found that varying the sidetone of the set has on the average 

 a definite effect on the talking volume of the speaker, the talking 

 volume being lowered as the efiiciency of the sidetone path becomes 

 greater. 



In a telephone conversation, there is also a tendency for a person 

 in talking to adjust his volume on the basis of the loudness with which 

 he hears the person at the other end of the circuit. If the voice of the 

 other person comes through weakly, he judges that the connection re- 

 quires loud talking and acts accordingly. If the listener indicates that 



