SEVENTY-FIVE YEARS OF THE TELEPHONE 227 



Siibjective Measurements 



We come now to the carrying out of Arnold's concept of using this refer- 

 ence system as a means of investigating the effects of distortion on the recog- 

 nition of reproduced speech sounds. This was first started in the Laboratories 

 under Crandall and then continued under Fletcher and his associates. This 

 work involved the use of people as meters, with the problems of their cali- 

 bration. 



For such tests, lists of monosyllables were prepared which went far be- 

 yond the simple lists proposed by Campbell. A large amount of work was 

 done in the determination of the basic sounds to be used, the most suitable 

 form of syllables and the arrangement of syllables in groups to have balance 

 with respect to their content of basic sounds.^ ° 



Starting with the earlier versions of the Master Reference System, the 

 effects were measured by these articulation tests of changes in loudness, 

 distortion and accompanying noise, on the understanding of the reproduced 

 sounds and syllables. This study of distortion included resonance such as 

 characterized commercial instruments and the variation of response with 

 frequency as encountered in commercial circuits. Also, extensive articulation 

 tests and analyses were devoted to the fundamental investigation of the 

 effects of bandwidth as provided by electric wave filters of the Campbell 

 type, and from these was derived a quantitative determination of the im- 

 portance of the different parts of the transmitted band on the recognition 

 of the reproduced sounds of speech. This work is described in Fletcher's 

 book "Speech and Hearing"^^ and in many papers Hsted in the bibliog- 

 raphv ^^' ^^' ^^' ^^' ^^' ^^ 



From this work there was developed also a procedure for computing the 

 articulation of a telephone circuit from the physical characteristics of the 

 circuit. With the availabihty of this computational method^^- ^^ it has been 

 practical to discontinue articulation testmg itself except for special purposes. 



Another factor which comes into telephony as an important effect in the 

 use of the telephone is "sidetone." The speaker's voice reaching his own 

 ear through the sidetone path of the telephone set reacts on his loudness 

 of talking, this loudness being decreased unconsciously as the sidetone is 

 increased. Also, in Hstening, sidetone introduces into the listening ear the 

 room noise picked up by the transmitter. Both of these effects of sidetone 

 were studied in the laboratory under controlled conditions, so that an ap- 

 preciation was obtained of the magnitude of their effects. 



There still remain the question as to applicability of the effects of vol- 

 ume, distortion, noise and sidetone as determined in the laboratory to com- 

 mercial telephony with the conditions uncontrolled at the telephone stations 

 and the users untrammeled in their habits and reactions. Information re- 



