SPEECH 



1713 



'delayed side-tone' may persist for a short period 

 after the latter is removed. This finding is not how- 

 ever confirmed (4). It is also noteworthy that there 

 is typically an increase of vocal intensity with in- 

 creasing increments of delay, most marked within the 

 range o to 0.09 sec. (8). This indicates that delayed 

 speech playback, apart from provoking articulatory 

 disorder, may give rise to significant change in both 

 the speed and intensity of speech. 



The results of a more recent study by Fairbanks 

 (35) have confirmed the incidence of changes in 

 both duration and articulation of speech. These 

 changes, unlike the concomitant changes in sound 

 pressure and frequency, are interpreted by this 

 author as primary effects of the time delay. He- 

 proposes an index of correct word rate as the si 



adequate combined measure. 



Although the modus operandi of aural feedback in 

 the control of speech is obscure, ii has been suggested 

 that the mechanism is similar in principle to that of an 

 aurally monitored manual task, such as operating .1 

 machine gun (75). If this analogy is sound, one ma) 

 ascribe the disturbance produced by delayed play- 

 back to the fact that the subject is endeavoring to 

 speak at his normal rate but that the dela) is sufficient 

 to enable him unintentionally to produce an 



additional phoneme or syllable. This view gains si ■ 



support from recent studies of stuttering (22, 23). 



SPEECH PRODUCTION 



The production of speech involves highly co 

 ordinated movements of the abdominal wall, thorax, 

 larynx, soft palate, tongue and lips. Although under 

 voluntary control, the movements involved in speak- 

 ing are normally carried out without conscious 

 direction and may be said to constitute highly auto- 

 matized patterns of reaction. In voiced speech, a 

 regulated expiration forces air past the vocal cords, 

 effectively chopping the air stream into a series of 

 rhythmical puffs (phonation). In the classic view, it is 

 supposed that the vocal cords set up in the adjacent 

 air a complex motion (cord tone), consisting of a 

 fundamental and a large number of overtones (90, 

 93). This spectrum is then modified in various wavs 

 by passage of air through the vocal cavities which are 

 traditionally assumed to operate as simple resonators 

 (vocal resonance theory). In the production of 

 vowel sounds, the tongue is so placed as to approach 

 the roof of the mouth or the back of the throat, thus 

 allowing air to vibrate freeh' in and out of the resonat- 



ing cavities. In the production of consonant sounds, 

 on the other hand, the tongue is placed in close 

 apposition to the throat or palate, thus restricting the 

 passage of air. A low turbulence is set up which 

 produces vibrations of considerably higher frequencv. 

 In general, the pitch and character of the sounds 

 emitted in phonation may be said to depend upon 

 the varying 'attitudes' of the vocal cords and of the 

 structures involved in articulation. 



The vibrating systems produced by the vocal cords 

 and the air columns above them have been subjected 

 to detailed experimental and mathematical analysis 

 (29, 30, 67, 76). Broadly, the results may be said to 

 favor the 'vocal resonance' theory which has also 

 derived support from direct studies of chest resonance 

 in man (67). It has also proved possible to construct 

 synthetic 'speech machines' which effectively dupli- 

 cate certain aspects of human speech production. 

 The earliest of these were mechanical (93, 1 05 1; the 

 more recent, electrical and electronic (30, 31, 47, 

 U4, 126). An adequate short account of 'synthetic 

 speech machines' has been given by Dunn (31). 



Respirator) Moi ments in Relation /<» Speech 



It has been pointed out that the rate of production 

 of the v .a inns speech sounds is limited not only by the 

 innervation of the muscle groups concerned but also 

 by the fact that speech is to be regarded as a function 

 'overlaid' upon respiration (64, 85). Indeed nervous 

 arrangements ol some complexity govern the relations 

 of respiration to speech. In normal speakers, there is 

 a fairly close correspondence between abdominal and 

 thoracic movements in breathing, a marked increase 

 in the relative length of expiration dining speech, 

 and a certain independence of breathing and vertical 

 movements of the larynx (132). These relations may 

 lie markedly altered in cases of speech disorder. In 

 stutterers, for instance, there is not uncommonly 

 opposition between thoracic and abdominal move- 

 ments in breathing, marked protraction of inspira- 

 tion, and vertical movement of the larynx synchronous 

 with the movements of respiration (132). It is there- 

 fore evident that stuttering, whatever its cause may 

 be, invokes widespread incoordination of the entire 

 speech-breathing apparatus. 



There is a considerable range of variation between 

 normal speakers in respiratory habit, with some 

 suggestion of a sex difference (67). At the same time, 

 it may be doubted whether peculiarities of voice, such 

 as carrying power, bear any constant relation either 

 to type of respiration or to vital capacity. Type of 



