182 



NA TURE 



[December 26, 1901 



■■"^^ EXPERIMENTAL PHONETICS} 

 TTHE movements of the organs of voice and speech are so 

 ■*■ complicated as to require for their elucidation the appHca- 

 tion of many methods of research. When one speaks there are 

 movements of the lips, tongue, soft palate and larynx, and some- 

 times movements of the muscles of expression. Then, again, 

 there are special characteristics about vowel sounds which 

 apparently distinguish these from the sounds of musical instru- 

 ments. Thus questions ari.se as to the true nature of vowel- 

 Sounds and as to what is the physical constitution of a word of 

 several syllables. It has also been suggested that language 

 might be recorded, not by letters or syllables, but by signs or 

 symbols which would indicate what had to be done by the vocal 

 and articulating organs for the production of any given sound. 

 There might thus be a physiological method of expressing speech 

 by a series of alphabetical symbols for sounds varying in pitch, 



r\ 



Special contrivances were devised for transmitting these move- 

 ments to three of Marey's tambours, so arranged as to record on 

 the surface of a blackened drum three superposed curves which 

 indicated the order of succession, duration and intensity of the 

 movements of the organs. The emission of air from the nostril 

 indicated movements of the soft palate, and these were signalled 

 by an indiarubber tube introduced into one nostril while the 

 other end was connected with a tambour, as in Fig. I. A small 

 electromagnetic apparatus was placed over the larynx, and by 

 making and breaking a current the vibrations of the larynx were 

 transmitted to another tambour. The movements of the lips 

 were recorded by a device which caused the pressures to act on 

 a third tambour, as is shown in the figure. 



This method was found to give characteristic tracings for 

 the sounds of consonants, but the records obtained from vowel- 

 sounds were all very much alike. It was also observed that if 

 one of the tambours did not act, say the one recording the 

 vibrations of the larynx, it wasdifficult to distinguish the tracings 

 of certain consonantal sounds. Thus p resembled b, so far as 

 the movements of the lips and the nasal pressures were con- 

 cerned, but with h there is a vibration of the larynx as well, 

 while this is absent in the tracing of ^ In Fig. 2 is shown a 

 table in which is depicted the traces obtained on uttering the 

 vowel a either before or after various consonants. In these 

 tracings, i' n indicates nasal pressure, v / vibrations of the 

 larynx, and .m / movements of the mouth. Five examples are 

 given of combinations of A with consonants. If there is no 

 emission of air from the nostrils, the line i' n is unbroken and 

 horizontal, but if there is emission then an elevation is 

 seen as in A 3 with a m, or m a, A sinuous line in V / 

 shows that the larynx vibrates, but if there is no laryjigeal 

 vibration the line is straight. It will be observed that in some 



Fig. r. — .\ method for recording simullaneoiisly the different acts of speech : 



of the larynx (s), and movements of the lips (3). 



ntensity and quality. It will be seen that experimental 

 phonetics constitutes a wide field of research, not only of great 

 scientific interest, but also one having practical aspects not at 

 first apparent. From the nature of the investigation, also, the 

 problems seem to be specially suited for the application of the 

 graphic method of research. 



In 1875, an investigation was carried out by Havet and 

 Rosapelly - in the laboratory of Prof Marey in Paris, in which 

 the pressure of the air in the nose, the movements of the lips, 

 and the vibrations of the larynx were simultaneously recorded. 



1 By Prof. John G. M'Kendrick, F.R.S. Read before the Section of 

 Physiology at the meeting of the British Association in Glasgow, Septem- 

 Ijcr 13. " Die Pbonetische Liter.ilur von 1876-1895." By Hermann Breymann 

 (Leipzig, 1897): "The ArticuL-ttion of Speech Sounds by Analphabetic 

 Symbols." By Otto Jespersen (Marburg, 1889): " L'Inscription des 

 Phiinomines Phonitique." By M. J. Marey (Rcvuc Gfnirale dcs Sciences, 

 iselsojuin. 1898): "Studies from the Yale Psychological Laboratory." 

 By E. W. Scripture (1899) ; " TbiSorie de la Formation des Voyelles." By 

 Marage (Paris, prix Barbier, 1900); "La Parole d'aprfcs le 'IVacd du 

 Phonographe." By H. Marichelle (Paris, 1897); "On Vowel Sounds." 

 ByJ.G. M'Kendrick and A. A. Gray, Schafers " Text-Book of Physio- 

 logy," vol. ii. p. 1206, in which the recent bibliography is given in detail. 



2 Rosapelly ; "Inscription des Movements Phoncliques," in "Travaux de 

 Laboiatoire de iM. Marey" (Paris, 1875). 



cases the larynx vibrates throughout all the experiment, as in 

 A 2, while in others there is an interruption, as in B i. The 

 movements of the lips in M /show a curve which varies in 

 amplitude and duration according as the lips are more or less 

 approximated and according to the duration of complete or 

 partial occlusion. 



These syllabic sounds inay be termed phones. This research 

 is an excellent example of the application of the graphic method 

 to the movements ol speech. The method has been much de- 

 veloped by Rousselot ' in the College de France, where there 

 now exists a special laboratory for research in phonetics. 



Prof Marey, whose earlier researches are well known to have 

 had much to do with the development of the kinemalograph, 

 employed, so long ago as 1S8S, chronophotography to catch 

 those evanescent changes of the countenance, the sum total of 

 which give expression to the face in speech. In Fig. 3 are seen 

 the changes of expression in a woman's face in speaking, during 

 a period of half a second. If these successive pictures are 

 projected by a lantern (Fig. 4) there is an animated face on the 

 screen. In this w.ay Marichelle succeeds in placing before the 



1 Rousselot: " Principes de PhoniSlique Expiirimeniale " (Paris, 1897). 



NO. 1678, VOL. 65] 



