wv 
 
 8^33^ 
 
 3 1822 01221 7865 
 
 
 
 
 
 THE 
 
 .NISM OF SPEECH 
 
 BY 
 
 ALEXANDER GMHAM BELL
 
 WV 501 B433 190* 
 UNIVERSITY OF CALIFORNIA SAN DIEGO B 
 
 3 1822 01221 7865 
 
 E^ 
 
 5^5 5 
 
 ! ■:'■ 
 
 611 
 
 4'^^yi>:^^^J L^<yia^/i£^^. 
 
 -f^^-PXJ'At/j 
 
 ^^'^'-^^^^^ "^^^i^A^tyl^
 
 LECTURES 
 
 UPON 
 
 THE MECHANISM OF SPEECH 
 
 BY 
 
 Alexander Graham Bell 
 
 Reprinted from the proceedings of the First Summer 
 Meeting of the American Association to pro- 
 mote the Teaching of Speech to the Deaf 
 
 FUNK & WAGNALLS COMPANY 
 
 NEW York and London 
 
 1906
 
 COPYRIGHT, 1906, BY 
 
 ALEXANDER GRAHAM BELL
 
 CONTENTS 
 
 The Thorax and Larynx ,.....! 
 
 The Pharynx and JVIouth in their Relation to Speech . 17 
 
 The Functions of the Epiglottis and Soft Palate . . 32 
 
 Visible Speech as Taught to the Deaf . . . 51 
 
 Consonants . . . . . . . -84 
 
 Vowels, Glides, and Combinations .... 100 
 
 Speech Reading . . . . . . .113 
 
 APPENDIX 
 Vowel Theories . . . . . • .117
 
 THE THORAX AND LARYNX. 
 
 The thorax is the treasure-house of the human body, — a verit- 
 able strong-room, girt about with walls of bone for the protection of 
 those precious organs the heart and lungs. Let us imagine our- 
 selves for a moment inside the thorax, but first, with your permis- 
 sion, let us empty this safe-deposit vault of its valuable contents, so 
 that we may have space for exploration. 
 
 We fmd ourselves in a dark room or vault with a door in the 
 roof. The floor of this vault, instead of being firm and solid, is a 
 soft membrane or muscle, — not flat like an ordinary floor, but dome- 
 shaped like the top of an open umbrella. The door above is a sort 
 of double trap door set at an angle instead of being flat, and open- 
 ing upwards. But the most extraordinary thing about this room is, 
 that the floor is in constant motion, heaving upwards and down- 
 wards in regular pulsations. The trap doors also are in motion ; 
 now they are opened so that a glimpse can be obtained of passages 
 above, and now they come together with a quivering motion, open- 
 ing and shutting with great rapidity, and causing a vibration that 
 makes the whole thorax tremble. The walls also are in motion, the 
 whole room alternately increasing and diminishing in size. 
 
 A membranous muscle when it contrads tends to become flat 
 and tense; and many of us have had the idea that the diaphragm or 
 dome-like lloor of the thoracic cavity, in contrading becomes flat 
 like the head of a drum. This idea is incorred, for the central 
 portion of the diaphragm is attached above by ligaments and tis- 
 sues to the bony walls of the thorax, so that it is incapable of 
 descent. The circumference or edge, also, is attached. When, 
 therefore, the diaphragm contracts, the dome-like lloor becomes 
 somewhat conical in shape. As I pifture the action in my mind, it 
 is as though the dome of the capitol in Washington were to change 
 
 into a cone somewhat like a blunt church spire. 
 
 1
 
 When, then, the diaphragm contracts, the thoracic floor 
 becomes tense and somewhat conical in shape, and the cavity 
 of the thorax is thus enlarged. When the muscular fibres relax, 
 the tense floor becomes loose and baggy, resuming its dome-like 
 shape, and the space within the thorax then becomes less. Con- 
 tinuing our explorations we find that the front or chest wall of the 
 thorax is capable of slight motion. By the operation of cer- 
 tain muscles, the ribs can be raised to a limited degree, so as to 
 cause an increase in the circumference of the chest, and thus an 
 expansion of the thoracic cavity. In animals which are prostrate 
 gravity helps the expanding aftion, but in man, on account of his 
 upright position, the weight of the bony framework renders a dis- 
 tinct effort necessary in order to elevate the chest wall, and relaxa- 
 tion of the muscles tends to collapse and consequent contraftion of 
 the thoracic cavity. 
 
 It will thus be seen that the interior capacity of the thorax can 
 be increased; (i) by the contraction and consequent depression of 
 the diaphragm; (2) by the elevation of the front wall of the chest; 
 or, (3) by both actions performed simultaneously. When the interior 
 capacity is increased, the air within the thorax expands to fill the 
 increased space, thus becoming rarified. If the trap doors are open 
 the denser air of the atmosphere then presses its way into the thorax 
 to supply the partial vacuum. The act of inspiration is completed 
 when the air pressures within and without the thorax are equal. If 
 now the interior capacity of the thorax be diminished, the contained 
 air by compression becomes denser than the air outside, and therefore 
 tends to rush out, and the act of expiration is completed when the 
 air pressures within and without are equal. 
 
 Exhalation can be effected; (i) by relaxation of the diaphragm 
 which rises into its dome-like shape; (2) by relaxation of the mus- 
 cles that raise the ribs, thus allowing the front wall of the chest to 
 fall ; or (3) by both processes performed simultaneously. We can- 
 not, however, by any of these means produce that forcible expul- 
 sion of air that is requisite for speech, for relaxing muscles cannot 
 exert much compressing power. What we need for speech is a 
 forcible compression of the thoracic cavity. This can be effected by 
 the abdominal or waist muscles. The contraction of these muscles 
 produces a compressing effect upon the viscera, just as though a 
 rope were passed around the waist and drawn tightly. This com- 
 pression forces the viscera upwards against the under side of 
 the diaphragm. The diaphragm is thus pushed up like a piston 
 into the thoracic cavity, compressing the contained air. In this way
 
 forcible emission of air is caused by the contradlion of the abdom- 
 inal muscles, and these are the muscles that we employ in throwing 
 out the voice. For example: — Prolong a vowel sound, suddenly 
 increasing the force into a shout, a number of times in succession, 
 without stopping the voice, thus: — ah, AH-ah-AH-ah-AH. At 
 every shout a forcible contraction of the abdominal muscles can be 
 felt by the hand, and the front wall of the chest is thrown upwards 
 by the force of the compressed air within the thorax, pulsating out- 
 wards with every shout. 
 
 Alternate inspiration and expiration, result from alternate expan- 
 sion and contraction of the thoracic cavity. This can be effected in 
 two ways. 
 
 We can expand the cavity; (i) by using muscles that tend to 
 raise the ribs and cause them to separate from one another slightly ; 
 and (2) by depressing the diaphragm. 
 
 We can contraft the cavity; (i) by allowing the chest wall to 
 fall, using muscles that tend to bring the ribs nearer together; and 
 (2) by employing the abdominal or waist muscles. 
 
 Of these two possible modes of ac^tion, it will be seen that one 
 involves the expenditure of less energy than the other. It is less 
 laborious to breathe by using the diaphragm and waist muscles, than 
 by moving the heavy bony framework of the chest. 
 
 When the diaphragm contrads, changing from the dome-like 
 to the conical shape, it presses downwards upon the viscera, thus 
 causing an expansion of the abdominal cavity. When the abdom- 
 inal muscles contra6l, the circumference of the waist diminishes. 
 Thus in natural breathing, produced by the alternate action of the 
 diaphragm and the abdominal muscles, the circumference of the 
 waist increases during inspiration, and diminishes during expiration. 
 
 I doubt the advisability of directing a pupil's attention to these 
 motions, for his attempts at reprodudion are often attended by 
 ludicrous results. The end desired would, I think, be better 
 attained by direding his attention to the chest, and not to the 
 abdomen. Get the pupil to expand the chest and keep it con- 
 tinuously expanded even when breathing out. If the bony frame- 
 work of the chest is kept raised and fixed, breathing can only be 
 performed by the diaphragm and waist muscles ; and, as the pupil 
 cannot help breathing, nature will work the proper muscles without 
 his knowledge or will. 
 
 This effort of continuous expansion can only be sustained for a
 
 few minutes at a time without fatigue by persons unaccustomed to 
 the exercise ; but if persevered in day after day the pupil can 
 acquire the power of sustaining the chest wall continuously during 
 his waking hours. The exercise usually results in a marked 
 increase in the capacity of the chest, I have known of instances 
 where the circumference of the chest has increased between two 
 and three inches after a month's practice. When we consider that 
 the thorax is the storehouse of the lungs, it is obvious that this 
 increased capacity will be beneficial to health. Deaf children, 
 especially, require exercises of this sort, because their lungs have 
 not been as fully exercised as those of ordinary children. 
 
 I consider this exercise of chest expansion as more beneficial 
 to pupils than the breathing exercises that are usually employed. 
 Conscious regulation of the breath is to be deprecated for the 
 following reason : the primary objeft of breathing is the oxygen- 
 ation of the blood, and the getting rid of the produds of com- 
 bustion in the lungs. We take in air to oxygenate the blood. We 
 exhale to get rid of carbonic acid gas and aqueous vapor. The 
 proper time to take in breath cannot be dicated to a pupil without 
 interfering with the primary fundion of the lungs. Nature gives 
 the signal for inspiration when the blood needs oxygenation, and 
 when we attempt to regulate the breath consciously we are apt 
 to interfere with the circulation of the blood. Breathmg exercises 
 should be stopped the moment dizziness is produced., for that is 
 nature's indication of a disturbance in the circulation. 
 
 Inspiration is utilized for the oxygenation ot the blood, an.j 
 expiration alone is employed in the production of speech. Observe 
 the breathing of a person engaged in conversation at a time when 
 he is unconscious of your observation. You will find that many 
 words are articulated between each inspiration. The time taken 
 for inspiration is instantaneous, whereas the duration of the expira- 
 tion is very long. The breath comes in quickly, and goes out 
 slowly. This means that the trap doors in the roof of the thoracic 
 cavity are opened widely during inspiration, and closed so tightly 
 during the aft of speech, that only a fine stream of air can escape 
 from the thorax. The prime requisite for speech is a store of com- 
 pressed air, which can be let out little by little, as wanted. It is 
 obvious that the air would escape with a gush unless restrained. 
 The trap doors already alluded to. constitute the chief means by 
 which a too rapid escape of air is prevented. These trap doors are 
 known as the vocal cords, and they are contained in the larynx.
 
 THE LARYNX. 
 
 Fig. I. 
 
 Vertical seftion of the Larynx 
 as seen from behind. 
 
 Vertical seAion of the Larynx 
 as seen from the side. 
 
 The larynx may be considered as the guard-house of the lungs, 
 — admirably adapted to protect them from injury. It consists essen- 
 tially of a box, surmounted by a movable lid, called the epiglottis, 
 which closes during the aft of swallowing, to prevent food from 
 
 entering the lungs. Inside this box are two pairs of valves. The 
 lower pair, called the true vocal cords, stand with their free edges 
 upward, and the upper pair, known as the false vocal cords, hang 
 with their edges downwards. The lower pair constitute the trap 
 doors to which I have before alluded.
 
 The specimens of the larynx which have been placed in your 
 hands by Dr. Hewson, will have shown you that the vocal cords do 
 not look like "cords"' or strings at all — nor perhaps are they very 
 suggestive of trap doors — they look more like lips. In fad, we may 
 consider that we have a pair of lips within the larynx capable of 
 approximation, with various degrees of tension. 
 
 The false vocal cords also resemble lips, but it is doubtful 
 whether they can be approximated sufficiently to touch one another. 
 If the lips of the throat are shut and the abdominal muscles are con- 
 traded, the air compressed within the thorax tends to blow the 
 vocal cords apart. The muscular tension may be so adjusted as to 
 allow the vocal cords to yield sufficiently to permit of the escape of 
 a puff of air — the aperture closing again immediately and remain- 
 ing closed until the pressure within the thorax causes the escape of 
 another puff. In this way a regular series of puffs may be produced, 
 the alternate opening and closing of the glottis constituting vibra- 
 tion of the vocal cords. The frequency of the vibration depends 
 upon the tension of the vocal cords. The more tightly they are 
 drawn, the more rapid will be the vibration produced. 
 
 With slow vibrations, distind puffs of air can be heard, pro- 
 ducing a sound known as "throat trill." If the glottis is opened 
 and closed more than thirty-two times in one second, the ear fails 
 to distinguish the individual puffs, and perceives only a continuous 
 effect of a musical charader denominated "voice." The pitch of 
 the voice rises as the vibrations become more rapid. 
 
 There are two ways of varying the pitch of the voice, just as 
 there are two ways of changing the pitch of a violin string. Ob- 
 serve the violinist tuning his instrument. He turns a peg at the 
 end, thus tightening the string. At each increase of tension the 
 pitch of the string becomes higher. When, however, he plays the 
 instrument, the pitch is varied in a different manner. He presses 
 his finger upon a string so as to permit only a portion of the string 
 to vibrate, instead of the whole — and the pitch becomes higher. 
 The shorter the vibrating portion, the higher is the pitch produced. 
 In this case the tension of the string remains uniform. 
 
 In a similar manner variations of pitch in the voice may be pro- 
 duced by allowing a portion only of the vocal cords to vibrate, 
 instead of the whole. Observations made by means of the laryn- 
 goscope, seem to indicate that the variations of pitch, in what is 
 termed the "head register" of the voice, are produced in this 
 manner; whereas, in the "chest register," the vocal cords vibrate
 
 as wholes, and the changes of pitch are produced by variations of 
 tension. 
 
 Vocal cords vibrating in part. Vocal cords vibrating as wholes.* 
 
 Dr. Hewson. during the course of his lecture, gave utterance to 
 rather a startling remark. If I understood him correftly, he ex- 
 pressed the opinion that some of the vowel sounds are formed in 
 the larynx, and not in the mouth. I cannot agree with him in this 
 opinion, although I am aware that he can quote authorities in sup- 
 port of his position. For example: Dr. E. M. Moore, of Rochester, 
 N. Y. . has published an account of the following case:f Some 
 years ago Dr. Moore attended a man who had attempted suicide by 
 cutting his throat. The cut was immediately above the thyroid car- 
 tilage, shaving off the epiglottis at its base. The wound resulted in 
 an oval opening, two inches long by three quarters of an inch wide. 
 The man was able to talk at any time bv bending his head forward 
 and temporarily closmg the opening. When the head was thrown 
 back he lost the power, but Dr. Moore noticed that under such 
 circumstances he could pronounce vowel sounds like ah, oh, etc. 
 The dod:or was surprised at the clearness and distinftness of the 
 vowel effeds, for the sounds seemed to emanate from the yawning 
 wound in the throat, and not from the mouth. Struck by this cir- 
 cumstance, the worthy doftor made a unique experiment. He 
 introduced into the wound a sheet of buckskin, so placed as to 
 prevent the possibility of any air passing from the larynx into the 
 mouth. The only outlet left for the breath, was the yawning 
 wound. He then asked the man to repeat the alphabet. A, B, C, etc. 
 Dr. Moore found that under these circumstances, certain vowel 
 sounds could be distinguished, and he came to the conclusion that 
 these voweic were formed in the larynx, and not in the mouth. 
 
 This conclusion would be more reliable if he had shown that 
 the man could pronounce these vowels with his mouth shut. Of 
 course, if the vowels heard were really produced in the larynx 
 alone, the closure of the lips would have made no difference in the 
 
 * These cuts are reproduced from " Voice, Song, and Speech." 
 f See Transactions of the New York State Medical Society, 1872, pp. 276-282.
 
 8 
 
 effeft. Whereas, if the ordinary theory is correct, that vowel qual- 
 ity is due to the resonance of the mouth cavities, then the closure of 
 the mouth passage, at both ends, would have been fatal to the 
 effec^t. 
 
 1 understand that during the course of the experiment, the man 
 moved his mouth as though he were speaking; and Dr. Moore simply 
 took precautions to prevent the passage of air through the mouth. 
 The mouth positions for the vowels were, therefore, assumed, and 
 the mouth would then, in effe6t, be a resonator, tuned to the vowel 
 positions, held near a sounding body — the larynx. Under such cir- 
 cumstances, resonance effefts should be produced without the 
 aftual passage of air through the mouth, just as an ordinary resona- 
 tor, when properly tuned, becomes sonorous when held near a 
 vibrating tuning fork. 
 
 I cannot look upon the experiment as by any means conclusive; 
 and 1 know of no other fa6ts to support the hypothesis that any of 
 the vowels are formed in the larynx, independently of the mouth. 
 
 A number of years ago Dr. McKendrick, of Glasgow University, 
 afforded me an opportunity of examining the speech of a man 
 whose larynx had been excised. The patient had been supplied 
 with an artificial substitute for the larynx, made, 1 think, of dentist's 
 rubber. As there were no vocal cords, the man could only speak 
 in a sort of whisper, which was barely audible. A small aperture 
 had been left in the front part of the rubber substitute into which the 
 man slipped a metal reed, taken from a harmonium or small parlor 
 organ. Upon then attempting to speak, the reed was thrown into 
 vibration by the air from the lungs, and a good sonorous voice resulted, 
 — resembling the natural voice to a remarkable and startling degree. 
 The patient spoke with a broad Scotch accent, and his articulation 
 was simply perfe6t. Very little peculiarity could be dete6ted in the 
 artificial voice excepting that it was monotonous and without inflec- 
 tion. The speech was so natural in quality that it was difficult to 
 realize that the source of sound was a metal reed inserted into the 
 throat. 
 
 The man was a machinist by trade, and he employed his spare 
 time in manufacturing reeds for himself. He had quite an assortment 
 for experimental purposes, and he let me hear the efife6l of reeds of 
 various sizes and materials. He could change his voice from bass to 
 tenor, and from tenor to soprano at will, by employing suitable 
 reeds. 
 
 The point to which 1 would direct your attention is this : — that the
 
 vowels were all perfectly produced, although the larynx had been 
 excised. 
 
 We have seen that a metal reed can be used in place of the 
 vocal cords; and I am inclined to think that the real larynx, if 
 detached from the body and operated by means of a wind chest or 
 organ bellows, would produce an effeft more resembling the sound 
 of a beating reed, than the human voice. The quality or "timbre" of 
 the human voice, I believe, is due in a very minor degree to the 
 vocal cords, and in a much greater degree, to the shapes of the 
 passages through which the vibrating column of air is passed. 
 As the shape of the passage above the vocal cords controls the qual- 
 ity or timbre of the voice, we may be sure that the false vocal cords 
 exert some influence upon the quality of the voice especially if they 
 are capable of approximation, a point 1 am unable to decide. The 
 ventricles, also, the spaces between the true and false vocal cords 
 on either side, should, theoretically, exert an intluence upon the 
 quality of the voice, for they constitute two small resonance-cham- 
 bers, situated close to the source of sound. In the howling monkey 
 the ventricles are expanded into pouches, and the chara6teristic 
 howl produced by the creature is due to the resonance of air in those 
 chambers. 
 
 In the case of the Scotchman at the Glasgow University, the 
 pitch of the artificial voice produced was undoubtedly due to the 
 reed employed, but the quality of the voice, and the consonant and 
 vowel effects were due to the passages above, through which the 
 vibrating column of air was passed. 
 
 I have already directed your attention to the case reported by 
 Dr. Moore of Rochester, New York, in which he claimed that 
 certain vowels are formed in the larynx and not in the mouth. Dr. 
 Moore directed my attention to the case of another patient of his 
 which seems to prove the converse proposition. * Dr. Moore had 
 performed upon this man the operation of tracheotomy. At the 
 time I saw the patient he had for over twenty years been dependent 
 for life upon air supplied through a silver tube inserted in the 
 trachea. The glottis had become completely closed and no air 
 could be forced through the larynx into the mouth. The strange 
 feature of the case was that under these circumstances the man 
 could talk. (Jf course the speech was peculiar on account of 
 
 *Case of Edward Matthewb. See Transactions o| the New York State Medu al 
 
 Society, 1872 pp. 2-;i)-z>^2.
 
 10 
 
 the absence of voice, but there was no dfficulty in understanding 
 it. In this case the air which was moulded into speech came not 
 from the lungs, but from the pharynx. If we close the lips tightly 
 and compress the air in the mouth and make an effort to blow, the 
 pharynx expands under the pressure of the confined air, just as a 
 rubber ball would expand if you were to blow into it. Upon 
 opening the lips the contra(5tion of the pharynx causes a sudden 
 puff of air. A puff of air of this charader can be produced even 
 though the glottis is closed. By long practice this man had 
 acquired such expertness in the use of the muscles about the 
 pharynx that he could produce explosive effects of this kind which 
 could be distinctly heard at a distance. When he spoke, the conso- 
 nants were formed with very great firmness, and the removal of the 
 consonant position resulted m a puff of air through the vowel posi- 
 tions assumed by the mouth, so that the puff had vowel quality 
 sounding like a whispered vowel. I can imitate the character of his 
 speech so that you may understand, more clearly than I can describe 
 it, the nature of his articulation. 
 
 The point that I would have you observe is, that in this case 
 vowel effeds were unmistakably produced, although no air passed 
 through the larynx into the mouth. 
 
 Dr. Hewson: Do I understand you to say that this individ- 
 ual who had the larynx closed, made audible speech ? 
 
 Dr. Bell: Yes, sir; quite audible speech. 
 
 Dr. Hewson: I merely wish to make a statement from 
 actual surgical experience that, in cases where the operation of 
 tracheotomy has been performed, it is impossible to make any sound 
 under such circumstances, unless the finger is placed over the tube 
 in the trachea. Now, of course, that may be due to the condition 
 which you have already indicated ; that is, that the speech was pro- 
 duced by this individual by long praftice. Now it is possible for 
 people to speak when that tube is closed, even though the principal 
 part of the vocal apparatus is lined with diphtheric membrane. 
 There is some sort of sound made, at least, but there is no audible 
 speech. The remark I made some time ago about the vowel sounds 
 being made in the larynx, I will amend by saying that of course 
 the experiments that you have had the opportunity of seeing, had 
 not come to the notice of either myself or the gentlemen who have 
 preceded me, in making statements in text-books from which I have 
 quoted. Your experience, I believe, is very unique; because I have 
 never heard of any one who had been able to see the experiments,
 
 II 
 
 made as you have detailed them to-day. With reference to the 
 closure of the upper part of the vocal apparatus by diphtheric mem- 
 brane, and the introduction of a tube into the trachea ; I may say 
 that in order that any sound may be produced at all, the tracheal 
 tube must be closed. Now, if the parts above, are almost closed 
 by the diphtheric membrane, no sound is produced. However, I 
 can readily see, from a knowledge of the muscular tissue surround- 
 ing the parts, that such sounds could be produced as you have 
 detailed in this individual. 
 
 Dr. Bell: In the case of this individual I have just men- 
 tioned, no air could pass up into the mouth, under any circum- 
 stances. The aperture in the windpipe remained open, and, all 
 the time he was speaking, air gushed out of the tube in his throat, 
 forming a whistling accompaniment to his speech. 
 
 Miss Yale: Dr. Bell, I have a number of questions here for 
 you to answer. 
 
 Dr. Bell: The first question is: "Is it possible to constrict 
 the false vocal cords ? " I think that Dr. Hewson will be more 
 competent than I am to answer that question. 
 
 Dr. Hewson: I cannot conceive of any muscular fibres con- 
 striding the ventricular bands, or false vocal cords. 
 
 Mr. Crouter: How was it with the Scotchman, Dr. Bell? 
 
 Dr. Bf.ll: There were no vocal cords in the case of the 
 Scotchman, the vocal cords were represented by a harmonium reed. 
 
 i nardly know how to commence on the stream of questions 
 you have set for me. They appear to be of very great impor- 
 tance, and I should be very glad if I can be of any assistance to 
 teachers here in answering them. I may, perhaps, group them so 
 as to answer two or three at one time. 
 
 (i) " Please illustrate the development of Jig." 
 
 (2) "Dr. Bell develops non-vocal r from ///; please demon- 
 strate." 
 
 (3) " How would you develop eh ?" 
 
 I will take up these in one group. In difficult cases you will 
 find manipulation of the tongue of great assistance: and I think that 
 this series of questions may be answered by showing you how to 
 manipulate the tongue; and by directing your attention to the 
 nature of the changes, you can produce by manipulation. You can 
 push a position further back, and you can enlarge an aperture by 
 manipulation ; but you cannot do the converse, if, then, your pu- 
 pil cannot pronounce a given sound, let him give a sound of similar
 
 12 
 
 formation but further forward in the mouth ; then push the position 
 back. For example; take the cases specified in the questions, (i.) 
 The pupil cannot pronounce iig. Now, suppose he can pronounce 
 71 — a sound of similar formation but formed further forward in the 
 mouth. Take a manipulator (for example; a paper cutter), and 
 hold it in the pupil's mouth so as to cover the top or front part of 
 the tongue. Now tell the pupil to say n. The point and front of 
 the tongue being fettered, the back of the tongue alone is free to 
 'rise; and the attempt to say // results in iig. Now give the pupil 
 a hand-mirror and tell him to keep his tongue still when you with- 
 draw the manipulator. In most cases the pupil is at once able to 
 pronounce iig, but in difficult cases it is advisable to get him to 
 manipulate his tongue for himself, watching the effed in a hand- 
 mirror. 
 
 (2) Your pupil cannot pronounce non-vocal r, but is able to 
 give th. Place the manipulator under the tongue, and gradually 
 lift the point of the tongue, while the pupil tries to sound ///. The 
 sound changes to a hiss, somewhat like s, then as the tongue is 
 raised higher the sound becomes more like sh; and if lifted still 
 higher, it becomes non-vocal r. When the correft position is 
 reached, the point of the tongue is against the inner part of the 
 upper gum— just where the palate commences to arch — with an 
 aperture over the centre. 
 
 (3) The pupil cannot pronounce sh. Let him pronounce th. 
 Place the manipulator over the tip of the tongue and push the 
 tongue gradually back. The sound changes first to s, and then to 
 sh. If pushed too far back, it becomesj-/^ (Ji in the word hue) ; and 
 if still further back, the German ch (back-centre aperture). 
 
 Another question is, "What would you do with a pupil who 
 gives tig too far back ? " I don't quite understand this question, for 
 you can't get w^.too far back. You can get k and g too far back, 
 and these are very common defeds. In such cases the back of the 
 tongue is placed against the back of the pharynx instead of against 
 the soft palate, but such a position cannot produce an ng sound, be- 
 cause the depression of the soft palate would not admit air into the 
 nasal passages, the shut position being below the soft palate. It is 
 very difficult to corred a position that is too far back. You can push 
 a position further back, but you can't pull it forward! I think the 
 best plan is to start anew. Take a position too far forward in the 
 mouth, and push the tongue backwards to the corred position. For 
 example; where you have k too far back, take /; let the pupil try to
 
 13 
 
 pronounce /, while you hold the manipulator overthe tongue so as to 
 prevent any portion from rising, excepting the back. In this way 
 you will be sure of a good k, if the pupil does not know what you 
 are aiming at. By vocalizing the k you get^. (Dr. Bell illustrated 
 his remarks by manipulating the tongue of a deaf pupil.) 
 
 The next question is, "Please demonstrate that intelligible 
 speech does not depend upon perfed vowel positions." I shall 
 read a few sentences from a book, substituting for each vowel 
 sound a mere indefinite murmur of voice. You observe that the 
 articulation, though, of course, very peculiar, is perfedly intelli- 
 gible.* We may learn from such an experiment as this, that con- 
 sonants are much more important elements than vowels. Intelli- 
 gibility of speech mainly depends upon the correft pronunciation 
 of consonants. We could manage to get along very well with only 
 one vowel sound, if indefinite enough, and yet make ourselves 
 understood. I don't mean to advise you to teach speech of this 
 character to your pupils, but many of you may be encouraged to 
 know that very imperfed vowel sounds will not prevent your pupils 
 from being understood by relatives and friends. Consonants are 
 much more easily acquired than vowels, and all pupils who can 
 pronounce the consonants corredly can acquire a useful articulation, 
 even though they murder the vowels. That is, their speech will 
 be intelligible to hearing people, and therefore useful as a means of 
 communication, even though it may not be very pleasant to hear. 
 Too much effort, I think, is made to impart a niceness of pro- 
 nunciation that is not appreciated by the outside world. 
 
 When 1 first entered upon the work of articulation teaching, I 
 was very proud of the pronunciation of some of my congenitally 
 deaf pupils. They had been drilled upon the elements and were 
 able to pronunce words and sentences written in Visible Speech 
 with absolute corredness, slowly, it is true, but with perfed ele- 
 mentary sounds. To my great mortification, however, I found that 
 
 * Read the following passage aloud, giving an indefinite murmur of the voice for 
 each dash, and the passage will be intelligible: — 
 
 - p-nt-d t- th- c-t -nd th-n t- -ts n-m -nd -nd-v-id t- m-k h-m 
 -n-rst-nd th- m-n-ng -v th- r-t-ng - -Is- t-t h-m t- sp-1 th- w-rd -n 
 h-s f-ngg-rs. -v-th-ng th-t h- d-d w-s p-rf-rmd w-th - p-nd-r-s s-rt 
 -V -mf-s-s th-t w-d h-v m-d - p-rs-n -nf-m-ly-r w-th th- d-f s-p-z 
 tli-t h- n- -1 -b-t -t — From the Aiiiuh for Jantury, i8()i,p. 4=;. 
 
 You may substitute for the (-) any large aperture vowel, such as u in up, er in 
 her, in on, or even a in cjI, without destroying intelligibility.
 
 14 
 
 visitors generally preferred the imperfect gabble of some semi-mute 
 to the elocutionary speech I had labored to impart. In those days 
 1 had a very poor opinion of the visitors, but 1 have since come to 
 the conclusion that 1 myself was somewhat at fault. My aim v/as 
 wrong. At that time I had the mistaken idea that it was hardly 
 worth while teaching articulation at all, unless I could teach 
 my pupils to speak -ucell. My ear was sensitive to every mispro- 
 nunciation, and 1 was constantly correcting my pupils for errors 
 that were not even noticed by visitors to the school. 
 
 Our obje(ft is not to train elocutionists, but to help deaf children 
 to make themselves understood. Intelligibility is of far greater im- 
 portance than perfection. By all means let us get as perfect a pro- 
 nunciation as we can, but do not let us spend much time on minor 
 details of pronunciation that are of no consequence to intelligi- 
 bility. Let us aim, not to get the speech of the elocutionist, hut the 
 speech of the people among ivhom the children live. Judged by the 
 standard of the elocutionist, how many of the hearing people of the 
 world talk well } Very few, and yet their speech is of inestimable 
 value to them in all the relations of life. 
 
 Now, I hardly think it worth while attempting to teach the mass 
 of the deaf to talk better than other people. Let us be satisfied with 
 speech like that which we ourselves employ without troubling our- 
 selves about niceties of articulation that might be all very well in an 
 oraior or public speaker, but which would be of no pradlical benefit 
 to a deaf child in his own home. Observe your own utterance and the 
 utterances of your friends, and you will find that half your syllables 
 are pronounced in a slipshod fashion that would make an articulation 
 teacher shudder if given in the same manner by a deaf child. Take 
 such little words as and or of. Who gives them their proper 
 didionary pronunciation in adual speech ? No one but the deaf 
 child. We pronounce them somewhat as though they were 
 spelled und or uv. You und I ; a cup uv tea. No one says thee 
 boy, Ann apple, or eh pear in conversation. The obscure sound 
 is heard, and our utterance is more like thu boy, iin apple, ti pear. 
 The same indefiniteness of vowel sound is charaderistic of all our 
 unaccented syllables. Why, then, need we be particular about the 
 pronunciation of the unaccented syllables in such words as, com- 
 fortable (kumf-t-ble), lesson (les-n), silent (sil-nt), sentence (sent- 
 nce), workman (workm-n), different (dif-r-nt), above (-buv), forgot 
 (f-got), etc. Any sort of indefinite vowel sound will pass muster 
 in these syllables if only softly uttered.
 
 15 
 
 We are training our children to talk to ordinary people, not to 
 elocutionists. There are certain points that must be attended to in 
 order that the speech may be satisfadory to ordinary people; but in 
 regard to other points, great latitude may be allowed. 
 
 Consonant elements and the vowels in accented syllables must 
 be properly pronounced, but the vowels in unaccented syllables 
 may be uttered in any sort of indefinite way without offending the 
 ordinary ear. Accent and rhythm, I think, are of more importance 
 than exact pronunciation. 
 
 In Visible Speech the voice symbol (i) is used to indicate an in- 
 definite vowel sound like er in the word her, or like the er sound 
 used by public speakers to fill up gaps in their sentences— when— er 
 —they are not — er — very— er—er— sure— er— what they want to say. 
 This indefinite vowel sign I consider a perfed God-send to the 
 teacher of articulation, enabling him to get rid of half the labor of 
 articulation teaching. In spelling phonetically the vast majority of 
 the vowel sounds in the unaccented syllables may be represented by 
 this indefinite voice mark; and it may also be substituted every- 
 where for glide r. I would recommend those who do not use 
 Visible Speech to use a dash. 
 
 Ordinary people who know nothing of phonetics or elocution 
 have difificulty in understanding slow speech composed of perfe6l 
 elementary sounds, while they have no difficulty in comprehending 
 an imperfed gabble if only the accent and ihythm are natural. Too 
 much labor is bestowed upon unaccented syllables. Any child can 
 give an indefinite vowel sound that may be combined rapidly with 
 consonants. I have seen a teacher puzzling herself over the word 
 comjot table. The first syllable gave no difficulty, but the second 
 syllable was not pronounced easily by the child. There was an 
 attempt on the part of the pupil to give long 6, as in pole, followed 
 by a consonant r of an exaggerated kind, and the teacher was try- 
 ing to get the pupil to substitute aw for o. With regard to the last 
 syllable the teacher was puzzled to decide whether the vowel should 
 have the sound of a in table, a in cat, or a in ask. Now the fad; is 
 that the exact vowel sounds in the unaccented syllables are of no 
 earthly consequence. An ordinary ear will accept any sort of inde- 
 finite sound as good speech, if the word is uttered rapidly with due 
 accent on the first syllable. Even an elocutionist would not ask for 
 a consonant r in the second syllable. He would simply demand a 
 gliding of the tongue towards the position for r; but ordinary peo-
 
 16 
 
 pie do not give an r of any sort in a large proportion of cases, and 
 it is certainly the case that glide r may be ignored in teaching a deaf 
 child and taught simply as voice, without any ordinary person notic- 
 ing any peculiarity, so that in the word "comfortable" the second 
 syllable may be given as, / followed by the indefinite voice sign. 
 (3i, /-) So also with the vowel in the last syllable. The whole 
 word, therefore, might be written a]B3iDi0CO or kumf-t-bl. Pro- 
 nounce the first syllable with due deliberation and care, and give 
 tbe others rapidly and carelessly, and it will be satisfadory. 
 
 Dr. Williams : With indefinite vowel sounds, is there not a 
 danger of carrying that too far so as to get indefiniteness .? 
 
 Dr. Bell: You don't want to do that with accented vowels. 
 However, it is the consonants that give intelligibility to speech. 
 You may give every vowel indefinitely; but if the consonants are 
 definite, you get intelligibility. If the accented vowels are given 
 corredly, the unaccented vowels may be jumbled up somewhat. 
 
 Dr. Williams : But if they get into the habit of giving the 
 vowel sound in that indefinite way, won't they carry it into the 
 accented vowels } 
 
 Dr. Bell ; Well, they might, just as we do. Of course, the 
 better articulation you can get the better; but allow me to say that 
 in ninety-nine out of a hundred persons you meet in ordinary 
 society, that indefinite sound is carried into a great many accented 
 syllables and also into nearly all the unaccented syllables. 
 
 The next question is, " In a whisper, are the vocal cords lax 
 or tense .? " In the case of a whisper there is a constridion ; there is 
 an obstruction to the passage of air ir the glottis. The vocal cords 
 are not adjusted so as to permit of a definite musical vibration, so 
 that the obstruction results in a rustling sound that we term 
 "whisper." 
 
 Prof. Binner : You mean constriftion of the vocal cords ? 
 
 Dr. Bell : Yes, a constridion of the glottis — the space be- 
 tween the vocal cords.
 
 THE PHARYNX AND MOUTH IN THEIR 
 RELATION TO SPEECH. 
 
 In my last lecture I told you about a man with a harmonium 
 reed in his throat, in place of vocal cords. Now, ordinarily, there 
 is vast deal of difference between the sound of a harmonium, and 
 the sound of the human voice, and yet in this case the reed pro- 
 duced the effed of a human voice when the man spoke. To the 
 ear, therefore, it made all the difference in the world, whether the 
 reed was vibrated outside or inside the man's throat. Now, we 
 have no reason to suppose that the thorax and lungs operated in 
 any different way from the wind chest of a harmonium. They 
 simply supplied air to set the reed in vibration. The difference oi 
 effed, therefore, must have been due to the parts above the reed. 
 In other words, the pharynx, mouth, etc., were the agencies 
 involved in changing the harmonium effe(ft into a human voice. 
 Consider for a moment, the nature of the difference between the 
 sound of a harmonium reed and a sound of similar pitch sung by 
 the voice. The same note may be played upon a piano, a violin, 
 a flute, or a trumpet, and yet each sound has an individuality of its 
 own. We can tell by the ear, at once, which instrument is used, 
 although all the notes may be alike in pitch, and equally loud. The 
 sounds differ from one another in "character," "quality" or "tim- 
 bre," and it will thus be understood that the pharynx, mouth, etc., 
 affe(fl the quality or timbre of the voice. 
 
 We can recognize that every sound possesses the elements of 
 pitch, loudness, and quality. It matters not whether the sound be 
 produced by the human voice, by a musical instrument, by the 
 rustling of leaves, or by a knock upon the door — it has a certain 
 
 >7
 
 i8 
 
 loudness, a certain pitch and a certain character, or timbre of its 
 own, by which we recognize it from other sounds of similar 
 pitch and loudness. 
 
 Now, when we study the production of voice, we find that 
 these three characteristics originate principally in three different 
 parts of the vocal apparatus. 
 
 (i) The pitch of the voice is determined by the 
 vocal cords. 
 
 (2) The loudness by the abdominal or expira- 
 
 tory muscles; and 
 
 (3) The quality or timbre by the parts above 
 
 the vocal cords. 
 
 1. The lips of the glottis open and close with great rapidity, 
 and the frequency of the vibration is mainly determined by the 
 tension of the vocal cords. 
 
 2. Air escapes from the lungs through this vibrating glottis in 
 a series of puffs, and the force of emission is chiefly determined by 
 the action of the abdominal or expiratory muscles. 
 
 3. The upper part of the larnyx, together with the pharynx, 
 nares, and mouth, constitutes a passage-way, or tube, of variable 
 size and shape, through which the vibrating current of air is passed. 
 It is here that the voice is moulded, so to speak, on its way to the 
 ear, and the shape of the passage-way largely determines the 
 quality or timbre of the voice. 
 
 You can produce a crude voice-like sound by the vibration of 
 the lips of the mouth. Press your lips very firmly together while 
 you blow air between them, so as to cause the edges to vibrate. 
 The sound produced is not very pleasant, and resembles more than 
 anything else the hum of a bee, or the buzz of an imprisoned fly. 
 But place the buzzing lips at the end of a tube — for example a 
 trumpet — and at once the quality changes. Out come the clear 
 ringing tones so familiar to us in a brass band! In this case the 
 source of sound is found in the vibration of the lips, but the timbre 
 or metallic quality is due to the trumpet. 
 
 In a somewhat similar manner the passage-way or tube, 
 through which the voice is passed, affeds the quality of the sound 
 produced by the lips of the throat; and if we could decapitate a 
 singer in the midst of a song, so as to hear the sound produced by 
 his vocal cords alone, I fancy we should find as great a change in
 
 19 
 
 the quality of the voice, as we do in the sound produced by the 
 lips when the trumpet is removed. The beauty of the voice would 
 be gone, and you would simply have a reed-like effeft. 
 
 In the case of the trumpet, the chara6ter of the tube affeds not 
 only the quality, but the pit:h of the sound produced. For example: 
 If you lengthen the tube, the lips vibrate more slowly, and the 
 sound becomes lower in ]. itch. In the instrument of speech, how- 
 ever, the lips of the glottis are so admirably adapted for independent 
 vibration, that changes in the passage-way do not affecl: their rate 
 of vibration, but simply chinge the quality of the resulting sound. 
 
 A number of years ago I visited a large school for the deaf, and 
 taught all the pupils to use their voices. In a few cases the effe6l 
 was decidedly unpleasant, the voice resembling somewhat the cry 
 of a peacock. The effect, indeed, was so unnatural and distressing 
 to the ear that some of the teachers expressed the opinion that the 
 vocal cords had been affected by the disease that had caused deaf- 
 ness. They thought, therefore, that it would hardly be worth 
 while attempting to teach these children to speak. Knowing 
 that the quality of the voice is chiefly determined by the shape of 
 the passage through which it is passed, I did not consider it 
 necessary to assume a defeat in the vocal cords, but rather sought 
 the cause of the peculiartiy in some constriction of the passage-way 
 higher up than the vocal cords. 
 
 I was careful to avoid discouraging the pupils by any expres- 
 sion of disapproval, so they were entirely unconscious of the fa6t 
 that their voices were unpleasant. They had no hesitation, there- 
 fore, in repeating the disagreeable sound as often as I desired ; and 
 I encouraged them to repeat it a great many times, so that I might 
 study the effeft and become familiar with the sound. I then found 
 it possible to imitate the effedi myself. This was proof positive 
 that the existence of the peculiarity was quite consistent with the 
 possession of perfeft vocal organs. Having acquired the ability to 
 repeat the effect, I set myself to work to find out what I did with 
 my mouth during the produdion of the sound. I could feel a con- 
 striction somewhere in the back part of the mouth, and therefore 
 examined my vocal organs in a hand mirror while I depressed the 
 tongue so as to exhibit the whole of the pharynx. At once the 
 cause of the peculiarity became manifest. The muscles constitut- 
 ing the side walls of the pharynx were seen to be forcibly con- 
 tracted, and they were approximated so closely together as almost 
 to touch. After a little pradice I found myself able to move these
 
 20 
 
 muscles at will without making any sound. Then I tested the 
 effed of the motion upon the quality of the voice. When the 
 muscles were relaxed and the cavity of the pharynx expanded the 
 quality of the voice was good, but the moment the side walls of 
 the pharynx commenced to approach one another (see dotted lines 
 in Fig. 2.), the character of the voice changed. It acquired a pecul- 
 
 Fig. 2.* 
 
 1. Soft palate. 2. Uvula. 3, 4. Anterior pillars of the soft palate. 5, 6. Pos- 
 terior pillars of the soft palate. 7,8. Tonsils. 9. Tongue. 10. Back of the pharynx. 
 
 The posterior pillars of the soft palate (5, 6,) are capable of approximation, as 
 shown by dotted lines. 
 
 iar metallic ring, somewhat like the tone of a brass musical 
 instrument. The effeft became more and more disagreeable as 
 the side walls approached, until the peculiarly distressing effect was 
 produced, which 1 have likened to the cry of a peacock. Having 
 gained this information I attempted to improve the voices of the 
 children. For this purpose I gave them hand mirrors and taught 
 them to depress their tongues so as to render visible the soft palate 
 and back of the pharynx. I then made them look into my mouth 
 while I silently contracted and expanded the pharvnx. After some 
 practice they were able to imitate the adion. 
 
 1 then placed my hands on my throat while 1 repeated the 
 exercise with voice. Their first attempts at reproduftion were 
 failures; the moment they sounded the voice,, a powerful contraction 
 
 ■Reproduced from " Voice, Song, and Speech.'
 
 .21 
 
 of the muscles about the pharynx became visible, and the usual 
 disagreeable effe6l was produced. By means of the mirror, I 
 direded their attention to the constri(ftion, and told them to expand 
 the pharynx, as they had done before when they made no noise. 
 At first they were unable to relax the muscles of the pharynx, with- 
 out stopping the voice, but, after some praftice, they succeeded in 
 doing this, and at once the voice became natural and pleasant in 
 quality. 
 
 The cavity of the pharynx may be roughly likened to a room 
 with four walls. The back part ot the tongue constitutes the front 
 wall of the chamber, and opposite to it is the back wall of the 
 pharynx. The side walls are formed by muscles that extend upwards 
 to the soft palate. The approximation of these side walls, as 1 have 
 already explained, imparts to the voice a disagreeable metallic 
 quality. The front and back walls, too, are capable of approxima- 
 tion, and in this case, also, the quality of the voice is injuriously 
 affe(5ted. For example: The tongue may be held so far back in the 
 mouth as to cause the base of the tongue to come almost into 
 contad with the back of the pharynx. The voice then acquires a 
 peculiar " guttural quality." I have heard this kind of voice pro- 
 duced by deaf children, but it is more common, I think, among 
 persons who hear. It is rarely heard during the aft of speech, hut 
 many persons affe6l this guttural quality of the voice when they 
 sing. The "metallic quality " of voice, on the other hand, is quite 
 common among the deaf, although it is rarely so marked as to be 
 painful to the ear. Many hearing persons also possess it in a greater 
 or less degree, — especially persons who use their voices much in 
 the open air. For example, the rasping voice of the street hawker 
 is of this description. 
 
 Another peculiarity of voice very common among the deaf, is 
 "nasal quality." This is occasioned by the habitual depression of 
 the soft palate. By means of a hand mirror, the cause may be 
 shown to a deaf child. 
 
 The soft palate is capable of elevation and depression. When 
 it is raised it fits closely against the back of the pharynx, forming a 
 ceiling to the pharyngeal cavity. When it is depressed, it hangs 
 down like a curtain, leaving a passage-way behind it, which leads 
 into the nares or nasal passages. 
 
 1 would recommend every teacher of articulation to learn to 
 control the movements of the soft palate and the muscles of the 
 pharynx, so as to be able to exhibit the acftion of the parts to pupils.
 
 22 
 
 The first point you have to learn, is to depress the tongue so as to 
 unveil the pharynx and soft palate. Many persons find difficulty in 
 doing this, but by persistent efforts before a mirror, all can acquire 
 the power. Now watch the soft palate while you breathe gently, 
 sometimes through the mouth, sometimes through the nose. At 
 first the soft palate appears to move about in a most mysterious man- 
 ner by itself, without any volition on your part. Now it goes up, 
 and then the next moment you see it hanging loosely down. By 
 watching these motions in a mirror, and attempting to control them, 
 you will soon find yourself able to elevate or depress the palate at 
 will. Now sound the voice continuously, and observe what effe£t 
 is produced upon its quality by the movement. You will notice 
 that the moment the palate falls, the voice acquires nasality, and 
 that this effed disappears when the palate is raised into contaft with 
 the back of the pharynx. 
 
 During the ad of speech, the soft palate is raised continuously, 
 excepting when the sound of w, ;/, and ng, are uttered. In order 
 to corred a nasal quality of voice, therefore, your pupil must raise 
 his soft palate. The question arises, however, — how are you going 
 to make him do it } 
 
 Various expedients may be resorted to, such as the common 
 one of telling him to blow an imaginary feather away from his 
 mouth while he speaks ; but these are all indirect methods, and do 
 not touch the root of the matter. I would recommend you to go for 
 the soft palate itself, direftly, with a hand-mirror. Teach your pupil 
 to elevate and depress it at will. Direft the adion with your hand. 
 
 When you raise your hand let him raise the palate (Cut A), and 
 keep it elevated till you give the signal for depression (Cut B). 
 
 Ill ll/l- lii' 1 
 (Cut B.)* 
 
 *These cuts are reproduced from " Voice, Song, and Speech."
 
 2 3 
 
 Then let him keep it depressed without motion, till you direct him 
 to raise it. Control over the vocal organs is gained not so much by 
 moving them as by keeping them still. Keep the soft palate de- 
 pressed and still for a long period of time, and then raised for an 
 equal length of time. Do this at first silently, and then afterwards 
 with voice. Elevate and depress the palate without stopping the 
 voice but retain the elevated or depressed position for a considerable 
 period of time. 
 
 When your pupil can do this without looking in the mirror you 
 may usefully vary the exercise by requiring him to raise or depress 
 the palate while at the same time he prolongs a vowel sound. (For 
 example: ah or ee or o.) Then let him rattle off a series of vowels 
 without stopping the voice. (For example: ah ee o ee; ah ee o ee, 
 etc.), elevating or depressing the palate as you dired. As the ulti- 
 mate objeft to be gained is ability to retain the soft palate in the 
 elevated position continuously during speech, there should be no 
 rapid alternations of elevation and depression. He should repeat 
 the series of vowels many times in succession with the soft palate 
 raised, and many times with it depressed, but the voice should not 
 be stopped excepting when it becomes advisable to take breath. • 
 
 The uvula, the pendulous extremity of the soft palate, seems 
 to have no special fundion in speech, at least in the English lan- 
 guage, and I have known of cases where it has been excised 
 without interfering with articulation. In teaching the deaf, however, 
 the uvula may be found of use as an index to the pitch of the voice. 
 A pupil may, perhaps, be made conscious of changes in the pitch 
 of the voice, by directing his attention to changes that simultan- 
 eously occur in the length of the uvula, in most cases, the uvula 
 hangs loosely down during the production of low tones and shrinks 
 in size as the pitch of the voice is raised, (See Cuts A, and B.) 
 When the pitch is very high the uvula shrinks up to such an extent 
 that it almost disappears (Cut C, page 24.) While this rule is not 
 invariable, the effed is so commonly produced, as at least to be 
 worthy of note. 
 
 Every change in the shape of the passage-way, through which 
 the voice is passed, occasions a corresponding change in the quality 
 of the voice, and I have pointed out the causes of certain disagree- 
 able effeds. in order to render the voice sweet and pleasant to the 
 ear, it is necessary that the soft palate should be raised into contad 
 with the back of the pharynx, and that the whole cavity of the
 
 24 
 
 pharynx should be expanded, so that the passage-way there should 
 be free and unobstructed. Any constriftion in the pharynx is fatal 
 to the beauty of the voice. 
 
 The mouth passage also affefts the voice, imparting to it 
 "vowel quality," and changes in the shape of the mouth-passage, 
 produced by the adion of the tongue and lips, occasion changes Of 
 vowel quality. In singing different vowel sounds the voice may have 
 the same pitch and loudness, and yet each vowel remains distinft to 
 the ear. Vowel differences, therefore, are differences in the quality or 
 timbre of the voice ; and vowels themselves are in reality qualities 
 of voice to which we have given specific names, and which we 
 employ as elements of speech. 
 
 I do not propose to-day to enter into any detailed description 
 of the positions assumed by the tongue and lips, during the produc- 
 tion of vowel sounds, as most of you, I know, are familiar with the 
 
 [III.' Il/it 
 (Cut C).* 
 
 subjeft. I shall rather attempt to show you why it is that changei; 
 in the shape of the cavities of the mouth, pharynx, etc., occasion 
 changes in the quality of the voice. 
 
 When we prolong a vowel sound without varying the pitch of 
 the voice, the effeft produced upon the ear is not simply that of a 
 single musical sound, but of a chord containing a number of musical 
 tones of different pitch. One of these tones is so much louder than 
 the others that it determines the apparent pitch of the whole combi- 
 nation. The other tones are so feebly produced, that it takes a 
 skilled ear to recognize them as musical effects at all; and the 
 untrained ear simply perceives them as the quality or timbre of the 
 sound. When a number of vowels are sung successively without 
 varying the pitch of the voice, a trained ear readily perceives that 
 
 This cut is reproduced from " Voice, Song, and Speech."
 
 the partial tones change in pitch with every change of vowel effed. 
 The loud fundamental is due to the vibration of the vocal cords, and 
 the " partial tones " are caused by the resonance of the air in the 
 cavities of the mouth. 
 
 " What do you mean by 'the resonance of the air in the cavities 
 of the mouth?'" I fancy some of you ask. In order to answer 
 this question 1 have brought a few empty bottles from the dining- 
 room table and from the kitchen of the hotel. Here we have a pep- 
 per-pot, a pickle-bottle, a mustard-pot, a vinegar-bottle from the 
 cruet-stand, and a few other bottles of different shapes and sizes. 
 Now let me blow into the mouth of one of these bottles. At once 
 you hear a musical tone something like that produced by an organ 
 pipe. I shall now blow into the mouths of the others. You .observe 
 that each bottle has a resonance tone of its own. In some cases the 
 pitch is high, in others low. Observe the pitch of the bottle 1 hold 
 in my hand. I shall now pour in a little water so as to reduce the 
 air space within. The bottle produces a tone of higher pitch than 
 it did before. I pour in a little more water and again the pitch rises. 
 In fa(5l, the smaller the cavity is made the higher does the pitch 
 become. Now you have in your mouth a bottle-shaped cavity, and 
 in this case also the air within has a tendency to vibrate at a 
 definite rate so as to produce a musical tone. When the size of 
 the cavity is reduced by elevating the tongue and bringing it further 
 forward in the mouth, the pitch becomes higher, just as the tone 
 produced by the bottle rose in pitch when I poured in water. I am 
 afraid you would hardly like me to demonstrate the truth of this 
 statement by blowing into your mouth as I did into the bottle! If 
 you are anxious to make the experiment you can blow into your 
 own mouth with a pair of bellows! A still simpler way, however, 
 of testing the effect is to blow air through the mouth from the 
 lungs. For example: whistle. The pitch of the whistle rises as 
 the tongue is advanced in the mouth. 
 
 Let me dired your attention once more to the bottle. The pitch 
 rose when I poured in water, and of course I can lower it again, if 
 I choose, by pouring out the water. Instead of doing this, however, 
 I shall change the pitch in another way, without varying the size of 
 the air space within. While 1 blow into the bottle I shall gradually 
 cover its mouth with my hand. The tone, you observe, falls in 
 pitch as the orifice is reduced. You see from this that you can vary 
 the pitch; (i) by varying the size of the cavity, and (2) by chang- 
 ing the size of the opening into it. Allow me to illustrate these two 
 ways with my mouth.
 
 26 
 
 1. I shall assume the position of the Visible Speech symbol 
 " Back center-aperture " (German r/« in the word nach); and then 
 glide the tongue gradually forward to the position for " Front centre- 
 aperture " ( h in the word hue) thus reducing the size of the cavity 
 in front of the tongue. The pitch of the sound rises as the tongue is 
 advanced in the mouth. 
 
 2. 1 shall now retain tne "Back centre-aperture position (Ger- 
 man ch) and gradually contraft the aperture between the lips until 
 the position for the English element wh is reached. The pitch falls 
 as the labial aperture is reduced. 
 
 In forming German rJi compressed air from the lungs escapes 
 through a very small aperture between the back of the tongue and 
 the soft palate, occasioning a rustling sound in the mouth. Although 
 this effect constitutes a noise rather than a musical tone, you have 
 no difficulty in recognizing that it has pitch. 
 
 1 think you should teach all your pupils to produce German 
 ch, because this position of the tongue enters into the composi- 
 tion of three English elements which are usually pronounced in a 
 very detedive manner. I allude to the sounds of wh, iv, and the 
 vowel 00 in the word foo. In these cases the aperture between the 
 lips IS so small as to prevent the pupil from observing the position 
 of the tongue, which position is essential to the production of the 
 sound. He imitates the labial aperture perfectly, but fails to give 
 the back tongue position, and hence produces only a crude approxi- 
 mation to the correft sound. 
 
 A short time ago I visited one of our best articulation schools, 
 and went through all the classes in search of a good oo. The only 
 children who gave the sound corredly were semi-deaf, or had 
 acquired speech by ear. There is really little difficulty in teaching 
 the sound if you commence with the lingual element by itself, and 
 then modify it by rounding the lips. Commence with German ch. 
 Round the lips and you have wh. Add voice and you have oo. For 
 all practical purposes this may be considered identical with w. 
 When the two elements are in juxtaposition the difference is readily 
 perceived by the pupil. For example : Pronounce the word woo. 
 The labial aperture is visibly smaller for the consonant than for the 
 vowel. The only trouble in teaching this sound arises from the faft 
 that a very slight change in the lingual position destroys the oo 
 effeft. If the tongue is only a little too far forward or a little too 
 far back, the position may yield a very respectable German r //, and 
 yet fail to produce a good oo when the lips are rounded and the
 
 27 
 
 voice is sounded. In such cases the pitch of the German ch will 
 tell you the nature of the defed, and how to remedy it. If the pitch 
 is too high the tongue is too far forward ; if it is too low the tongue 
 is too far back. You can obtain your standard for comparison in 
 the following way : Pronounce a good oo. Convert it into wh by 
 substituting breath for voice, and then force your lips apart so as to 
 obtain the effect of the lingual position alone. Observe the pitch of 
 the German ch thus produced. If the pitch of the sound produced 
 by your pupil is higher than this, dired: him to place the tongue 
 further back ; and if it is lower, tell him to bring the tongue forward. 
 
 The pitch of the mouth can be brought out by other means 
 than by blowing air into or through the cavity. Resonance is caused 
 whenever a sound of similar pitch is produced in the neighborhood. 
 For example: Here is a tuning fork, and upon the table is a bottle 
 which has the same pitch. I hold the vibrating prongs of the fork 
 over the mouth of the bottle, and at once its resonance tone is loudly 
 evoked. Here is another bottle, but it remains silent when the fork 
 is rpolied. Upon blowing into it you perceive that the pitch is too 
 low. Let me tune it by pouring in water. It still fails to respond — 
 the pitch is nov/ too high. Upon pouring out a little water the bot- 
 tle resounds, but very faintly. It has almost the same pitch as the 
 fork — but is still a little too high. I pour out a few more drops, and 
 now you hear the full and loud response made when the fork is ap- 
 plied. Let me hold the fork in front of my lips while my mouth is in the 
 position of "wh. You have no response, because the proper tone of 
 the mouth cavity is different from that of the fork. Upon tuning 
 the cavity by shifting the position of the tongue, the mouth resounds 
 as the bottle did a few moments ago. 
 
 In these cases you have resonance produced by "sympathetic 
 vibration." If you have in the same neighborhood two bodies that 
 tend to vibrate at the same rate, set the one vibrating and the other 
 vibrates of itself — out of "sympathy" as it were! I shall show you 
 another case. Here we have a piano. 1 shall depress the pedal so 
 as to release all the strings, and then sing into the instrument. When 
 1 stop singing you will observe that the piano echoes my voice. 
 That string of the piano that had the same pitch as my voice was 
 set sympathetically into vibration. A similar effe(f\ is produced in 
 the case of vibrations which are too slow to produce the sensation 
 of sound. For example: If two clocks having pendulums of 
 similar length are attached to the same wall you need only set 
 one of them going, for by and by the other will go by itself. Of
 
 28 
 
 course we cannot suppose that the pendulums had any particular 
 "sympathy " or afFe(5\ion one for the other — or that the string of the 
 piano experienced any emotion at the sound of my voice! Every 
 mechanical effeft must have a mechanical cause — and the fads of 
 " sympathetic vibration " require explanation. Let us consider the 
 case of an oscillation slow enough to be followed by the eye. For 
 example: Imagine one of your pupils to be upon a swing. Stand 
 behind him and give him a shove. The swing moves forward a 
 little way and returns upon its path. It oscillates backwards and 
 forwards at a definite rate for a long time before it comes to rest. 
 Indeed, were it not for frittion and resistance of the air it wouldn't 
 stop at all. Inertia would keep it going; and one shove would be 
 sufficient to set it vibrating forever! In spite of the resistance of air, 
 the effeft of a single push is retained through many vibrations. If, 
 then, you push the swing again at the proper time, the motion is 
 increased. Very slight efforts of the hand will suffice to set the 
 swing into full vibration if the impulses are properly timed to the 
 movements of the swing so as always to come during the forward 
 motion alone. Under such circumstances the effect of each succes- 
 sive shove is added to that of the last, and a cumulative effect is 
 produced. The amplitude may become very great even though the 
 individual impulses are slight. If the shove is given at the wrong 
 -time, that is, when the swing is moving on its backward path, then 
 every impulse tends to stop the vibration. Each push retards the 
 motion to a certain extent, and a series of very slight impulses will 
 bring the swing to rest. For example: The resistance of the air 
 stops it in time if you leave it alone. The air, in effeft, shoves the 
 swing at the wrong time at each vibration. Make it push at the 
 right time, and the converse effedl will be produced. You could 
 set the swing going, for instance, by puffs from a pair of bellow ! 
 
 Now the strings of a piano are in effed swings, each tuned to 
 vibrate at a different rate from the other ; and puffs of air from my 
 lungs set them going when I sang into the instrument. At each 
 opening of the glottis a puff of air escaped from the lungs, and all 
 the strings of the piano received a shove. The first shove started 
 them all swinging, but th» second caught some of them on the 
 return path and stopped their motion. The motion of others was 
 retarded later on ; but that string which had the same rate of vibra- 
 tion as my vocal cords, received each shove always at the right 
 time, and was thus set into vigorous vibration. It continued sound- 
 ing for some time after I stopped the voice, just as a swing
 
 29 
 
 continues in vibration after you stop pushing it. The same kind of 
 adion took place when the tuning fork was held over a bottle of 
 similar pitch to its own. At each descent of the prong the air in 
 the bottle received a shove ; and the air was thus set into vibration, 
 as you set a swing into vibration by the hand. 
 
 A similar aftion took place when I held the vibrating fork in 
 front of my lips. At first no sound was produced, but when 1 
 shifted the position of the tongue, so as to tune the mouth-cavity 
 to the pitch of the fork, resonance resulted, and you all heard the 
 effed. I have no doubt that the Scotchman with the artificial 
 larynx could have produced the same effed, if he had slipped a 
 tuning-fork into his throat in place of the harmonium reed. Imagine 
 a multitude of tuning-forks of different pitch to be massed together 
 in front of the mouth and all simultaneously to be set in vibration. 
 It should then be possible, by shifting the position of the tongue, to 
 reinforce the tone — now of one fork, now of another — at will. 
 Indeed under such circumstances, it would hardly be possible to 
 assume a position of the mouth, that would not reinforce some 
 fork — at least in a greater or less degree. Imagine the mass of tun- 
 ing-forks to be placed in the Scotchman's throat, and similar effeds 
 would result. 
 
 Now the vocal cords like the hypothetical forks, produce a 
 number of feeble tones of different pitch ; when we pronounce a 
 vowel sound, the mouth cavity reinforces, by resonance, that " par- 
 tial tone" of the voice which is nearest in pitch, to the proper tone 
 of the cavity. The effect produced we call the " vowel quality." 
 The loud fundamental tone of the voice, so distrafts the attention of 
 the untrained observer, that he finds difficulty at first in hearing the 
 resonance tone produced by the mouth. The best way to train the 
 ear is to commence by observing the pitches of non-vocal sounds. 
 Then listen for similar effects when the voice is sounded. 
 
 If you whisper the vowels ah, aiv, oh, 66, I think you will have 
 little difficulty in recognizing the fa<ft that the pitch of the whisper 
 falls as the lips are approximated. More difficulty will be experi- 
 enced in determining the relative pitches of other vowels. For 
 example : whisper the vowels in the words eel, ill, ale, ell and 
 shall. The pitch changes with each vowel, but how does it change ? 
 I attempted to determine the point a good many years ago, and 
 came to the conclusion that the vowels formed a descending musical 
 scale, ee having the highest, and a the lowest pitch. To my surprise, 
 however, my father was unable to agree with me in this result. To
 
 30 
 
 his ear, the vowels formed an ascending series, ee having the lowest, 
 and a the highest pitch. The faft is, we were both right, for these 
 vowels have a double resonance. The passage-way for the voice 
 extends from the vocal cords to the lips, and if you constrid it at 
 any point, you divide it into two parts forming bottle-shaped cavi- 
 ties placed neck to neck. (See diagram. Fig. 3.) There is a cavity 
 in front of the point of constriction and another behind it. 
 
 In forming ee, the front cavity is very small, and the pitch con- 
 sequently high; but the back cavity is low in pitch, because it 
 possesses an extremely narrow neck at the point of constridion. 
 In pronouncing the other vowels of the series mentioned, the front 
 cavity increases in size, and, therefore, falls in pitch ; but the pitch 
 of the back cavtiy rises because the neck at the point of constric- 
 tion is enlarged. I can demonstrate the double resonance of these 
 vowels by a simple experiment. 1 shall hold the side of a lead 
 pencil against my cheek, and tap it forcibly with my thumb nail, so 
 as to agitate the air in the front cavity, while I whisper the vowels 
 ee, i, a, e, a. You perceive at once a descending series of sounds, 
 in which ee is the highest, and a the lowest pitch. 1 shall now hold 
 the pencil against my throat so that each tap may agitate the air in 
 the back cavity. Upon whispering the same vowels, the taps pro- 
 duce an ascending series, ee having the lowest pitch and a the 
 highest. The effed is improved by closing the glottis so as to con- 
 vert the back cavity into a bottle closed at the bottom. The front 
 cavity also yields a much louder effeft if it is shut off completely 
 from the back cavity, by allowing the soft palate to fall into contact 
 with the back of the tongue {ng position). In these cases, of 
 course, the vowel positions must be silently assumed. When we 
 pronounce these vowels aloud, feeble " prartial tones," due to the
 
 31 
 
 resonance of the air in the back and front cavities of the mouth, 
 mingle with the tone of the voice, and produce in our ears the 
 sensation of "vowel quality." 
 
 Helmholtz has not only resolved vowels, by a process of analy- 
 sis, into their constituent musical elements, but has produced vowel 
 sounds artificially by a synthetical process. In place of voice, he 
 caused a tuning fork to vibrate continuously in front of a tuned 
 bottle or " resonator," thus producing a loud musical tone. He then 
 selected two forks having the pitches of the partial tones he had 
 recognized as charafteristic of the vowel ee. One was very high, 
 and the other low. (They represented, indeed, the front and back 
 cavities of the mouth in forming the vowel.) These forks were then 
 placed in front of bottles, tuned almost but not quite to their own 
 pitch, so that the sounds produced should be very faint. The simul- 
 taneous vibration of the three forks in front of their respective 
 resonators, or bottles, produced one loud sound, and two feeble 
 partial tones. The effeft upon the ear was that of the vowel ee. 
 
 Those of you who desire to pursue this subject further, may 
 consult a paper of mine upon "Vowel Theories," which was read 
 before the National Academy of Sciences, April 15, 1879, and was 
 published in i\\e Journal of Otology, Vol. 1, July, 1879. This paper is 
 printed as an Appendix in this volume.
 
 THE FUNCTIONS OF THE EPIGLOTTIS AND 
 
 SOFT PALATE. 
 
 The instrument of speech consists essentially of a collection of 
 tubes or passage-ways connefted together somewhat as shown in 
 the following crude diagram, which I have found of assistance in 
 explaining to pupils the functions of the epiglottis and soft palate. 
 (See Fig. 4.) 
 
 Fig- 4- 
 Diagram illustrating the aftion of the epiglottis ( k) and soft palate (n). 
 
 32
 
 33 
 
 Key to diagram, Fig. 4. 
 
 There are three entrances into the vocal organs; a, the mouth, 
 and b, c, the nostrils. Following these passages downwards we 
 find they unite in one passage, d, the pharynx. Below this point 
 the passage-way splits up into two tubes, e, f, the sesophagus and 
 the windpipe. The windpipe, f, bifurcates lower down into the 
 bronchial tubes, g, h. These in their turn split up into multitudi- 
 nous smaller tubes, ramifying through the lungs. The aesophagus, 
 e, passes downwards between the bronchial tubes, g, h, through 
 an aperture in the diaphragm, i, into the stomach. 
 
 In this apparatus we find two valves one, k, (the epiglottis) 
 hinged at 1, and capable of shutting against m. The other, n, (the 
 soft palate), hinged at o, and capable of shutting against p, 
 and q. 
 
 These valves are largely for the pr6te(5tion of the lungs. We 
 all know how important it is that foreign bodies should be kept out 
 of the lungs. The New York doftor who recently inhaled a cork 
 has died, in spite of all that science could do to aid him. Equally 
 serious results might follow were particles of food to find their way 
 into the lungs.
 
 34 
 
 The pharynx, d, forms a common passage-way through which 
 both food and air pass, and the valves, k, n, prevent the passage of 
 food into the wind-pipe, and permit breathing to take place with 
 safety during the process of mastication. If we were obliged to 
 breathe through the mouth-passage, a, while the mouth contains 
 partly masticated food, it would be almost impossible to prevent 
 particles from being drawn into the lungs with the breath. The 
 valve, n (the soft palate), obviates such a catastrophe by shutting in 
 the contents of the mouth during the process of mastication, by 
 closing against p (the back of the tongue), as shown by dotted 
 lines. Breathing can be carried on safely behind the soft palate 
 through the nasal passages, b, c. When, however, the process of 
 mastication is completed, a new danger threatens the lungs. The 
 food, on its way to the stomach through the sesophagus, e, must 
 pass the upper end of the wind-pipe, f. The valve, k, (the epiglottis), 
 closes tightly against m during the aft of swallowing, and thus 
 prevents the possibility of food obtaining access to the wrong pas- 
 sage-way. The larynx constitutes a sort of box on top of the 
 wind-pipe, of which the epiglottis k, forms the lid. In the diagram, 
 I have represented the lid as shutting down on the top of the box, 
 but in the aftual instrument of speech the box also shuts up against 
 the lid. Place your hand against your throat, and you can feel the 
 larynx rise when you make the a6l of swallowing. In speech, the 
 soft palate, n, is used for the purpose of directing the breath through 
 the mouth or nasal passages, as desired, When it shuts against p, 
 ( the back of the tongue), air from the lungs passes up behind it 
 through the nasal passages, b, c, and no air can escape through the 
 mouth, a. When it shuts against q, (the back of the pharynx), air 
 from the lungs passes in front of it, through the mouth, a, and no 
 air escapes through the nasal passages, b, c. When it hangs half 
 way down, as shown in the position o, n, air from the lungs escapes 
 through all three passages, a, b, c, simultaneously 
 
 METHODS OF STUDYING THE MECHANISM OF SPEECH. 
 
 I have already direfted your attention to the fad; that many 
 words, and even sentences, may be pronounced during a single 
 effort of expiration. The muscles concerned in expiration produce a 
 continuous, steady pressure upon the air in the thorax, during the 
 aft of speech. In flift, we play upon the instrument of speech as 
 the piper plays upon the bagpipe. The piper's arm continuously
 
 35 
 
 squeezes the bag all the time he is playing. He does not jerk his 
 arm in time to the music. A continuous, steady pressure exists all 
 the time his fingers are moving. When an amateur tries to play 
 upon the bagpipe for the first time, he is apt to give a fresh squeeze 
 for every note, producing an effe6l somewhat like the jerky utter- 
 ance of deaf children who have been taught word-by-word 
 articulation. 
 
 When children are taught at first to pronounce each word by 
 itself, with a distind and separate effort of emission, an intermittent 
 adion of the abdominal muscles is apt to become habitual, even in 
 rapid utterance. 
 
 We have had the opportunity of listening to pupils from Mil- 
 waukee, Boston, and Philadelphia, and 1 have been delighted to 
 observe that they do not exhibit this fault. Whatever defeds of 
 speech they possess, they do not exhibit the fundamiental error of 
 word-by-word, syllable-by-syllable articulation. I would certainly 
 recommend to your notice the exercises that have been so success- 
 ful in producing continuity of utterance in their cases. 
 
 The material from which speech is made, is a store of com- 
 pressed air within the thorax, which is let out little by little, in a 
 continuous, steady stream, and moulded into the various sounds of 
 speech. No compression is possible unless the escape of air is 
 restrained. In order that we should have a continuous, steady 
 stream, it is necessary that emission should take place through 
 only a very fine orifice. You all probably know the effed of restrain- 
 ing the emission of fluid from a pipe, under pressure. If you 
 haven't made the experiment, partially plug the orifice of a faucet 
 with your finger, while the water is running. At once the slow, 
 silent stream, is converted into a rushing torrent, which spurts out 
 with great noise. Instead of a large quantity of water coming out 
 slowly, you have a small quantity rushing out quickly. Instead of a 
 silent flow you have noise. In the produdion of noise a little 
 water goes a great way, and a steady stream can be sustained for a 
 long period of time, without the expenditure of much fluid. 
 
 In a similar manner, noise is produced by partially plugging the 
 air-passage from the lungs ; and all the elements of speech result 
 from constriftions of some kind. In studying the mechanism of 
 speech-sounds, therefore, it is necessary to determine the location 
 and nature of those constridions that produce and modify the 
 sounds.
 
 36 
 
 Now how are you going to describe a constridion ? My father 
 has pointed out that the principal organs concerned in the produc- 
 tion of speech, naturally group themselves into two classes, adive 
 and passive. As a general rule, the lower organs are aftive, and the 
 upper passive. For example: In forming the sound of A the point 
 of the tongue is the adive agent involved, and the upper gum is the 
 passive. In this case the two organs are approximated together in 
 such a manner as to completely close the passage-way between 
 them. 
 
 Irv describing other constridions also, it is usual to designate: — 
 
 (a) The adive organ employed ; 
 
 {b) The passive organ; and 
 
 {c) The condition of the passage-way between them. 
 Whether or not we adopt my father's classification into adive 
 and passive organs, it is certainly the case that we are obliged, in 
 defmmg accurately the location and nature of a constridion, to dis- 
 tinguish three associated elements, viz: — 
 
 -I f Two organs which are approximated together; and 
 (c) The condition of the passage-way between them. 
 
 A constridion is usually termed "a position of the vocal 
 organs." When two or more positions of the vocal organs are 
 simultaneously assumed, the effed upon the ear is that of a single 
 sound. In such a case the passage-way is constrided at more than 
 one point at the same time. For example: Take the vowel oo, in 
 such a word as too. In forming this sound three distind positions 
 (P, F, P",) are simultaneously assumed. (See Fig. 6.) 
 
 P. A labial position. 
 
 -,, ^ -, (a The under lip. 
 
 The two organs ? -j ^^ The upper lip. 
 
 Passage-way between ? c A small central aperture. 
 
 P'. A lingual position. 
 
 rr, ^ :s (a' The back of the tongue. 
 
 The two organs ? -j ^, ^j^^ ^^^^ p.^,^^^ 
 
 Passage-way between ? c' A small central aperture. 
 
 P*. A laryngeal position. 
 
 ( a" 
 The two organs ? j i » The two vocal cords. 
 
 Passage-way between .? c" A slit- like aperture.
 
 37 
 
 Fig. 6. Positions of the vocal organs in forming the vowel ^ in " too." In this 
 diagram a side view of the mouth is given in sedion, but a front view of the larynx so 
 as to show both of the vocal cords. 
 
 The last position (P") represents the condition of the glottis in 
 forming "voice." The passage of air through the slit-like aperture 
 occasions a vibration of the vocal cords. The glottis then alter- 
 nately opens and closes with great rapidity, causing the emission of 
 a series of puffs from the lungs. These puffs follow each other 
 with such rapidity that the ear fails to distinguish the individual im- 
 pulse, and recognizes only a continuous effeft of a musical char- 
 after, which we term "voice." 
 
 When we come to study the method of symbolizing positions 
 of the vocal organs devised by my father, Prof A. Melville Bell, and 
 called by him "Visible Speech," we shall find great advantage from 
 considering the symbols as algebraical signs. Positions, (like P. P', 
 and P"), which are simultaneously assumed, may be considered as
 
 38 
 
 added together (P+P'+P"); but the three elements (a b c) which 
 compose each position, must be taken as multiplied into one an- 
 other (axbxc.) Thus: — 
 
 Position for 00= P+P'+P":=a b c + a' b' c' + a" b" c". 
 
 Position symbols placed in juxtaposition without any connec- 
 tive sign between them must be taken as successively, not simul- 
 taneously, assumed. Thus P P' P" means : Assume first the posi- 
 tion P alone, then P' alone, and then P" alone. 
 
 If the positions (P, P', P",) are assumed separately, instead of 
 together, they yield sounds that are quite unlike the vowel 00. For 
 example: If the labial position (P) be assumed without any other 
 constriction in the passage-way, a sound results which is not an 
 English element of speech. English-speaking children, however, 
 give the sound when they blow upon their porridge to make it cool. 
 
 If the lingual position (P') alone be assumed, the resulting sound 
 is the German ch in such a word as nach. 
 
 If the vocalizing position of the glottis, (P"), be assumed with- 
 out any other constridion higher up in the passage-way, an indef- 
 inite vowel sound results like the er-er-er of a hesitating speaker, or 
 like the vowel heard in such words as her, sir, word, etc. 
 
 If two of the positions (P, P', P",) are simultaneously assumed 
 without the third, still other sounds result which neither resemble 
 the vowel 00, nor the sounds I have just described. For example: 
 
 If the positions P and P' are simultaneously assumed without 
 P", the English consonant wh is produced. 
 
 If the positions P and P" are simultaneously assumed without 
 P', the sound heard is that of the German w, in the word wie. 
 
 If the positions P' and P" are simultaneously assumed without 
 P, the result is the Gaelic vowel in the word '' laogh." 
 
 These fads may be arranged in tabular form as follows : 
 
 ITIONS ASSUMED. 
 
 RESULTING SOUNDS. 
 
 P 
 
 Blowing to cool. 
 
 P' 
 
 German r/? in "nach." 
 
 P" 
 
 The vowel er in " her." 
 
 P+P' 
 
 The consonant wh in " what." 
 
 P+P" 
 
 German w in "wie." 
 
 P' + P" 
 
 Gaelic " aohg" in " laogh." 
 
 P + P' + P" 
 
 The vowel 00 in "too." 
 
 What we term an "element of speech" may in reality, like 
 the vowel 00, be the result of a combination of positions. The true
 
 39 
 
 element of articulation, 1 think, is a constridion or position of the 
 vocal organs rather than a sound. Combinations of positions yield 
 new sounds just as combinations of chemical elements yield new 
 substances. Water is a substance of very different character from 
 either of the gases of which it is formed ; and the vowel oo is a sound 
 of very different character from that of any of its elementary positions. 
 When we symbolize positions, the organic relations of speech- 
 sounds to one another can be shown by means of an equation, for 
 example: — 
 
 English wh = P+P\ 
 German ch = P'. 
 Hence German ch = English wh- --P. 
 
 The equation asserts that the English wh without the labial con- 
 striction (P) is the German rh. 
 
 1 performed this equation upon the mouth of Mr. Lyon during 
 the course of my last le<5ture. While Mr. Lyon was prolonging the 
 sound of wh, I forced his lips apart with my fingers, and you then 
 heard the sound of German ch. Take another case: — 
 
 The English vowel oo in " too " = P + P' + P". 
 The Gaelic vowel " aogh " in " laogh " = P' + P". 
 
 Hence " aogh" =oo — P. 
 
 That is: The English vowel " oo," without the labial constric- 
 tion (P) is the Gaelic vowel "aogh." 
 
 If then you desire to pronounce the Gaelic vowel ''aogh, "sing 
 the vowel oo while some one else forces your lips apart. This is a 
 diredion that will enable any English-speaking person to convert 
 the known vowel oo into the unknown Gaelic sound "aogh," with- 
 out the aid of hearing. In a similar manner, the sounds known to 
 the deaf child can be converted into the unknown sounds of the 
 English language. Indeed, manipulation succeeds better with a 
 deaf person than with one who hears, because the hearing person 
 attempts to retain the sound, whereas the deaf child simply tries to 
 retain the position. 
 
 The symbols of Visible Speech bear the same relation to pho- 
 netics that chemical symbols do to the science of chemistry. In 
 dealing with the mechanism of speech, it is as necessary now-a- 
 days to make use of my father's symbols, as it is to use chemical 
 symbols in treating of the composition of matter. 
 
 As many of you are already familiar with the subjed, it will 
 not be necessary for me to enter into any detailed description of 
 Visible Speech. 1 shall, therefore, to-day, simply attempt to give a
 
 40 
 
 general idea of the nature of my father's method of symbolizing 
 positions of the vocal organs, so as to enable those who are unfa- 
 miliar with the subject to follow me intelligently in my use of the 
 symbols. 
 
 The fundamental characters represent the vocal organs and the 
 various kinds of apertures employed in the production of speech; 
 and these can be combined into a compound form to express a 
 position of the vocal organs. 
 
 A character shaped like a tube ( C ) is used to denote a very 
 small aperture for the escape of breath; and when this tube is 
 plugged up at one end ( CI ) the symbol then indicates complete 
 closure of the passage-way. 
 
 The symbols for the principal organs of speech are shown in 
 Figure 7. 
 
 The active organs of the mouth, viz: the under lip, the point 
 of the tongue, the top or "front" part of the tongue, and the back 
 of the tongue are represented by curves that form, in such a dia- 
 gram as that shown, the outlines of the organs themselves. This 
 pictorial basis forms an element of great value in teaching the deaf. 
 
 The upper or "passive" organs, to which the lower or 
 " aCtive " organs are usually applied, are represented by the same 
 symbols, written upon a small scale. 
 
 Thus, the symbol for the upper lip is a miniature of that for the 
 lower lip; the upper gum, to which the point of the tongue is 
 usually applied, has the point-tongue sign upon a small scale; the 
 top of the hard palate is expressed by the symbol for the top or 
 "front" of the tongue; and that part of the soft palate to which 
 the back of the tongue is applied, is denoted by the back-tongue 
 symbol upon a reduced scale.
 
 41 
 
 Fig7- 
 
 ) Under lip. 
 
 •^ Point of tongue. 
 
 --» " Top " or " Front " part of tongue. 
 
 ( Back of tongue. 
 
 \ Back of soft palate. 
 
 I Slit-like aperture between the vocal cords. 
 
 3 Up|>er lip. 
 
 " Upper gum. 
 
 <> Top of hard palare. 
 
 c Front part of soft palate. 
 
 ' Back of pliarynx behind sort palate.
 
 42 
 
 T3 
 
 t/3 
 
 3 
 
 E 
 
 on 
 
 C 
 
 C3 
 
 U) 
 
 u 
 
 O 
 
 > 
 
 c 
 
 
 o 
 
 
 
 
 <4— > 
 
 
 
 
 (A 
 
 
 o 
 
 
 a. 
 
 
 n 
 
 
 00 
 
 
 t/2 
 
 
 <u 
 
 
 ;_ 
 
 
 O- 
 
 
 X 
 
 ^d 
 
 o 
 
 ■4-1 
 
 bb 
 
 t/2 
 
 
 « 
 
 E 
 
 
 
 o 
 
 • J 
 ^ 
 
 00 
 
 bij 
 
 
 
 oo 
 
 "•■* 
 
 i- 
 
 -5 
 
 ^ 
 
 _c 
 
 
 
 «^J 
 
 
 k. 
 
 
 « 
 
 c 
 
 
 ^ 
 
 
 o 
 
 <u 
 
 -C 
 
 00 
 
 00 
 
 <u 
 
 
 ^ 
 
 00 
 
 •4— • 
 
 r^ 
 
 
 o 
 
 bJD 
 
 
 c 
 
 '§. 
 
 
 
 c 
 
 ■4-1 
 
 ^ 
 
 t/0 
 
 E 
 
 o 
 
 o 
 
 u 
 
 «J 
 
 
 
 a; 
 
 V- 
 
 j= 
 
 O 
 
 ■♦— ' 
 
 (U 
 
 bjj 
 
 "O 
 
 c 
 
 O 
 
 |n 
 
 ti^ 
 
 
 
 o 
 
 (L> 
 
 X 
 
 J= 
 
 E 
 
 H 
 
 
 oo 
 
 
 >> 
 
 
 X: 
 
 o 
 
 c 
 
 OLE 
 
 13 
 
 X 
 
 ttl 
 
 c 
 o 
 
 
 &-0 
 
 r; 
 
 bo 
 
 O 
 
 'u. 
 
 
 
 
 
 
 
 
 
 -C 
 
 
 OJ 
 
 
 
 ^ 
 
 OJ 
 
 o 
 
 
 r-; 
 
 G. 
 
 (A 
 
 X 
 
 
 tu 
 
 
 o 
 
 o 
 
 ■4—1 
 
 a- 
 
 — r- 
 
 1- l- 
 
 C 
 _ — <U 
 
 (D <U U 
 
 ■a ci. , 
 
 3 3^ 
 
 OJ <D V3 
 
 o 
 
 
 t/5 
 
 s;; t>c 
 
 •t-i 
 
 Xi 
 
 o 
 
 O r: 
 
 — O 
 
 Xj -t-i 
 
 (U 
 
 JZ 
 
 H 
 
 I 
 
 biD 
 rz 
 oo 
 oo 
 
 a. 
 
 u 
 
 a 
 
 c 
 o 
 
 ■4—' 
 
 o^ 
 
 u - 
 
 3 
 
 ■4-' 
 
 a. 
 
 o -5 c 
 
 :^ O = 
 
 Xi CO W 
 
 (1) OJ t 
 
 
 o 
 
 C/3 
 
 o 
 
 3 
 
 be 
 
 C 
 
 C 
 
 -4-> 
 
 D 
 Xi 
 
 t/2 
 
 c 
 be 
 
 o re 
 
 o 
 
 I 
 
 bo 
 
 0) 
 
 
 oo 
 
 r- '^ 
 
 1^ _: 
 
 "O . ■- 
 03 j::^ 
 
 "S £ 
 
 E:^ 
 
 — 03 
 
 J;; D 
 ^ a, 
 
 (U 00 
 
 
 (U 00 
 
 o 
 
 3 — 
 ■^ 03 
 <J 
 O 
 
 
 1^ i' 
 
 ~C3 
 
 "o3 X ' 
 
 o 
 
 00 
 
 O 
 
 00 
 
 C 
 
 _ bD 
 
 be 
 
 c 
 
 « 
 
 
 o 
 ■*-> 
 
 c 
 
 ■4-1 
 CD 
 
 03 
 I 
 
 bo 
 
 03 
 00 
 00 
 
 «
 
 43 
 
 The various combinations of these positions tabulated upon 
 page 208, may be symbolized as follows : 
 
 POSITIONS ASSUMED. 
 
 Expressed Expressed 
 
 upon plan shown upon plan shown 
 
 in Fig. 6. in Fig. 7. resulting sounds. 
 
 P D^ Blowing to cool. 
 
 P' C*= German ch in " nach." 
 
 P" I The vowel i^r in "her." 
 
 P + P' O^ + C*= The consonant tvh in ' ' what. " 
 
 P+ P" 0^ + 1 German w in " ztie." 
 
 P+P" C^ + l Gaelic aogh in " laogh." 
 
 P+P'+P" O^' + C^ + l . .The vowel 00 in "too." 
 
 The equations to which 1 have direded your attention may be 
 thus expressed: 
 
 1. German ch = English zch — P. 
 
 C^ = (D^ + C<:) — D^ 
 
 2. Gaelic aogh = English 00 — P. 
 
 C*^ + I = (O^ + C^ + I) — O^ 
 
 Three positions (P, P', P"), have been shown in Fig. 6, but a 
 fourth position, which has not hitherto been noticed, is also indi- 
 cated in that diagram. 
 
 The soft palate is in contad with the back of the pharynx, thus 
 closing the entrance to the nasal passages. Visible Speech affords 
 us a means of expressing this position if we so desire. The portion 
 of the soft palate that fits against the back of the pharynx, like other 
 active organs, is represented by its own outline {S ), in such a dia- 
 gram as Fig. 7; and the part of the pharynx with which it makes 
 contad could, consistently with the notation, be represented by the 
 same symbol in miniature ('). 
 
 Combining these with the symbol for passage-way closed 
 (CI), we could form the compound charader D' (soft palate shut 
 against back of pharynx). In English utterance this position is con- 
 stantly assumed during the ad of speech, excepting when the 
 sounds of m, n, and ng occur. As a matter of convenience the 
 position of the soft palate is not noted excepting when these sounds 
 occur. 
 
 When the soft palate is depressed, as in Fig. 7, a passage-way 
 exists between it and the back of the pharynx, through which air 
 escapes into the nasal passages.
 
 44 
 
 This position may be thus expressed : 
 
 p, j " A central aperture between the soft palate and the 
 } back of the pharynx." 
 
 In forming /;/ we shut the lips and pass voice through the 
 nose. This sound, therefore, results from three positions which are 
 simultaneously assumed. 
 
 Position for /» = D^ + C + I (See Fig. 8). 
 
 That is: "Under lip shut against upper lip" plus "central 
 aperture, between the soft palate and back of the pharynx," plus 
 "slit-like aperture between the vocal cords." 
 
 Position for « = D" + C + I (See Fig. 9). 
 
 That is, "point of tongue, shut against upper gum" plus 
 " central aperture, between the soft palate, and back of the pharynx,' 
 plus " slit-like aperture between the vocal cords." 
 
 Position for w^ = Q*^ + C + ' (See Fig. 10). 
 
 That is: "Back of the tongue shut against the soft palate" 
 plus " central aperture, between the soft palate and back of the phar- 
 ynx " plus " slit-like aperture between the vocal cords." 
 
 RESUME OF ELEMENTARY POSITIONS. 
 
 = Under lip, shut against upper lip. 
 
 D^ = Point of tongue, shut against upper gum. 
 
 Q'^ = Top (or "front") of tongue, shut, against top of hard 
 palate. 
 
 Q*^ = Back of tongue, shut against soft palate. 
 
 □ ' = Soft palate, shut against back of pharynx. 
 
 O^ = Central aperture, between the under lip and the upper 
 lip. 
 
 U*^ = Central aperture between the point of the tongue and 
 the upper gum. 
 
 O" = Central aperture between the top (or "front") of the 
 tongue and the hard palate. 
 
 C*^ = Central aperture between the back of the tongue and the 
 soft palate. 
 
 C = Central aperture between the soft palate and back of the 
 pharynx. 
 
 I = Slit-like aperture between the two vocal cords. 
 
 An elementary position is expressed by a symbol composed of 
 three associated characters. For example: — 
 
 0= z= :p or P = a b c. (Fig. 6). 
 
 The sign for the constridion (c) is united with that for the
 
 4? 
 
 Fig, 8. Position for M 
 
 l + tl' + D'zr-S 
 
 Fig. 9. Position for N. 
 
 i + c'Do+^ce 
 
 Fig. 10. Position for NG.
 
 46 
 
 active organ (a) so as to form one charader ; and the passive organ 
 (b) appears as a diacritical mark. 
 
 In order to express the mechanism of speech-sounds with accu- 
 racy, as many symbols are required as there are elementary positions 
 to be represented. Hence many sounds like the vowel oo (Fig. 6), 
 require at least three position-symbols to express their formation. 
 (See also positions for iii, n, and iig; Figs. 8, 9, 10.) 
 
 This method of symbolization, though admirable for the pur- 
 poses of scientific analysis, would be unsuitable for current use as a 
 phonetical representation of speech. For example: It would 
 hardly be convenient to write the word " moon " in the following 
 way! 
 
 M 00 N 
 
 (D^ + C + I) (O^ + C^ + I) D" + C + I) 
 
 See Fig. 8. See Fig. 6 See Fig. 9. 
 
 In order to fit the symbols for use as a phonetical alphabet, my 
 father saw that it was necessary, or at all events advisable, that each 
 sound should be represented by only one character; and that there- 
 fore, associated positions should be combined into a compound form 
 capable of use like a letter of the alphabet. 
 
 By the adoption of certain principles of abbreviation this has 
 been accomplished without interfering with the symbolic character 
 of the notation. 
 
 ABBREVIATIONS. 
 
 I. Passive organs may in general be implied and not written. 
 This plan gets rid of the diacritical marks save in exceptional cases. 
 For example: D (•' Point shut"') alone, without any representation 
 of the passive organ, is to be taken as meaning D" ("Point shut 
 against the upper gum ") for the upper gum is the usual place of 
 application for the point of the tongue. The symbol expresses the 
 position for /. 
 
 If, however, a deaf child should pronounce / by placing the 
 point of the tongue against the top of the hard palate, then the 
 passive organ must be shown, thus D". So also when we repre- 
 sent the /-like sound produced during the ad of spitting, D^ 
 (" Point of the tongue, shut, against the upper lip.") 
 
 In forming the Sanscrit "cerebral /" D<= ("Point of the tongue, 
 shut against the soft palate ") the passive organ, being abnormal, 
 must also be indicated. 
 
 In general^ the passive organ may be omitted without ambi- 
 guity, by adopting the rule that in such cases the adive organ shall
 
 47 
 
 be considered as applied to that passive organ which is represented 
 
 by its own symbol in miniature. For example: 
 
 = 3^ U = U^ O = O" C = C^ C = C 
 D = D^ D = D^ Q^Q"" a = a^ □ = D' 
 
 2. I have already alluded to the abbreviation employed to 
 represent the vocalizing position of the larynx. The slit-like aper- 
 ture (I) alone, is used for the full position — the organs themselves 
 (the vocal cords) being implied. This simple symbol (I) is admirably 
 adapted for combination with other signs into a single charader. 
 
 For example: — 
 
 D^ + I = D + I = 
 
 D" + I = D + I = CD 
 
 Q" + I = Q + I = Q 
 
 ac + i = a + i = G 
 
 D^ 4- C^ + I (Fig. 6) = D + C + | = 3 + C = 0-|-e = 3 + G 
 
 3. The symbol S ("soft palate") is used to indicate C (centre- 
 aperture, between the soft palate and back of the pharynx "). This 
 can be combined with other symbols into a single charader. For 
 example: — 
 
 (Fig. 8) D^ + C' + l = D + ) + l = D + l = D + )=a (Position for 
 
 (Fig. 9)0"4-£:' + l=D + ) + l = C5 + I = CD + ) = ?C (Position for 
 
 ;/). 
 (Fig. 10.) 3D + C' + l = a + i + | = G + l = a + j = G (Position 
 
 for iig). 
 
 4. The most difficult case arises when two mouth-positions are 
 simultaneously assumed. Fortunately the curves to be combined 
 are usually of opposite kind so that one can be hooked on to the 
 end of the other. 
 
 Thus C + D = O or a 
 
 The resulting character, however, is of so awkward a shape 
 that another hook is added for the sake of symmetry. Curves of 
 this kind are what my father terms "mixed" symbols. For 
 example: — 
 
 C + ^ = ^ a labial position modified by the back of the 
 tongue (English it'//). 
 
 Or = c; a back tongue position modified by the lips. (A labi- 
 alized German r/i). 
 
 O -f U = Q a front-tongue position modified by the point of 
 the tongue (English s/i).
 
 48 
 
 Or = 23 'T point-tongue position modified by the top or "front " 
 of the tongue (English 5). 
 
 5. When two or more positions are simultaneously assumed, 
 the sound may be considered as originating at the point of greatest 
 constridion — the other constrictions merely modifying the effed. 
 
 In representing associated positions, therefore, the point of 
 greatest constriction is selecfted as the base for the compound sym- 
 bol, and the other positions are indicated in a subordinate manner. 
 
 For example, take the positions shown in Fig. 6. In this case 
 we have three associated positions, P, P', and P". 
 
 Now if P, (the labial position), should happen to be the point of 
 greatest constriction, a rustling noise will be perceived originating 
 at the labial aperture. 
 
 This kind of sound is characteristic of air under pressure, 
 escaping through a fine orifice. We can recognize by ear many 
 varieties of the sound for which we have no name. It varies, 
 according to the size of the orifice, and the degree of pressure, 
 from a simple rustling sound — like the rustling of leaves upon a tree, 
 to an intense hiss — like the noise of steam escaping from a loco- 
 motive. When the vocal organs yield a noise of this character, we 
 call the effeCt a " consonant sound." 
 
 If, then, the 00-like effeCt produced by the positions shown in 
 Fig, 6, is accompanied by a rustling sound at the labial orifice (P), 
 we call the result a labial " consonant," and not a " vowel" although 
 the voice is heard. The labial position becomes the base for the 
 compound symbol which is then written as follows: — 
 
 O^ + C*= + 1 = 3 (English w) a labial position modified by 
 the back of the tongue and the throat. 
 
 If P', (Fig. 6), be the point of greatest constriction a rustling 
 sound is also heard, but in this case it originates within the mouth, at 
 the back-tongue position (German r//), and the labial aperture 
 simply modifies the effeCt. This rustling noise characterizes the 
 sound as a " consonant " although the voice is also heard. In this 
 case the back-tongue position becomes the base and the other posi- 
 tions are indicated in a subordinate manner. 
 
 C*^ + O^ + I = C (German ch modified by the lips and 
 vocal cords. 
 
 If P" (Fig. 6), be the point of greatest constriction, the sound 
 heard originates in the glottis. Instead of a continuous rustling 
 noise or hiss, an intermittent effeCt is produced by the vibration of 
 the vocal cords. The air escapes in a series of p'.;rf5 t'r.at succeed
 
 49 
 
 one another with such rapidity as to produce upon the ear the effe(5t 
 of a musical tone. Voice alone is heard without any rustling or 
 whistling accompaniment in the mouth. This charaAerizes the 
 sound as a "vowel". The sound originates at the position P", and 
 the mouth positions P, P', merely modify the effect 
 
 The slit-like aperture between the vocal cords is therefore 
 made the base for the compound symbol; and the mouth positions 
 P, P', are indicated in a subordinate manner as follows: — 
 
 I -|- C*^ + 0*= = i (vowel 00), a laryngeal position modified 
 by the back of the tongue and the lips. 
 
 The vertical line or "vowel stem," represents (1) the slit-like 
 aperture between the vocal cords (P", Fig. 6), the tlack dot indi- 
 cates the back-tongue position C^ (P') • and the horizontal cross- 
 bar the labial position O^ (P). 
 
 The three sounds represented by the symbols 59. G and 1, 
 result from almost identical positions of the vocal organs, and in 
 teaching the deaf the sounds themselves may be considered as iden- 
 tical. The subtle distinctions, however, recognized by the ear are 
 faithfully depicted to the eve, in the shape and general appearance 
 of the symbols. 
 
 The symbols for w (53), and the vowel oo (\), may be taken as 
 typical of consonant and vowel symbols in general. A curve (/. e. 
 a mouth-position), is the charaderistic feature of a consonant sym- 
 bol; and a straight line (I) (the slit-like aperture between the vocal 
 cords), forms the basis of the vowel notation. 
 
 In conclusion I may say that the symbols for all the English 
 consonants and vowels have been abbreviated to single signs, and 
 that the phonetical alphabet thus produced is admirably adapted for 
 use in schools for the deaf. The following example of abbreviation 
 may be of interest: 
 
 M 00 N 
 
 (D^ + C' + l) (C^ + D^ + l) (D^ + C'-l-l) =SlCD 
 Fig. 8. Fig. 6. Fig. 9. 
 
 In my next lecture 1 shall present the symbols of Visible Speech 
 in the way they are taught to the deaf. 
 
 Dr. Bell then took charge of some children and illustrated his 
 method of teaching. 
 
 Dr. Bell : You must understand that while I claim the privi- 
 lege of telling you in the forenoon what 1 want to tell you, I want 
 to do in the afternoon just what you want. 
 
 (Dr. Bell here gave an exhibition of the clicks with Miss Black's 
 little girl pupil.)
 
 50 
 
 He said: "It does not matter what sound you get from a child 
 as long as you get a sound. The plan is to follow the child up and 
 symbolize the different sounds made, and get him to remember and 
 repeat the varieties that occur. I go from the known to the un- 
 known. The queer sounds children make are the known sounds to 
 them and the English sounds the unknown. Children like the 
 process, and this to my mind is a proof that it is suited to their con- 
 dition. There is something wrong about a process that gives pain 
 to a child. It grieves my heart to visit schools for the deaf and find 
 little children constantly corre6led for minor defefts of pronuncia- 
 tion. The nagging process interrupts the flow of thought through 
 speech, and is apt to dishearten the child in his attempts to speak. 
 I would accept all sounds with approval, and utilize defe(5live sounds 
 in the way I have suggested above."
 
 VISIBLE SPEECH AS TAUGHT TO THE DEAF. 
 
 The following Charts are employed for the purpose of explain- 
 ing to deaf children the meaning of my father's "Visible Speech" 
 symbols. 
 
 The elementary symbols shown in Chart 1. are compounded in 
 Charts II. and HI. to express positions of the vocal organs which 
 yield consonant sounds. In Chart IV. we have other elementary 
 symbols which are combined in Chart V. to express vowel positions. 
 Chart VI. illustrates symbolically the positions of the vocal organs 
 in uttering English consonants, and Chart VII. symbolizes positions 
 that yield English vowels. 
 
 CHART 1. 
 
 The teacher seleds some member of her class, and pretends to 
 draw upon the blackboard the profile of the pupil's f;ice. She then 
 looks into the pupil's mouth and proceeds to draw a picture of the 
 interior of the mouth. The whole piAure when completed, con- 
 stitutes a diagram like that shown in Chart 1. The teacher then 
 proceeds to test the children's comprehension of the drawing. She 
 points to different parts of the diagram, for example the forehead, 
 nose, upper lip, lower lip, chin, lower part of jaw, throat, etc. The 
 children indicate their comprehension of the diagram by touching 
 the corresponding parts of their own faces. Attention is then 
 directed to the interior of the mouth, and the teacher points to the 
 picfture of the upper teeth, upper gum, top of the hard palate, soft 
 palate, etc. The children touch or attempt to touch the correspond- 
 ing parts of their own mouths. So with the lower organs, — the 
 under teeth, the point of the tongue, the top or " front " part of the 
 tongue, the back of the tongue, etc. 
 
 5'
 
 52 
 
 When the comprehension of the class has been weh tested, the 
 teacher erases from the blackboard all those parts of the diagram 
 which are shown by dotted lines in Chart I., leaving the Visible 
 Speech symbols in position as shown by the heavy lines. 
 
 The teacher points to the fragmentary remains of the pidlure 
 upon the blackboard, and the pupils recognize the symbols as "the 
 nose," the "under lip," "the point of the tongue," "the top, or 
 front of the tongue," "the back of the tongue," and "the throat." 
 The arrow-head, which represents a sudden emission, or puff, of air 
 from the mouth, is indicated by a sudden motion of the hand away 
 from the mouth. 
 
 The next step is to have the pupils recognize the symbols inde- 
 pendently of their position on the blackboard. The symbols are 
 therefore written in one line below the fragments of the head (see 
 Chart 1.) The heavy lines alone are written, the dotted lines not 
 appearing at all. 
 
 The pupils then compare these symbols with the fragments of 
 the drawings above and identify them, — as (i) the throat, (2) the 
 back of the tongue, (3) the top, or front part, of the tongue, (4) 
 the point of the tongue, (5) the under lip, (6) the nose, and (7) 
 puff of air from the mouth. 
 
 Finally the upper drawing is entirely removed from the black- 
 board, and the lower line of symbols alone is left. Each pupil 
 describes these as follows: (i) he touches his throat; (2) he points 
 backwards into his mouth with a little jerk of the hand, indicating 
 a part of the tongue further back in the mouth than he can well 
 touch with his fmger; (3) he touches the top, or front part, of his 
 tongue; (4) he touches the tip, or point, of his tongue; (s) he 
 touches his under lip; (6) he touches his nose; (7) he places his 
 hand near his mouth to indicate a sudden emission, or puff, of air. 
 
 After these have been mastered, two new symbols, shown at 
 the bottom of Chart I. are introduced. Here again it should be 
 noticed that the symbols drawn on the blackboard consist only of 
 the parts in heavy lines, the parts in dotted lines being omitted. 
 The first of these new marks as you already know, symbolizes a 
 pipe or passage through which air may pass. In the second case 
 the pipe is shut, or stopped up> at one end. The first indicates a 
 narrow central aperture or passage, somewhere in the mouth ; the 
 second indicates the complete closure or shutting of the mouth- 
 passage at some part. The idea is of too abstract a charafter to be 
 explained at once to a deaf child who knows no language; hence
 
 CHART I 
 
 ( 
 
 ) 
 
 
 I C ^ u ) s 
 
 c a
 
 54 
 
 these symbols are taught arbitrarily as positions of the fingers with- 
 out any attempt being made to explain their significance. As a 
 matter of fad, deaf children come to understand their meaning when 
 applied to the explanation of positions of the mouth. 
 
 The pupils are taught to indicate the first symbol at the bottom 
 of Chart I. by holding the thumb and forefinger of the right hand 
 near to one another without touching. This sign we may translate 
 as "centre aperture." The second or "shut" symbol, is shown by 
 bringing the thumb and forefinger together with a shutting adion. 
 
 We may here notice that the straight line indicating a slit-like 
 aperture between the vocal cords, is used in the sense of " voice." 
 When a deaf child places his hand upon the throat of his teacher 
 he can feel a vibration or tremor in the throat, whenever the voice 
 is sounded. Hence he soon comes to associate the throat sign with 
 a vibration of the vocal cords, and he indicates " voice " by touch- 
 ing his throat. 
 
 It should also be noticed that the " nose " sign is really pictorial 
 of the pendulous extremity of the soft palate, and it indicates, as 
 you have already learned, "soft palate depressed" so as to allow 
 air to pass into the nasal passages. When a deaf child places his 
 finger against the nose of his teacher while she pronounces ///, ;/, or 
 ng, he can feel a vibration or tremor of the nostrils, and to him the 
 soft palate symbol means voice or breath passing through the nose. 
 
 The symbols shown upon Chart 1. are capable of being com- 
 bined into compound forms, some of which are shown in Charts II. 
 and III.. Before proceeding, however, to the analysis of the com- 
 pound characters on these Charts it may be well to assign brief 
 names to the elementary symbols of Chart 1. : these we can use to 
 designate the gestures or signs employed by the deaf child which 
 have been explained above. 
 
 In the following Charts, I shall refer to the symbols at the bot- 
 tom of Chart 1. as — 
 
 I. Voice. 2. Back. 3. Front. 4. Point. 5. Lip. 6. Nose. 
 7. Puff of air. 8. Centre-aperture. 9. Shut. 
 
 CHART II. 
 
 The symbols on this Chart are named by the deaf child by 
 analyzing them into the elementary symbols of which they are com- 
 posed. We may translate his signs as follows : — 
 
 First line.— 1. Lip centre-aperture. 2. Point centre-aperture. 
 3. Front centre-aperture. 4. Back centre-aperture.
 
 55 
 
 Second line. — i. Lip centre-aperture. Voice. 2. Point centre 
 aperture, Voice. 3. Front centre-aperture, Voice. 4. Back 
 centre-aperture, Voice. 
 
 Third line. — i. Lip centre-aperture, Nose. 2. Point centre- 
 aperture, Nose. 3. Front centre-aperture, Nose. 4. Back 
 centre-aperture, Nose. 
 Fourth line. — i. Lip centre-aperture. Voice, Nose. 2. Point 
 centre-aperture, Voice, Nose. 3. Front centre-aperture, 
 Voice, Nose. 4. Back centre-aperture. Voice, Nose. 
 Fifth line. — i. Lip shut. 2. Point shut. ^. Front shut. 4. Back 
 
 shut. 
 Sixth line. — i. Lip shut, Voice. 2. Point shut, Voice. 3. 
 
 Front shut, Voice. 4. Back shut, Voice. 
 Seventh line. — i. Lip shut, Nose. 2. Point shut, Nose. 3. 
 
 Front shut, Nose. 4. Back shut, Nose. 
 Eighth line. — i. Lip shut, Voice, Nose. 2. Point shut, Voice, 
 Nose. 3. Front shut. Voice, Nose. 4. Back shut. Voice, 
 Nose. 
 Long before a class has finished describing these symbols, the 
 pupils begin to obtain the idea that the symbols are directions to do 
 something with the mouth. For example, when they describe the 
 first symbol in the fifth line, " Lip shut," some of them usually shut 
 their lips. After the whole Chart has been described, it then 
 becomes the teacher's duty to make the children understand that 
 the compound symbols they have been describing indicate positions 
 of the mouth. The teacher direds attention to her mouth while 
 she assumes some of the positions symbolized. For example, she 
 describes seriatim the symbols in the first line. 
 
 1. "Lip centre-aperture." She places her lips close together 
 leaving a small aperture between them. She then takes a pupil's 
 hand and blows through this small centre-aperture against his hand. 
 The resulting sound is not an English element of speech, but is the 
 sound produced by blowing to cool something. 
 
 2. She describes the next symbol, "Point centre-aperture." 
 With her hand she lifts up the point of her tongue and brings it 
 into position against the upper gum, and makes the pupil look into 
 her mouth and observe that there is a small aperture or hole between 
 the point of her tongue and the upper gum. She then, without 
 moving her tongue, blows through the point centre-aperture against 
 the pupil's hand. The resultant sound is that of the French /', in 
 the word theiUrc, or the English /-, (non-vocal), in the word irce.
 
 56 
 
 In a similar manner she shows that in pronouncing thf third 
 symbol "Front centre-aperture," the tongue is humped up in the 
 middle, leaving a small centre passage or channel over the front of 
 the tongue, through which she can blow against the pupil's hand. 
 The resultant sound is that of the letter h in the word hue. 
 
 4. In pronouncing the fourth symbol she pushes her tongue 
 towards the back part of her mouth with her hand, and shows that 
 her tongue remains back when her hand is removed. She then lets 
 the pupil feel that air can be blown upon his hand without moving 
 the tongue. The resulting sound is that of the German ch in the 
 word nach. 
 
 Proceeding next to the second line: — 
 
 1. She shows that the first symbol, "Lip center-aperture, 
 Voice," is the same as the first symbol in the first line, " Lip centre- 
 aperture," excepting that a straight line is placed within the curve. 
 She shows then that the lips are in the same position, but that a 
 tremor or vibration can be felt in the throat which could not be felt 
 when the other symbol was sounded. She takes the two hands of 
 her pupil and places one against her throat, and holds the other in 
 front of her mouth while she produces "Lip centre-aperture, 
 Voice." The pupil sees the small centre-aperture between the lips, 
 and feels the emission of air against his hand, and also perceives the 
 trembling of the throat when the voice is sounded. The resulting 
 sound is the German iv in the word wie. 
 
 2. In a similar manner, keeping one of the pupil's hands on 
 her throat and the other in front of the mouth, she produces the 
 second symbol in the second line, "Point centre-aperture, Voice,'' 
 contrasting it with the second symbol in the first line, which has 
 no voice. He sees the centre-aperture over the point of the tongue, 
 and feels the vibration of the voice and the emission of air from the 
 mouth. The resulting sound is that of the letter r in the word run. 
 
 3. In a similar manner she exemplifies the third symbol in the 
 second line, " Front centre-aperture, Voice." The resultant sound 
 is that of the consonant y in you. In teaching the deaf, this may 
 be considered identical with the vowel ee. 
 
 4. The fourth symbol in the second line, " Back centre-aperture, 
 Voice," is shown to be the same as the German ch (Back centre- 
 aperture), excepting that a vibration is felt in the throat. 
 
 Proceeding next to the eighth line: — 
 
 I. The teacher describes the first symbol, " Lip shut, Voice^ 
 Nose." In forming this sound the lips are shut and the voice is
 
 CHART 11- 
 
 ) U O C 
 
 3) Ci,) (T) € 
 
 D O O Cs 
 
 ^ Ci^ CD Gs 
 
 D D Q O 
 
 3 Q CI Q 
 
 ? 
 
 ) O Q ( 
 
 s 
 
 B G^ CD G
 
 58 
 
 passed through the nose. She places one of the pupil's hands 
 against her throat, and the other against her nose, and produces the 
 sound of the letter in. The pupil sees the closure of the lips and 
 feels a vibration in the throat and nose. 
 
 2. The second symbol in the eighth line, "Point shut, Voice, 
 Nose," represents the position of the organs in forming the letter n. 
 The pupil sees the point of the tongue shut against the upper gum 
 and feels a vibration in the throat and nose. 
 
 4. The last symbol in the eighth line, "Back shut. Voice, 
 Nose," expresses the position of the organs when producing iig in 
 such a word as sing. Here the pupil sees that the back of the 
 tongue is raised, and feels a vibration in the throat and nose. The 
 objed of this exemplification is simply to make the pupils under- 
 stand what the symbols mean, and not to get them to make the 
 sounds themselves. Still, the children generally try to imitate what 
 the teacher does, and of course, in some cases they fail because they 
 have not yet acquired control over their vocal organs. As it is not 
 the objed of their teacher at this stage to cause the pupil to make 
 sounds, she should not take any notice of their f^iilures for fear of 
 discouraging them. She should be satisfied with evidences of com- 
 prehension as to the meaning of the symbols. Most children are 
 able to take Charts 1. and II. in one lesson. After reviewing these 
 at a subsequent time the third Chart is explained. 
 
 CHART 111. 
 
 The pupil's attention is directed to the symbol "Lip centre- 
 aperture" (see the first symbol in Chart II.), which he describes by 
 touching the under lip and then holding the thumb and forefinger 
 close together without touching. The teacher then directs attention 
 to the mouth, and shows that there is only one small hole through 
 which the air passes. She then holds her lips together in the mid- 
 dle and allows air to escape through two side apertures, one at each 
 corner of the mouth, showing the pupil that now there are two 
 holes through which the air escapes instead of one. This fad she 
 symbolizes by writing two "Lip centre-aperture " symbols one above 
 the other, ^ thus, forming a charader somewhat like the Arabic 
 numeral 3. This the pupil describes by touching his lip, and then 
 holding near the thumb two tlngers, instead of one alone, indicating 
 that the aperture is divided into two parts. Thus the thumb and 
 forefiiiger held together indicate one central aperture, and the thumb 
 held near th'^ fore and middle fingers indicates "divided aperture." 
 
 Turning now to Chart III. the symbols are described as follows:
 
 59 
 
 First line. — i. Lip divided-aperture. 2. Point divided-aperture, 
 
 3. Front divided-aperture. 4. Back divided-aperture 
 Second line. — i. Lip divided-aperture, Voice. 2. Point di- 
 vided-aperture, Voice. 3. Front divided-aperture. Voice. 
 
 4. Back divided-aperture, Voice. 
 
 The second symbol in the second line, " Point divided-aperture, 
 Voice," expresses the position of the tongue in forming the sound 
 of / in such a word as love. The point of the tongue is placed 
 against the upper gum, and the voice is passed through two side 
 apertures, one on each side of the tongue. The symbols in the 
 third, fourth, fifth, and sixth lines are what my father terms "mixed" 
 symbols, involving two positions of the organs assumed simulta- 
 neously. The first svmbol in the the third line is composed of a 
 large "Lip centre-aperture" symbol with a small "Back centre- 
 aperture " hooked on to one end of the curve. For the sake of 
 symmetry another small " Back centre-aperture " is attached to the 
 other end of the curve, but this has no organic significance. This 
 compound symbol expresses the position of the organs in sounding 
 the English element represented by the letters wh in such a word as 
 ■whistle. The back of the tongue is in the position for the German 
 ch (Back centre-aperture), while at the same time a small centre- 
 aperture is formed by the lips. The labial aperture being more 
 obstrudive than the back aperture, characterizes the sound as a 
 labial letter. For this reason the " Lip centre-aperture sign is made 
 the most prominent part of the compound symbol. Deaf pupils 
 describe this symbol as "Lip centre-aperture, Back centre-aper- 
 ture." 
 
 Proceeding now with the description of the remaining symbols 
 upon Chart 111. we have: — 
 
 Third line. — i. Lip ceritre-aperture, Back centre-aperture. 2. 
 Point centre-aperture. Front centre-aperture. 3. Front 
 centre-aperture, Point centre-aperture. 4. Back centre- 
 aperture, Lip centre-aperture. 
 Fourth line. — i. Lip centre-aperture, Back centre-aperture, 
 Voice. 2. Point centre-aperture, Front centre-aperture. 
 Voice. 3. Front centre-aperture, Point centre-aperture, 
 Voice. 4. Back centre-aperture. Lip centre-aperture, Voice. 
 Fifth line. — i. Lip divided-aperture. Back centre-aperture. 2. 
 Point divided-aperture. Front centre-aperture. 3. Front 
 divided-aperture, Point centre-aperture. 4. Back divided- 
 aperture. Lip centre-aperture.
 
 6o 
 
 Sixth line. — i. Lip divided-aperture, Back centre-aperture, 
 
 Voice. 2. Point divided-aperture, Front centre-aperture, 
 
 Voice. 3. Front divided-aperture, Point centre-aperture. 
 
 Voice. 4. Back divided-aperture. Lip centre-aperture, 
 
 Voice. 
 
 Numerous other compound symbols might be built up out of 
 
 the elementary signs shown in Chart I., expressing both possible and 
 
 impossible positions of the organs. The forms shown in Charts II. 
 
 and III. are not intended to be pronounced by the pupil, but are 
 
 given simply as exercises in analysis. If the pupil can be made 
 
 to understand the meaning of the compound symbols by analyzing 
 
 them into their elementary forms, Visible Speech become a symbolic 
 
 language, whereby any imaginable position of the vocal organs may 
 
 be expressed, so as to be understood by the children. 
 
 The remaining symbols on Chart III. seventh line, are throat 
 symbols. They picture various conditions of the glottis. 
 
 1. The first charafter, shaped like the letter O, piftures a wide 
 aperture in the throat. The vocal cords are wide apart, leaving a 
 large opening between them through which air may freely pass 
 without obstrudion. This is the condition of the glottis in uttering 
 the letter h, and all non-vocal or breath consonants. The letter h 
 may, indeed, be considered as the non-vocal or breath form of a 
 vowel. It has just as many different sounds as there are vowels. 
 Pronounce such words as he, hay, ha, hoe, and who; it will be 
 observed that the mouth-position for the sound of h is different in 
 each word. H only occurs as an element of speech before a vowel. 
 Under such circumstances the mouth position for h is the same as 
 for the succeeding vowel, but the opening in the glottis is so wide 
 as to allow the breath to pass into the mouth without sensible 
 obstruction in the throat. 
 
 2, The second symbol in the seventh line, pictures a smaller 
 aperture in the throat than the first. The vocal cords are brought 
 near enough together to obstruft in some degree the passage of air 
 between them, giving rise to a rustling sort of sound which is uni- 
 versally denominated "whisper." This is the condition of the 
 glottis when we whisper vowel sounds. This position of the 
 throat also may be assumed in uttering consonants, thus giving rise 
 to the "whispered" consonants, which in some languages are 
 significant elements of speech, quite distinct in meaning from the 
 "breath " and " voiced " consonants of similar formation occurring 
 in the same languages.
 
 6i 
 
 3. We have already become familiar with the third symbol in 
 the seventh line, as the representative of voice. It pictures a still 
 smaller aperture in the throat than either of the preceding. The 
 vocal cords are placed parallel to one another, and the aperture 
 between themjs reduced to a mere slit (pictured by a straight line). 
 In this condition of the glottis the passage of air through the slit- 
 like aperture occasions a vibration of the vocal cords, producing 
 voice. This is the condition of the glottis in uttering vocal conso- 
 nants and vowels. 
 
 4. The fourth symbol in the seventh line, piftures complete 
 closure of the glottis. The vocal cords are pressed together so as 
 completely shut the aperture between them, and prevent the escape 
 of air. This is the condition of the glottis aimed at by singers in 
 pra6licing what is called the "coup de glotte.'' It also occurs as an 
 element of speech in certain diale(?ts. For example: In the Scotch 
 dialeft as spoken in Glasgow, "Throat shut" is substituted for/ 
 (Point shut) in such words as butter, water, etc. In English also it 
 occurs as an unrecognized element of speech in words commencing 
 with vowels. In ordinary utterance every syllable really commences 
 with a consonant. When words are supposed to begin with 
 vowels, the "Throat shut" consonant really precedes the vowel 
 sound, although it is not usually recognized as an element of speech 
 by orthoepists. Pronounce with considerable force the names of the 
 five vowel letters a, e, i,o, u. A closure of the glottis takes place 
 before each vowel, excepting the last. The "Throat shut" conso- 
 nant precedes the vowels a, e, i, and 0; but 11 is preceded by the 
 consonant y. Indeed, the name of the vowel might have been 
 spelled you without affefting the pronunciation. The ''Throat 
 shut" consonant, followed by a forcible emission of air from the 
 lungs, is familiar to every one in the form of a cough. 
 
 The meaning of the throat symbols shown in the seventh line, 
 is explained to deaf children in the following way: 
 
 1. Touch the throat, and then hold the two hands together palm 
 to palm, curving the fingers so as to cause the space between the 
 hands to assume the shape of the first symbol. The idea to be 
 conveyed is, that the aperture in the throat is somewhat of that 
 shape, and very large. 
 
 2. Touch the throat, and then hold the hands together palm to 
 palm, as before, but reduce the space between the hands so as to 
 cause the aperture to assume the shape of the second symbol. The
 
 CHART III 
 
 3 U o C 
 
 3 GJ rn C 
 
 D V n C 
 
 3 W n 
 
 D (J n 
 
 
 
 c3 gJ r^ e 
 
 O I
 
 63 
 
 idea to be conveyed, is that the aperture in the throat is more con- 
 trafted than in the former case. 
 
 3. Touch the throat, and hold the hands together palm to palm, 
 as before, so that the aperture between the hands is reduced to a 
 mere slit. At the same time give a quivering or trembling motion 
 to the hands. The idea to be conveyed, is that the aperture in the 
 throat is a mere slit, and that a trembling or quivering motion occurs 
 in the throat which the pupil may perceive for himself by placing 
 his hand upon the teacher's throat while the teacher produces voice. 
 
 4. Touch the throat, and then press the two hands together 
 palm to palm, with a shutting adion, causing the hands to assume 
 the appearance of the fourth symbol in the seventh line. 
 
 We may translate these gestures into words, and give names to 
 these symbols, in the following manner: — 
 
 Seventh line. — i. Throat open. 2. Throat contrafted. 3. 
 Throat a-slit (Voice). 4. Throat shut. 
 
 CHART IV. 
 
 When we compare the symbols shown on Charts II. and III. 
 with those on Chart V., we notice a radical difference between them. 
 The most prominent feature of the symbols on Charts II. and III. is 
 a curve of some sort, whereas the chara6teristic of those on Chart 
 V. is a straight line. By reference to Chart I. it will be seen that a 
 curve is indicative of some part of the mouth, and that a straight 
 line represents voice. The symbols on Charts II. and III. represent 
 positions of the organs that yield consonant, and those on Chart V., 
 positions that yield vowel, sounds. The generic difference between 
 consonants and vowels is thus portrayed in the symbols. In conso- 
 nant symbols the mouth position is made the charaderistic feature 
 of the symbol, the voice where it occurs being written subordin- 
 ately by a straight line within the curve. In vowel symbols, on the 
 other hand, the voice sign is made the charaderistic feature, and the 
 mouth position is represented subordinately by curves, or dots or 
 other marks appended to the voice line. Chart IV. is used for the 
 purpose of explaining to deaf children the meaning of these appen- 
 dages. The chief parts of the tongue employed in forming vowel 
 sounds are the back and the front parts of the tongue. When we 
 drawa vertical line centrally through the diagram on Chart IV., we 
 find that a dot or other mark on the right-hand side of the line rests on 
 the front part of the tongue, whereas a mark on the left-hand side 
 of the line rests on the back of the tongue. In vowel symbols a
 
 •CHART IV 
 
 1 1 J nx L r i
 
 65 
 
 mark on the right-hand side of the voice line indicates the front part 
 of the tongue, a mark on the left indicates the back of the tongue, 
 and a short horizontal line drawn across the vowel stem indicates 
 that the lips aie employed. Thus the symbols at the bottom of 
 Chart IV. indicate (i) the voice; (2) the back of the tongue; (3) 
 the back of the tongue; (4) both back and front of the tongue 
 used simultaneously; [this is what my father terms a "mixed' 
 position]. (5) Back and front ["mixed"]; (6) back and front 
 ["mixed"]; (7) the front of the tongue; (8) the front of the 
 tongue; (9) the lips. 
 
 it will be observed that the appendages are placed sometimes 
 at the top of the vowel stem, sometimes at the bottom, and some- 
 times at both ends. This pictures the elevation of the tongue in the 
 mouth. When the mark is at the top of the vowel stem the part 
 of the tongue indicated is placed high up in the mouth, leaving a 
 small aperture between the tongue and the palate; when the mark 
 is at the bottom the tongue is low with a large aperture; and when 
 the mark is at both ends the tongue occupies an intermediate posi- 
 tion with an intermediate aperture. Reading again the symbols at 
 the bottom of Chart IV., we have (i) the voice; (2) back of the 
 tongue high ; (3) back of the tongue low ; (4) back and front both 
 high ["high mixed"]; (s) back and front both mid positions ["mid 
 mixed"]; (6) back and front both low ["low mixed"]; (7) front 
 low ; (8) front high ; (9) this symbol means not only that the lips 
 are used but that the aperture between them is of a rounded form. 
 
 The deaf child is taught to indicate the small aperture formed 
 by the high position of the tongue, by holding his thumb and fore- 
 finger close together without touching. (This is the same sign for- 
 merly described as meaning "centre aperture.") The low tongue 
 position with large aperture, is indicated by holding the finger and 
 thumb far apart. ; and the intermediate position is represented by a 
 half-way position of the thumb and forefinger. Thus, degrees of 
 aperture are indicated by degrees of separation of the thumb and 
 the forefinger. 
 
 We are now prepared to analyze the symbols on Chart V. 
 
 CHART v.— Vowels. 
 
 The vowels on Chart V. may be divided into four groups of 
 nine symbols each : — 
 
 First Group. — Primary bowels. 
 
 Reading downwards we have: —
 
 First line.— 1. High Back. 2. Mid Back. 3. Low Back. 
 Second line. — i. High Mixed. 2. Mid Mixed. 3. Low Mixed. 
 Third line. — i. High Front. 2. Mid Front. 3. Low Front. 
 
 Second Group. — Wide yowels. 
 
 Reading downwards we have: — 
 
 First line. — i. High Back Wide. 2. Mid Back Wide. 3. Low 
 
 Back Wide. 
 Second line. — i. High mixed Wide. 2. Mid Mixed Wide. 3. 
 
 Low Mixed Wide. 
 Third line. — i. High Front Wide. 2. Mid Front Wide. 3. 
 
 Low Front Wide, 
 
 Third Group. — Primary Round yowels. 
 
 Reading downwards we have: — 
 
 First line. — i. High Back Round. 2. Mid Back Round. 3. Low 
 
 Back Round. 
 Second line. — i. High Mixed Round. 2. Mid Mixed Round. 3. 
 
 Low Mixed Round. 
 Third Line. — i. High Front Round. 2. Mid Front Round. 3. 
 
 Low Front Round. 
 
 Fourth Group. — Wide Round Vowels. 
 
 Reading downwards we have: — 
 
 First line. — i. High Back Wide Round. 2. Mid Back Wide 
 
 Round. 3. Low Back Wide Round. 
 Second line. — i. High Mixed Wide Round. 2. Mid Mixed 
 
 Wide Round. 3. Low Mixed Wide Round. 
 Third line. — i. High Front Wide Round. 2. Mid Front Wide 
 Round. 3. Low Front Wide Round. 
 ■ Wide vowels differ from primary vowels by a slight widening 
 of the oral passage; for example: Contrast the "high front" vowel 
 {ea in the word eat), with the "high front wide" vowel (/in the 
 word //). The oral passage for the latter is slightly larger than for f^, 
 and Prof. Melville Bell believes also that the back part of the mouth, 
 or the cavity of the pharynx, is more expanded in wide vowels 
 than in primary. Widening the oral passage is indicated by a hook 
 instead of a dot. Groups 111. and IV. are rounded vowels, that is, 
 the passage between the lips is of a rounded form. 
 
 Deaf children describe these symbols by using the signs already 
 mentioned in describing Chart IV., and we may translate their signs 
 for the symbols on Chart V. as follows:
 
 67 
 
 First Group. — Primary Vowels. 
 
 Reading downwards we have: — 
 
 First line. — i. Voice, Back small-aperture. 2. Voice, Back 
 
 mid-aperture. 3. Voice, Back large-aperture. 
 Second line. — i. Voice, Back small-aperture, Front small-aper- 
 ture. 2. Voice, Back mid-aperture. Front mid-aperture. 
 3. Voice, Back large-aperture, Front large-aperture. 
 Third line. — i. Voice, Front small-aperture. 2. Voice, Front 
 mid-aperture. 3. Voice, Front large-aperture. 
 Second Group. — Wide Vowels. 
 
 In teaching deaf children, the symbols of this group are con- 
 sidered as identical with those of Group 1., and are described in the 
 same manner. When the pupils have become familiar with the 
 analysis of Visible Speech symbols, they are shown, by means of 
 the thumb and forefinger, that the position symbolized in Group II. 
 have a slightly wider aperture than the corresponding positions in 
 Group I. 
 
 Prof. Melville Bell's conception of the expansion of the pharynx 
 during the utterance of wide vowels, is a difficult one to convey to 
 deaf children who know no language; I have, therefore, not at- 
 tempted to do more than convey the idea that the mouth passage 
 for wide vowels, is slightly wider than for primary vowels, so that 
 the primary and wide symbols, taken together, represent six degrees 
 of aperture; for example: Take the front vowels, commencing 
 with the smallest aperture and ending with the largest, we have the 
 following series of apertures: — 
 
 1. High Front. 
 
 2. High Front Wide. 
 
 3. Mid Front. 
 
 4. Mid Front Wide. 
 
 5. Low Front. 
 
 6. Low Front Wide. 
 
 Third Group. — Primary Round Vowels. 
 
 Reading downwards wc have: — 
 
 First line. — 1. Voice, Back small-aperture, Lip, small-aperture. 
 2. Voice, Back mid-aperture, Lip mid-aperture. 3. Voice, 
 Back large-apertuie, Lip large-aperture. 
 Second line. — i. Voice, Back small-aperture, Front small-aper- 
 ture, Lip small-aperture. 2. Voice, Back mid-aperture, 
 Front mid-apertuie, Lip mid-apertuic. 3. Voice, Back 
 large-aperturc, Front large-aperture. Lip large-aperture.
 
 CHART V 
 
 nt Jit
 
 69 
 
 Third line. — I. Voice, Front small-aperture, Lip small-aperture. 
 
 2. Voice, Front mid-aperture, Lip mid-aperture, 3. Voice, 
 
 Front large-aperture, Lip large-aperture. 
 
 The labial apertures described are of a rounded form, but as the 
 
 pupils can see for themselves the shape of the labia! apertures, it 
 
 has not been considered necessary to give them a distinct sign for 
 
 a rounded aperture; they simply describe the size of aperture by 
 
 the separation of finger and thumb. 
 
 Fourth Group. — 'Vide Round bowels. 
 
 In teaching deaf children, the symbols of this group are con- 
 sidered as identical with those of Group 111., and are described in a 
 similar manner. The differences are explained later on. The sym- 
 bols of Group IV. bear the same relation to those of Group III., that 
 the symbols of Group 11. bear to those of Group I. (See note above 
 relating to Group II. 
 
 CHART VL 
 
 Chart VI. shows the mechanism or the English consonants as 
 explained to the deaf. 
 
 First line: — 
 
 (1) "Lip shut," followed by a "puff of air." We have here 
 two symbols, the first of which (Lip shut), represents/), as in put, 
 cup, etc. It is not advisable to teach "shut" consonants as sepa- 
 rate elements. They are best taught in connexion with vowels. 
 The most elementary form of p taught, is the final p, as in cup, 
 where the "Lip shut" position is followed by a puff of air, as 
 shown in the Chart. 
 
 (2) "Lip shut, Voice," followed by "voice." The first of 
 these symbols (Lip shut, voice), represents h in but, cub, etc. This 
 is not taught elementarily, but in connexion with a vo.wel. The 
 simplest form is that shown in the Chart where the "Lip shut, 
 Voice" position is followed by an indefinite murmur of voice, form- 
 ing a syllable somewhat like bir in bird. 
 
 (3) "Lip shut, Voice, Nose," represents m in man, come, etc. 
 
 (4) " Lip divided-aperture," represents /in file, luff, etc. The 
 upper organ in this case is the edge ot the teeth, instead of the 
 upper lip. 
 
 Second line: — 
 
 (I) " Point shut," followed by "a puff of air." The first sym- 
 bol (Point shut), represents / as in to, not, etc. When / occurs as a 
 final letter, as in not, the "Point shut" position is followed by a 
 purf of air, as shown in the Chart.
 
 70 
 
 (2) "Point shut, Voice," followed by "Voice." The first 
 symbol (Point shut, Voice), represents d, as in do, nod, etc. In the 
 symbols shown in the Chart, the "Point shut. Voice" position is 
 followed by an indefinite murmur of voice, thus representing a syl- 
 lable somewhat like dir in dirk. 
 
 (3) " Point shut. Voice, Nose " represents n, as in no, nun, etc. 
 
 (4) "Lip divided-aperture. Voice" represents v, as in vie, 
 love, etc. 
 
 Third line: — 
 
 (i) " Back shut" followed by a "puff of air." The first sym~ 
 bol (Back shut), represents k, as in key, sick, etc. When k occurs 
 as a final letter, as in sick, the " Back shut" position is followed by 
 a puff of air, as shown in the Chart. 
 
 (2) "Back shut, Voice," followed by "Voice." The first o 
 these symbols ( Back shut. Voice, represents g, as in go, log, etc. 
 The " Back shut, Voice" position is followed by an indefinite mur- 
 mur of Voice, forming a syllable somewhat like gir in girl. 
 
 (3) "Back shut Voice, Nose," represents ng, as in limg, 
 tongue, etc. 
 
 (4) " Lip centre-aperture, Back centre-aperture," represents wh, 
 as in whet. It is taught to the deaf as "Back centre-aperture" 
 (German ch), with the lips rounded as in the aft of whistling. In 
 obtaining this sound from a deaf child, it is found essential to direct 
 attention to the " Back centre-aperture position." 
 
 Fourth line : — 
 
 (i) "Point divided-aperture, Voice" represents /, in /////. 
 
 (2) "Point divided-aperture. Front centre-aperture " represents 
 //? as in thin, kith, etc. 
 
 (3) " Point divided-aperture. Front centre-aperture. Voice " rep- 
 resents th as in then, with, etc. 
 
 (4) "Lip centre-aperture, Back centre-aperture, Voice" repre- 
 sents zv in the word ivet. In teaching the deaf it is essential to direft 
 attention to the "Back centre-aperture" position, and the sound is 
 taught as identical with the vowel 00 in pool. 
 
 Fifth line: — 
 
 (i) "Point centre-aperture, Front centre-aperture" represents 
 5. as in sown, hiss, etc. 
 
 (2) "Point centre-aperture, Front centre-aperture, Voice, rep- 
 resents I in ^one, and 5 in his.
 
 CHART VI 
 
 C> 31 B 3 
 
 0> 01 ® 3 
 
 o> ai G D 
 
 o o on 
 
 (jO U (jO 3 
 
 13 lij ^ C^ 
 
 \m/ \^ GO O O 
 
 O (T) U CO 
 
 O
 
 72 
 
 (3) "Front centre-aperture. Point centre-aperture" represents 
 sh as in she, and 5, in assure. It also occurs after " Point shut" in 
 such a word as church (tshurtsh). 
 
 (4) "Front centre-aperture, Point centre-aperture, Voice" rep- 
 resents 5 in measure, and { in a:(ure. It is heard in / and in g soft, 
 after " Point shut Voice" in such a word as jud^e (dzhudzh). 
 Si'xlh line: — 
 
 (i) " Front centre-aperture " represents the sound given to h in 
 hue, for which we have no letter. As it is the non-vocal form of 
 the consonant y, it may be represented by yh. It also occurs after 
 non-vocal consonants, as in few {{yhoo), tune (tv^oon), cue 
 ( kyhoo). 
 
 (2) "Front centre-aperture. Voice" represents ji^ in the word 
 vou. In teaching the deaf it is considered as identical with the 
 vowel e. 
 
 (3) "Point centre-aperture." This sound has no letter, and 
 may be represented by rh, as it is the non-vocal form of r. The 
 deaf are taught that the letter r has this sound when it comes 
 after a non-vocal consonant, as in pry (prhy), try (trhy), cry 
 ( crhy). 
 
 (4) "Point centre-aperture, Voice" represents r in such a 
 word as run, also r after a vocal consonant, bride, dry, etc. 
 Seventh line: — 
 
 "Throat large-aperture" represents h in such words heat, 
 hit, hate, head, hat, hoot, hook, hope, hall, hot, half, hurl, hut» 
 high, how, hoist, etc. 
 
 CHART VII. 
 
 The symbols in Chart Vll. represent the positions for English 
 vowel sounds. 
 First line: — 
 
 (i) "High Back Wide, Round " represents the vowel heard in 
 the following words : foot, put, etc. 
 
 (2) "High Back, Round" represents the vowel heard in pool, 
 move, through, true, flew, etc. 
 
 (3) "High Front represents the vowel heard in eel, eat, field, 
 hey, sei^e, etc. 
 
 (4) "High Front Wide" represents the vowel heard in ill, 
 build, etc. 
 
 Second line: — 
 
 (i) " Mid Back, Round," followed by a glide towards "High 
 Back, Round," represents the diphthongal vowel heard in pole, coal, 
 soul, dough, bowl, etc.
 
 73 
 
 (2) "Mid Front," followed by a glide towards "High Front," 
 r presents the diphthongal vowei heard in ale, ail, eight, great, lay, 
 they, etc. 
 
 Third line: — 
 
 (i) "Low Back Wide, Round" represents the vowel heard in 
 doll, what, etc. 
 
 (2) " Low Back, Round " represents the vowel heard in all, 
 paul, paiv, thought, etc. 
 
 (3) " Low Front " represents the vowel heard in shell, head, 
 said, etc. 
 
 (4) " Low Front Wide " represents the vowel heard in shall, hat, 
 can, and, etc. 
 
 Fourth line: — 
 
 (i) "Low Back Wide" represents the vowel heard in ah, 
 father, etc. 
 
 (2) "Mid Back Wide" represents the vowel heard in ask, 
 path, etc. 
 
 (3) "Low Mixed Wide " represents the vowel heard in her, 
 pearl, girl, fur, etc. 
 
 (4) "Mid Back" represents the vowel heard in gull, come, 
 rough, etc. 
 
 Fifth line: — 
 
 (i) " Mid Back Wide," followed by a glide towards "High 
 Front," represents the diphthongal vowel heard in pile, sleight, buy, 
 eye, etc. 
 
 (2) "Mid Back Wide," followed by a glide towards "High Back 
 Round," represents the diphthongal vowel heard in cow, bough, 
 round, etc. 
 
 (5) "Low Back Round, " followed by a glide towards "High 
 Front," represents the diphthongal vowel heard in oil, boy, etc. 
 
 The sound of h only occurs before a vowel, and it is advisable 
 to give the deaf pupil the idea that there are as many sounds of h as 
 there are vowel sounds. Defective pronunciation results from the 
 attempt to give a uniform value to the sound. The deaf pupil is 
 taught that the mouth position for h is always the same as that of 
 the succeeding vowel; in faft, that h is the breath form of the 
 succeeding vowel. For example: Contrast h in he, with that in
 
 CHART VII
 
 75 
 
 who (hoo). In the former case the mouth position lor h is the 
 same as that for the vowel ee, in the latter it is the same as that for 
 the vowel oo. 
 
 Do you describe the word "Mixed?" 
 
 We do not use the word "Mixed" in teaching the deaf, but 
 describe in detail the positions that are mixed. Thus, we describe 
 my father's "Lip-mixed" consonant as "Lip centre-aperture. Back 
 centre-aperture." 
 
 What do you mean by "divided-aperture?" 
 
 An aperture divided in the middle so as to leave two orifices. 
 For example: In assuming the position for the letter / (CO), the point 
 of the tongue is placed against the upper gum, and the air passes 
 out over both sides of the tongue. 
 
 Can a person realize by any feeling the muscular condition 
 represented by your symbols? 
 
 Certainly. Familiarity with the organs through the use of a 
 mirror leads to a perception of muscular feeling of the positions 
 assumed by the vocal organs. Indeed, in talking we are all guided 
 more or less by muscular feeling. For example, we can talk with- 
 out making any noise so that a deaf pupil can understand what we 
 say by watching the mouth. How do we know that our vocal 
 organs are in the correft position when we make no sound? Surely 
 by muscular feeling. The deaf child also, through training, becomes 
 conscious of the movements of his vocal organs and can tell by 
 muscular feeling exactly what he does. 
 
 Why do you begin with lip positions instead of back positions? 
 
 The lip positions are the most visible. The deaf child under- 
 stands what the symbols mean when applied to the lips, because he 
 can see the positions assumed. This knowledge he applies to the 
 interior positions that cannot be so easily seen. 
 
 Now in teaching a deaf child you present to him the symbol 
 for some difficult sound. If he has been taught to analyze the sym- 
 bols in the manner shown, the symbol conveys to his mind a 
 direction what to do with his mouth. That is what your pupil has 
 to aim at, but in ninety-nine cases out of a hundred he may not 
 get it, at least at the first shot. Now what are you going to do? 
 Are you going to say "No, no! that's not right. Try again?" Let 
 him try once more and the chances are that he fails again to give 
 the sound intended. The No-No method only aggravates the 
 difficulty by discouraging the pupil and disgusting him with 
 articulation.
 
 76 
 
 Learning to speak is like learning to shoot. Now, suppose you 
 aim at a target for the first time, and fail to hit it, and you are 
 simply told "No, no; that's not right. Try again." Well, suppose 
 you do try again. The chances are that you fail, and if you were 
 simply told once more that you didn't hit the bull's eye, you are no 
 further advanced than you were before. That's not the way to 
 learn to shoot. You must know where your bullet struck when you 
 failed, so as to see the relation between the point struck and the 
 point you intended to hit. The knowledge of that relation will 
 guide your next shot. For example: If you know that you hit too 
 far to the right, why your next shot is aimed more to the left, and 
 perhaps flies clear of the target on the other side. If, then, you are 
 told the result of this shot also, you make due allowances the next 
 time you try. You may fail a hundred times. Now you go a little 
 to the right, now a little to the left, sometimes too high, sometimes 
 too low; but your knowledge of the elTeft of each shot causes you 
 to make unconscious allowances, so that, little by little, you come 
 nearer the bull's eye until at last you hit it. When you can hit the 
 bull's eye every time, you have mastered your instrument — the gun — 
 and can hit any other bull's eye with equal ease. 
 
 The No-No method, besides discouraging the beginner fails to 
 give the very information that is necessary for his progress. The 
 deaf child must know what he did when he failed, and the relation 
 of the position struck to the bull's eye. The knowledge of that 
 relation will guide him in his next attempt. For example: If he 
 knows that his tongue was too far forward in the mouth, in his 
 next attempt he aims at having his tongue further back and probably 
 gets too f^ir in that direction. If, then, he is told the result of this 
 attempt also, he makes due allowances the next time he tries. He 
 may fail a hundred times. Now the position may be a little too far 
 forward, now a little too far back, or the tongue may be too high or 
 too low, but his knowledge of the effect of each effort causes him 
 to approach more and more closely to the exaft position desired, till 
 at last he gets it. The time spent in studying and representing the 
 incorreft positions is not wasted, for it gives the pupil mastery over 
 the instrument of speech itself, and the struggle to get exaditude of 
 position with one difficult sound gives him power to get any other, 
 just as the ability to hit one bull's eye qualifies a man to shoot at 
 any mark.
 
 77 
 
 I will now answer some of the questions that have been pro- 
 pounded to me. 
 
 The first question is: What is accent? 
 
 I have found in my past experience that accent is length. At 
 least that we get a more natural effect from a deaf child if we give 
 him the idea of making the accented syllable longer than the others 
 rather than louder. The attempt to make the accented syllables 
 louder often leads to a jerky utterance very unlike the effeft we desire. 
 
 To make my meaning clear 1 will say that 1 do not think 
 that we give a jerk of the abdominal muscles for every accented 
 syllable, any more than the piper gives a jerk of the arm to mark 
 the accented notes. The pressure upon the bag is continuous 
 and the rhythm of the music is brought out by the dif- 
 fering durations of the notes. Of course the music may be 
 made louder or softer by increasing or diminishing the pressure 
 upon the bag, but this effeft corresponds more to emphasis than 
 to accent. 
 
 The fa£t that the effeft of accent can be produced by length- 
 ening the duration of a syllable, without any change in the loud- 
 ness or volume of the voice may be demonstrated by a simple 
 experiment. 
 
 Let a deaf child prolong the voice while you manipulate his 
 lips so as to produce B\3\B\B\, etc. (Mama, mama, etc.) Now, 
 although he makes no variation in the loudness of his voice you 
 can with your fingers produce the effect of accent by prolonging 
 the syllable you desire to bring out. For example: If you. prolong 
 the open position you can make him say 9133 SiB], etc. (Mama, 
 pronounced in the English way with the accent on the second syl- 
 lable) or 93l 9391, etc., with the accent on the first syllable, as is 
 very commonly heard in America. 
 
 "But," you may ask, "May not syllables containing short 
 vowels be accented, and how can you prolong the syllable if the 
 vowel is short ?" 
 
 Certainly, syllables of this kind can be accented. In such cases 
 you do not prolong the vowel but the succeeding consonant. For 
 example: You can by manipulation cause your pupil to say 91939 
 91939, etc., (mamum, mamum, etc.), by prolonging the closed 
 position of the lips instead of the open. 
 
 There is a great point here. Short vowels are succeeded by 
 long consonants; for example: the consonant position is retained
 
 78 
 
 for a much longer time in such words as come, cuff, it, sin, look, 
 than in such words as, calm, calf, eat, seen, Luke. You can dem- 
 onstrate this prolongation of the consonant position by emphasizing 
 with great force the word " not " in the familiar quotation, "To be, 
 or NOT to be; that is the question." The hiatus caused by 
 the prolongation of the shut position (D ) of the / is so great as to 
 occasion a perceptible silence in the midst of the sentence. 
 
 Let your pupils pronounce with precision the accented syllables 
 of words and slur over the others, articulating them rapidly, with 
 indefinite vowels, and the effeft will be much more natural than a 
 precise articulation of every syllable with loudness for accent. 
 
 Miss McCowen: Have you ever thought of there being a differ- 
 ence in pitch ? 
 
 Dr. Bell: 1 don't see how that could come into the case. In 
 natural speech the pitch of the voice is constantly varied in both 
 accented and unaccented syllables. You can distinguish the element 
 of accent even though the voice is on a monotone and without 
 variation of loudness. 
 
 The next question on my list is: " Please imitate Helen Keller's 
 voice." 
 
 I am sorry 1 have not a sufficiently distinft recollection to do 
 that. Perhaps Miss Fuller can. 
 
 Miss Fuller: I don't think I can. 
 
 Dr. Bell: 1 will say it was distind; and perfedly intelligible to 
 persons not accustomed to the deaf 
 
 The next question on my list is: "How would you teach r 
 and/.?" 
 
 1 have experienced so little difficulty in teaching r and / (Ci). 
 and CO) that 1 am inclined to think that the very defecJtive manner in 
 which these sounds are sometimes given by deaf children must be 
 due to the mode of teaching. 1 always commence with the non- 
 vocal forms U and CO. 1 would recommend you to adopt this plan 
 generally in dealing with consonants as the non-vocal forms are 
 usually more easily acquired than the others. It is especially advis- 
 able in the case of r and /, as you then avoid the common fault of 
 too large aperture. When the non-vocal forms have been mastered, 
 the vocal forms follow as a matter of course by the addition of voice 
 
 In forming O (non-vocal r) the point of the tongue is applied 
 to the upper gum just at the part where the palate begins to arch, 
 and the breath is allowed to escape through a central aperture. Deaf
 
 79 
 
 children acquire the sound very readily by imitation, as the whole 
 mechanism can be seen in a mirror. In any case of difficulty 1 
 manipulate the sound from U (fh) or ^ (5) in the manner described 
 to you the other day. It is quite unnecessary to attempt a "trill" 
 that is, to cause a vibration or trembling of the point of the tongue. 
 Such an effect is un-English. 
 
 The defecftive variety of r most commonly met with results 
 from placing the point of the tongue too far back in the mouth. For 
 example: It is often applied to the top of the hard palate (Ci)"); and 
 in some cases it is coiled up within the mouth so as to approximate 
 the soft palate (OJ^). 
 
 The defective form of / so common in schools for the deaf 
 results from an exaggerated narrowing of the tongue (COv) (too large 
 aperture) and from opening the jaws too widely. 
 
 Sometimes the tip of the tongue is placed against the upper 
 teeth and the under side of the tongue is actually protruded from the 
 mouth. 
 
 The correct position for CO (/) is so nearly the same as that for 
 CC (/;) that the deaf have difficulty in distinguishing one from the 
 other by the eye. This leads some children to substitute CO for CO (n 
 for /). 
 
 In forming CO (/), the point of the tongue should be placed 
 against the upper gum, and air allowed to escape through two side 
 apertures. The lingual position for CD (//) is the same, excepting 
 that the two side apertures are closed. Thus, the tongue appears 
 slightly broader for CC ( // ) than for CO (/). The exaggerated narrow- 
 ing of the tongue so commonly associated with / results in side 
 apertures that are much too large. This defed is avoided, if you 
 commence by teaching the sound non-vocally (CO) with quite small 
 side apertures. Even pupils who give n for / readily acquire the 
 non-vocal form (CO). After this has once been mastered, the vocal 
 form follows as a matter of course by the addition of voice (CO). 
 
 Even though CO and CO ( r and /) may be given corretWy as ele- 
 mentary sounds, deaf children produce an effect that is not heard in 
 ordinary utterance when they attempt to give these sounds after 
 non-vocal consonants. Thus, DCOl (tree), QCOrD> (sleep), etc., 
 sound as if there were two syllables in each word, and if any defed 
 of combination exist, the vocality of the r or / causes the introduc- 
 tion of a voice glide after the non-vocal consonant. Thus, CClCOl, 
 QlCOlD> (turee, sCileep), etc.
 
 8o 
 
 I have found that deaf children give the proper vernacular effedt 
 when they attempt to make the sounds non-vocally. Thus, DUl, 
 U2jrD> (tree, sleep). The same may be said of the sounds 59 and 
 (^ (iv and J') after non-vocal consonants; for example, when deaf 
 children attempt to sav DJ3tG3Dr. QC^i (twenty, cue, /. e., kyou), 
 they are apt to give DilCCOf, Oli (too-enty, kee-oo), although the 
 distindtion between the consonant and vowel forms J3— i and (f)—L 
 (w-oo and y-ee) may have been fully explained to them. The ver- 
 nacular effeft, however, is at once produced when they try to give 
 the sound non-vocally, 10 and O (wh, yh). For example, DJOCCCOf 
 and aoi (twh-enty and k-hue). I would adopt the rule of teach- 
 ing deaf children to give r, /, zv, and y without voice (O, CO, D, and 
 O), where they follow non-vocal consonants in the same syllable, 
 for example, in such words as pry, try, cry, try, thrice, shrine; play 
 clay, flay, slay; twin, queen, swim; pew, tune, cute, few, thews, 
 sue. 
 
 Vocal consonants, when they occur as final elements, are pro- 
 nounced by the deaf in such a manner as to offend the ordinary ear, 
 and I think it worth while, therefore, to dired your attention to a 
 simple expedient by which the effeft may be much improved. For 
 example: — take such words as love, nose, smooth, rub, good, bag, 
 etc., when they occur by themselves or at the end of a phrase. 
 
 The vocality of the last element produces an effed that is at 
 once recognized as peculiar. The effeft is much improved when 
 the pupil is taught to finish off with the non-vocal form of the con- 
 sonant softly uttered. Thus C0]33> (luvf) Ci5}16Sy> (noze) 239l&5U> 
 (smoodhth) Ci))DD> (rubp) QiQO (goodt) Dl€ia> (bagk), etc. 
 
 When two vocal consonants end the last syllable uttered, it is 
 better to give the last consonant non-vocally. Thus: ICDQ>, (edge), 
 instead of IQ3S^; (edsh, instead of edzh); OC3y>, instead of 0]365, 
 (duvs instead of duvz), etc. When such words occur in the middle 
 of a phrase the latter pronunciation is correct. 
 
 Miss Barton: How do you get pupils to give long e easily ? 
 
 Dr. Bell: I always teach first the non-vocal form O (^ in 
 hzie). If a pupil does not give this readily, I manipulate it from IS 
 (fh) or U (s) in the manner I have already described. When O has 
 been well fixed — that is, when a pupil can give it readily without 
 manipulation, I add voice. 
 
 At once we get I (ee) or, what is practically the same thing,
 
 8 1 
 
 The next question on my list is this: — 
 " Please demonstrate the teaching of tn in cotton." 
 in this word, (QJ-OCD) the sound of n alone (CC) constitutes a 
 distind; syllable. Indeed 3 (w) CD (;/) S {iig) and also CO (/,) when 
 prolonged, are in reality vowels. That is, the aperture through 
 which the voice is passed is so large as to be non-obstru(5live. We 
 fail to hear any rustling or hissing or puifing sound from the mouth 
 position. The fricative noise which is charafteristic of a consonant 
 is not heard excepting at the moment of the relinquishment of the 
 position. These sounds can be used both as consonants and vow- 
 els. If the positions are assumed only momentarily so that the 
 sound of the removal of the position is the chief effed; perceived, 
 then we recognize the sounds as consonant elements of speech. If 
 on the other hand the chief effeft perceived is due to the retention 
 of the position, and* not to its removal, we hear only a quality of 
 voice, that is, a vowel sound, and this sound may constitute a dis- 
 tinct syllable by itself. 
 
 In the English language 03 («) and CO (/) are often employed as 
 vowels. 9 {m) is more rarely used, €5 {ng) not at all. A vowelized 
 S (;«) is not usually recognized as constituting a distin6l syllable by 
 itself, but surely such words as Ct)l2^D (rhythm) and 2^f2i5B (schism) 
 are as really dis-syllabic as lOCD (eaten), QJOCC (cotton), or IDCO 
 (apple). The termination, "ful," which is so often murdered by 
 deaf children is pronounced by most people simply as 3C0 (fl). For 
 example: J3C0 (awH), Q3UIQ53C0 (dreadful), D(T)iOl3C0 (beautiful), 
 etc. Surely the word i3C0 (awfl) would be more acceptable to 
 ordinary ears than the J3l3*ICt) pronunciation commonly heard from 
 the deaf. 
 
 It is difficult to teach the sound of tn in such a word as cotton 
 without the use of symbols. The pupil associates the single char- 
 after / with a double adion of the organs (0>). Hence, he tries 
 to give this double adion to the / in cotton; that is, D>CD for OCD. 
 The vocality of the CD (;;), however, usually causes him to fail in his 
 aim, so that the puff of air (>) is vocalized (I). Thus, QJOICD. In 
 our pronunciation of the word the point of the tongue is not 
 removed from the upper gum between the positions for / and n, 
 ajOCD. The point-shut position is common to the two sounds DCD. 
 The soft palate is closed against the back of the pharynx (D) during 
 the production of the / and drops (D) during the production of the
 
 82 
 
 n (see dotted lines in Figure i6), thus allowing the voice to pass 
 through the nose. 
 
 Fig. i6. Aftion of the soft palate in foiming tii (DCC) in "cotton." 
 
 First Position. 0+0=: D 
 
 Second Position. C4-D=C5 
 
 D + (Dq=OC5 
 
 Repeat the sound of tn a number of times without voice 
 (DDODDCJ, etc.), and you will feel that the whole action consists in 
 the alternate elevation and depression of the soft palate [DCDC etc]. 
 The point of the tongue is not moved at all, but is shut continuously 
 against the upper gum. DC50CJOC5=D+(acacaC, etc.). 
 
 There is only one point-shut position. Now, such an aftion 
 presents no difficulty to a child who has been taught to elevate and 
 depress his soft palate at will, but does present enormous difficulties 
 to one who has not, for the whole adion is invisible. All you can 
 do in such a case is to use symbols, and show that the shut-position 
 (D) of / (D>) is alone assumed, followed by CC (//) without moving 
 the tongue. The combinations DCC (///), CCD> (///), ©CC (,///), ©=-0 
 {nd), DCi30> (/;;/), and ©CDCD> {dud), occur in English words, some 
 of them quite frequently; and I think therefore that deaf children 
 should be taught to control the movement of the soft palate so as to 
 be able to produce these combinations at will. For example : The 
 words: IDCC (eaten), ICiiO> (ant), SIQCC (madden), XCCCD> (and), 
 DIDCCD=> (patent), and StCDCD0> (maddened), involve these actions, 
 actions. The word " abandoned," as pronounced by some people 
 IDICCG5aj0> adually involves four point-shut sounds successively 
 uttered, or rather one point-shut position, and four successive posi- 
 tions of the soft palate. CD0CD® = 0-t-(CDCa>). 
 
 In difficult cases I would recommend the following plan : Give
 
 ^3 
 
 your pupil a hand mirror and teach him to elevate and depress his 
 soft palate (QCDC, etc.), in the manner described in my second 
 ledure. 
 
 When he can do this at will without looking in the mirror, ask 
 him to repeat the exercise with his lips shut all the time. This 
 results in D+(a£:acaC, etc.)=DDDDDD, etc. 
 
 Then repeat the exercise with the point of the tongue shut 
 against the upper gum : 
 
 D+(acaEac, e{c.,)=oaD^oa, etc. 
 
 Then with the back of the tongue shut against the soft palate: 
 
 a+(DcacDC)=aGaGaa, etc. 
 
 Then let him repeat these exercises with the voice sounded 
 intermittently through the nose: 
 
 DSD9, etc. (pm, pm, etc.) 
 
 DCCDCC, etc. (tn, tn, etc.) 
 
 QGOG, etc. (kng, kng, etc.) 
 Then with the voice continuously sounded: 
 
 D9D9, etc. (bm, bm, etc.) 
 
 0CCQCD, etc. (dn, dn, etc.) 
 
 eeee, etc. (gng, gng, etc.) 
 
 If a pupil is taught to control the movement of his soft palate 
 at will, such combinations as that of in in cotton will present no 
 difficulty. 
 
 [Dr. Bell here illustrated his method with Mr.Kiesel].
 
 CONSONANTS. 
 
 I propose in my lefture to-day to deal with the mechanism of 
 speech. In my demonstrations I shall make use of Monroe's school- 
 room charts, "Sounds Of The English Language," which contain 
 good diagrams of the positions of the vocal organs in uttering 
 English sounds. As you are professionally familiar with the subject 
 it will be unnecessary for me to describe the correcft positions unless 
 in answer to questions. I shall therefore consider the mechanism 
 of common defeats. We shall consider first the defeds of shut 
 consonants. 
 
 DEFECTS OF SHUT CONSONANTS. 
 
 * 
 
 The labial letters, p, b, m, are not liable to errors of position. 
 In forming /, d, ;/, slight changes of position do not offend the 
 ordinary ear and may therefore be passed lightly by. The point of 
 the tongue should be placed against the upper gum. but it may be 
 shut against the teeth, even to the th position without producing a 
 defeft sufficient to attraft the attention of any one but an articulation 
 teacher. In forming h, g, ng, also, considerable latitude may be 
 allowed so long as the position is too far forward. When it is too far 
 back the defeft at once attracts attention, and should be corrected. 
 If you try to form a k, with your mouth opened as widely as pos- 
 sible, you will find it very difficult to raise the back of the tongue 
 into contact with the soft palate. It is much more easv to produce 
 the shutting adion lower down by the approximation of the base of 
 the tongue to the back of the pharynx. This produces the defective 
 sound of k to which I have alluded (Q<), a sound quite commonly 
 given by deaf children. I am inclined to think that the defeat is due 
 to the mode of teaching. The teacher is very apt to open her mouth 
 as widely as possible to show her pupils the position of the tongue. 
 They imitate the action, and this naturally leads them to give too 
 
 84
 
 85 
 
 low a position. It is very difficult to corred a position that is too 
 far back. I think the best way is to teach the sound anew. Take 
 a position which is too far forward, for example, / (D), then manip- 
 ulate the tongue. The same remarks, of course, apply to g (Q). 
 
 The shut consonants, though generally pronounced well by deaf 
 children, are liable to a defeft of a very extraordinary nature. In 
 nearly every school for the deaf some pupils may be found who 
 give clicks in place of these consonants. For example: p will be 
 pronounced like the sound of a kiss, and / like the clicking sound 
 we make as a sign of impatience, or like the cluck with which we 
 hurry up a horse. 1 may not be able to tell you exadly what to do, 
 but of one thing you may be sure, — the first step in the correcffion 
 of a defe(5t is to understand the mechanism of the defective sound. 
 "Knowledge is power," and when we know the nature of a defe(5l, 
 ingenuity will find a remedy. 
 
 The first step then is to study the mechanism of the defective 
 sound. How are you to investigate it } Imitate the defed ive sound 
 yourself and then study your own vocal organs. 
 
 Your pupil, we shall suppose, gives a kiss instead of the sound 
 of p. Let us study the mechanism together. 
 
 When the lips are opened you observe that air goes into the 
 mouth instead of coming out. Let us examine into the cause. 
 
 But first, let us express by means of symbols and diagrams the 
 condition of our knowledge at each stage of the investigation, so 
 'hat we may realize as clearly as possible what we are about. 
 
 Fig. II. 
 First Position. Second Position. 
 
 D< 
 
 The first diagram in Fig. ii. illustrates the closure of the lips 
 (D). The second shows, by means of an arrow-head, the direction 
 of the air when the lips are opened (<). 
 
 Can it be that the pupil makes an effort of inspiration while he 
 is trying to say p .^ How can we satisfy ourselves on this point }
 
 86 
 
 Shut your lips then open them with an effort of inspiration. Thus, 
 (D<). At once you notice that the effeft is very different from the 
 sound of a kiss. In forming the kiss, then, the air does not enter 
 the lungs (<), but only goes into the mouth as far as some point [<] 
 yet to be determined, [The symbol D< now becomes D<], 
 
 Repeat the kissing sound many times — D< D< D<, etc., — while 
 you observe what you do with your lungs. You will find that you 
 can go on kissing for any length of time without stopping to take 
 breath. You can breathe freely through the nose. 
 
 What conclusion can we draw from this .^ [i] The soft palate 
 is depressed [S]. 
 
 [2] You can breathe freely through the nose thus showing that 
 no constaiftion exists between the soft palate and the lungs. 
 
 Let us express this knowledge upon the diagrams we have 
 made. Figure 11 now becomes Figure 12, and the symbol D< 
 becomes S -\- [D<]. 
 
 £- 
 
 Fig. 12. 
 First Position. Second Position. 
 
 [) + D] [) + <] = )[D<] 
 
 Is anything further to be discovered ? Think for one moment. 
 If no other constriction exists than is shown on the diagram, then 
 there must be an open passage-way into the mouth from the lungs, 
 and air should escape through the mouth as well as throug the nose. 
 Does it do so ? We know it does not, for when the lips are opened 
 air enters the month in just the opposite diredion [< and not t>]. 
 The passage-way, then, must be closed somewhere between the 
 soft palate and the lips. What organs are there in the mouth by
 
 87 
 
 which the closure could be effeded ? We are limited in our choice 
 to the point of the tongue, the "top" or "front" of the tongue, 
 the back of the tongue, or to intermediate parts. 
 
 If you repeat the sound of a kiss, 1 think you will feel that the 
 concealed shut position must be pretty far back in the mouth. Cer- 
 tainly the point of the tongue is not involved, and we are limited 
 therefore, to the top or back of the tongue, with the probabilities in 
 favor of the back. How can you decide the matter ? Make a 
 hypothesis, and then experiment upon your mouth to test the truth 
 of your assumption. For example: Assume that the back of the 
 tongue is shut against the soft palate (Q). Fill in this position upon 
 the diagrams. Figure 12 now becomes Figure 13, and the expres- 
 sion S + (D<) becomes G + (D<) 
 
 Fig. n. 
 First Position. Second Position. 
 
 (j + a + D) () + a + <) = G + (D<) 
 
 Now study the diagrams and the symbols and try to establish 
 some relation between the hypothetical position (G) and some sound 
 of known formation. Then experiment upon the mouth to see if 
 that relation holds good. 
 
 Now we know that G + 1 = &(iig). 
 
 If then your hypothetical position (G) is corred, you should 
 get G (iig), by adding voice to a kiss. Test the matter. Sound the 
 voice continuously while vou repeat the sound of a kiss: — 
 i+a+(D< D<-D<- etc.) = G+ (D< D< D< etc). 
 
 You at once recognize the familiar effect 6 (//«), continuously 
 sounded — like the drone of a bagpipe — accompanied by the equally 
 familiar sound of kissing. This is proof that your assumption is 
 correct.
 
 88 
 
 A kiss:^G + (D<). 
 The defed, then, consists in the assumption of the ng position 
 without voice (G) while the pupil is trying to say D> (/>). 
 
 But why does the air go into the mouth when the lips are 
 opened ? The f:i6l indicates that a partial vacuum exists there. This 
 means that the cavity of the mouth had been enlarged while the 
 shut positions were assumed, thus causing rarefaftion of the con- 
 tained air. The tongue, therefore, must have been moved before 
 
 Fig. 14. 
 
 i + Q+I+D 
 
 5+a+L+D 
 G+[D + (rT)T] 
 
 or G+ (D<) 
 
 i+a+i+D 
 1+ l+o 
 
 D+ra+dDL] 
 
 orD+(a<) 
 
 the lips were opened. Any movement of the tongue that will en- 
 large the cavity, will produce a partial vacuum in the mouth, and 
 thus lead to the produdion of the sound. The symbol < (air going 
 into a cavity) expresses the effect independently of the exa6l
 
 89 
 
 positions assumed. Hence: G-f (D<), is a general expression and 
 covers any change of position inside the mouth that will produce a 
 partial vacuum there. 
 
 As a clear understanding of the cause of the click effe6l in this 
 case will throw light upon the nature of clicks in general, it may 
 be well to show some specific movement of the tongue that will 
 produce the effect. For example: Suppose that the front of the 
 tongue is elevated in the position for e in eel thus f (high front) 
 when the shut positions are assumed. (See Fig. 14). If then the 
 front of the tongue is lowered into the position for e in pet (as 
 shown by a dotted line in Fig. 14) thus \ (low front), without 
 changing the other positions, the cavity of the mouth is enlarged. 
 As a partial vacuum then exists, air will rush in if an opening is 
 made anywhere. For example: — 
 
 1. If you keep the back of the tongue closed against the soft 
 palate and open the lips, air will rush in between the lips ; G + 
 (D<). 
 
 2. If you keep the lips closed and open the passage-way be- 
 between the back of the tongue and the soft palate, air will rush in 
 the cavity from behind D + (a<). 
 
 Suppose again that instead of starting with the tongue elevated 
 you commence with it depressed I (as shown by dotted line Fig. 14,) 
 and then elevate the tongue into the position for e in ee\ f, the cavity 
 between the two shut positions is reduced in size and the contained 
 air compressed. Then : — 
 
 3. If you keep the back of the tongue closed against the soft 
 palate and open the lips, air will rush out of the cavity through the 
 labial aperture G + (D>) ; or 
 
 4. If you keep the lips closed and open the passage-way between 
 the back of the tongue and the soft palate, air will rush out of the 
 cavity into the pharynx SO + (Q>). 
 
 Numbers i and 2 are suction clicks. Numbers } and 4 are ex- 
 pulsion clicks. Numbers i and 3 are both given by deaf children 
 instead of p (D>), 
 
 In order to have a click sound it is necessary that you should 
 have a cavity in which the air is of different density from that out- 
 side. There must, therefore, be two constricftions of the passage- 
 way which we may call x, y, with a cavity between them. If the 
 air in that cavity is of less density than the air outside, the opening 
 of the passage-way at either end will result in a sudden in-rush of 
 air, forming a suction click.
 
 90 
 
 It the air in the cavity is of greater density, an expulsion click 
 will be produced. Double positions are, therefore, capable of pro- 
 ducing two sudion clicks, and two expulsion clicks which may be 
 thus symbolized : — 
 
 Su^ion Clicks. 
 
 2. y-\-{x<)\ 
 
 , / [ [ Expulsion Clicks. 
 4. y +{x>)\ 
 
 In my use of the syrribols — 
 
 > means air going out from the lungs ; 
 
 < means air going into the lungs; 
 
 > means air going out from a cavity ; 
 
 < means air going into a cavity. 
 
 I have alluded to two labial clicks made by deaf children instead 
 of P. There is still a third which is quite common X + (D>)- 
 
 Fig. 15. 
 
 X + D +(D<) 
 X +(□<)+ D 
 (X<)+ D + D 
 
 Sii^ion Clicks. 
 
 4. X + a +(D->) 
 
 5. X +(a>)+ D 
 
 (X>)+ D + D 
 
 The inner shut position is produced by the c.osure of the glottis 
 ( X ) (See Fig. 15), and the cavity between the two shut positions is 
 
 \ Expulsion Clicks.
 
 91 
 
 larger than in the case shown in Fig. 14. The sound therefore has 
 a lower pitch than any of the clicks resulting from the positions 
 shown in Fig. 14. 
 
 The pupil also is unable to breathe through the nose while pro- 
 ducing the sound. 
 
 The closure of the soft palate against the back 01 the pharynx 
 (D) is of course assumed though not expressed in the symbol 
 X + {0>). There are really in this case three shut positions, and 
 theory therefore indicates the possibility of producing three expulsion 
 and three suction clicks from the positions shown. Below Fig. 15 
 I give the symbols for the six clicks alluded to for the benefit of 
 those who desire to study them. 1 need not describe them further 
 in detail, as we meet with only one of them — the fourth — in our 
 work. 
 
 The fourth click X + (D->) is sometimes given by deaf children 
 in place of D> (p). 
 
 We meet with three click forms of D> (/), G+(D<) ; G+ (D>) ; 
 
 and X + (0>) 
 
 The inner shut positions in each case is the same as that already 
 noted for p. 
 
 K is subjed to only one click X+ (Q>)- 
 
 zM is often given as a vocalized kiss G+ (D<) — such a word as 
 "mamma " for example, being pronounced as two kisses with the 
 voice passing continuously through the nose. 6+ (D<D<). 
 
 U^ is liable to a click of similar formation e+(D<), but ng is 
 never clicked. 
 
 'B, d and g are subject to a very curious form of sudion click 
 resulting from the attempt to teach these sounds elementarily instead 
 of in combination with vowels. A teacher, for example, will pro- 
 nounce the sound of Z' without opening the lips (3). and place her 
 pupil's hand upon her throat so that he may feel the vibration pro- 
 duced by the vocal cords. The voice cannot be prolonged because 
 the closure of the lips prevents the escape of air. The pupil there- 
 fore feels only a momentary impulse of voice; and in attemoting to 
 imitate this effed, he makes a sudden muscular effort. 
 
 if you place your hand upon the pupil's throat while he makes 
 the sound, you will observe a strong muscular effort causing an ex- 
 pansion of the whole throat. The cavity between the vocal cords 
 and the lips is thus enlarged, creating a partial vacuum within. 
 The slit-like aperture of the glottis is not large enough to permit air 
 to freely enter the cavity from the lungs so as to restore the pressure.
 
 92 
 
 upon opening the lips, therefore, air enters the cavity from 
 outside, thus forming a sudion click, 1 + (D<). 
 
 D and g are liable to a similar defedl. For example : d = 
 -f- (D<), _«^ = I + (a<). These clicks, unfortunately, are of very 
 common occurrence. 
 
 TABLE OF CLICKS GIVEN BY DEAF CHILDREN. 
 
 For/). a+(D<) G+(D>) I + (D>) 
 For /. G + (D<) G + (D>) X + (0>) 
 
 For k. X + (a>) 
 
 For /;/. e + (D<) 
 
 For «. G+(D<) 
 
 For nff. 
 
 For b. I + (D<) 
 
 For d. ■ I + (D<) 
 
 For^. l + (a<) 
 
 If you have followed me so far, you will recognize the fatt 
 that clicks result from double positions of the vocal organs. When, 
 therefore, you hear a click, you know that there is a constriction 
 somewhere which is concealed from dired observation. In study- 
 ing the defeat, therefore, your first object should be to discover where 
 that concealed position is. The clicks most commonly given by deaf 
 children result either from the closure of the back of the tongue 
 against the soft palate (O), or from the closure of the glottis (X). 
 
 If the pupil can breathe through the nose, you may assume at 
 once that the concealed position is G {ng, without voice). If he 
 does not breathe through the nose, the location is more uncertain 
 (either Q or X)- The pitch of the sound may help you here, for a 
 click which is due to the closure of the glottis is lower in pitch 
 than one due to the closure of the back of the tongue against the 
 soft palate, because the cavity is larger. Perhaps the most certain 
 plan of ascertaining the location is to cause the pupil to repeat the 
 click for as long a time as possible without stopping. 
 
 If he is not breathing through the nose, nature will sooner or 
 later force him to relinquish the concealed position in order to take 
 breath. Watch for that moment. At the moment of relinquishment 
 a sound will be heard which will enable you to determine by ear the 
 location of the concealed position. For example: If the back of the 
 tongue is involved you will hear the sound of k (Q< or a>).
 
 9} 
 
 pronounced either with the air going in or out of the lungs. If the 
 glottis is closed you will hear ( X> or X<) the effeft of throat shut 
 followed by a puff of air. If you are uncertain whether or not the 
 pupil can breathe through the nose, hold the nostrils closed 
 with your fingers until the pupil's breath gives out. If the 
 concealed position was G {ng without voice) you should hear the 
 sound of A (a> or a<) for Cf> — 'i—(J. 
 
 The first step in the correction of a defeft is a knowledge of the 
 cause. With this knowledge the teacher can not only devise means 
 (i) of correcting the defed, but (2) of utilizing it in the production 
 of other sounds. Let us take as a typical case of a click defeat, the 
 sound of a kiss given instead of^. How can we correal it and how 
 can we utilize it ? 
 
 (1) Correction of the defeSt. In this case the cause consists in 
 the assumption of a shut position (G) which prevents the breath 
 from reaching the lips. If then, we can devise any method of caus- 
 ing air from the lungs to press against the lips the assumption of the 
 concealed position becomes impossible. For example: You can 
 make your pupil blow feathers or pieces of paper away from his lips, 
 or inflate his cheeks while trying to pronounce/). 
 
 These are simple expedients that are usuallv successful; but 
 they may fail because it is perfe6tly possible to produce inflation of 
 the cheeks and expulsion of air from the mouth, and yet have the 
 back of the tongue shut against the soft palate when the lips are 
 opened. 
 
 The characteristic puff heard during the production of a click is 
 necessarily of very short duration. Then get your pupil to make a 
 continuous effort of expiration. For example: Let him shut his lips 
 and blow continuously through a very fine orifice between them 
 (DDi) as though he were blowing to cool something. 
 
 In producing a prolonged emission of this kind the air can only 
 come from the lungs, and the assumption of any interior shut posi- 
 tion is therefore impossible. It is true that the inflation of the cheeks 
 suggested above, and the too small aperture between the lips just 
 alluded to, themselves constitute defeCts; but they are easily cor- 
 redled, because the aCtions are visible. A skillful teacher. will not 
 hesitate to substitute a defed that is easy of corredion for one that 
 is more difficult.
 
 94 
 
 (2) UtiIi{atiOfi of the defeSlive sound. The moment you realize 
 that the concealed shut position is G you will recognize the possi- 
 bility of producing ng (6) from a kiss. For example: Get the 
 pupil to add voice to the kiss. 
 
 [a+(D<)]+l = S+(D<) 
 
 and at once you obtain ng combined with a labial aftion. By ma- 
 nipulation of the pupil's mouth you can prevent the lips from closing, 
 
 [65+ (D<)] - (D<) =e 
 
 and you then get ng alone. If the pupil has not already acquired 
 the sound of ng, you can thus utilize the kiss as a means of teaching 
 it to him ; and if he already has the sound and knows its symbol 
 (6) then the presentation of the symbol for the kiss will convey to 
 his mind an idea of the mechanism of the click. In corre(5ling de- 
 feds, it is surely advisable, if possible, that the pupil as well as the 
 teacher should know the cause and understand the mechanism of 
 the defe(5live sound. 
 
 If then your little pupil should happen to give the sound of a 
 kiss instead of the letter p, don't frown at him and say, "No, no, 
 that is not right." Give him the symbol 0+ (D<) and encourage 
 him by a sign of approval. 
 
 He has tried his best and it is not his fault that he failed to give 
 the sound you wanted. All sounds are but positions to him and he 
 was right in his attempt — not wrong — for he imitated corredly the 
 position which alone he could see — the position of the lips. Why 
 then should we express disapproval } He had done nothing worthy 
 of censure. The "No-no method " does not help him to corre6t 
 the defeft — and it does throw cold water upon honest attempts to 
 please. Give him a sign for his sound, and reward his effort by 
 approbation. If you do not know how to write the sound properly 
 give him x as a provisional symbol — or invent a chara(5ter to repre- 
 sent it. Say, "That is what you did, now do it again." The 
 "No-no method" gives him the idea that it is wrong to make a 
 noise of that kind. On the contrary, encourage him to repeat the 
 sound so that you may study it and find out how best to utilize it 
 in his instruiftion. If the sound is unfamiliar to your ear and you do 
 not know how it is formed, that itself is a reason why you should 
 hold on to it and not throw it away. When you have analyzed its
 
 95 
 
 composition you may find it to contain gold where you only asked 
 for lead. Let your pupil repeat the sound until you can imitate it 
 yourself. Then study your own mouth. In the meantime do not 
 let him forget the sound. Fix it by reference to the letter x, or some 
 other mark, and when you have satisfied yourself how it should be 
 expressed, substitute for x the proper symbol. 
 
 The meaning of the symbol need not be explained to a young 
 child. It may be treated as an arbitrary sign. The expression G+ 
 (D<) need only mean to him, " That's what you did, now do it 
 again." The deaf child soon comes to understand the application 
 of the symbols even though he may not understand their full mean- 
 ing. For example: When he knows that G-|-(D<) represents the 
 sound he makes, then if you change the symbol to G+(D<) he will 
 at once attempt to vocalize the kiss. He will do this even though 
 he may be unable to analyze or understand the full significance of 
 the expression. 
 
 The symbols of Visible Speech are invaluable as a means of cor- 
 reding and utilizing defedive sounds. Indeed, I think their chief 
 value lies in their ability to express the mechanism of the sounds the 
 children make, so as to show in a graphical manner their relation to 
 the English sounds we wish them to give. 
 
 The methods suggested above are applicable to the correction 
 of all the clicks of t and k. (See Table of Clicks). 
 
 DEFECTIVE COMBINATIONS OF P, T, K. 
 
 Pupils who are taught by means of Visible Speech have many ad- 
 vantages over those taught only by means of Roman letters and 
 diacritical marks. Defeds of combination, which are inevitable upon 
 the latter plan, and which require the expenditure of much time and 
 labor on the part of the teacher in order to correct them, need not 
 arise at all when symbols are employed, and if they do arise are 
 easily correded. 
 
 Let me illustrate by a common case. 
 
 The position for p (D) by itself yields no sound, because the 
 lips are closed. It is usual, therefore, to teach it in combination 
 with an open position. Thus, D>. The lips are first shut and 
 then opened to allow of the escape of a puff of air. 
 
 Here we have two successive positions represented by only 
 one character p. This leads at once to a defect when p is com- 
 bined with other letters, for the child naturally attempts to give
 
 96 
 
 both positions (D>) wherever/) occurs. Thus, ps becomes 0>U in- 
 stead of DU (a puff of air appears between the/) and s). So also 
 with / and h. 
 
 Ts becomes D>^ instead of 0^. 
 
 Ks becomes 0>13 instead of QO. 
 
 Tsh{ch) becomes D>Q instead of DQ, etc. 
 
 These defeds are inevitable upon the Roman letter plan. 
 Without the use of symbols it is difficult to explain the nature of 
 the defe(?t to a deaf child who knows no language. The teacher 
 usually imitates the defedive sound and exclaims, " Don't say D>Q, 
 but DQ (tsh), " trusting to the quickness of the pupil's eye to discern 
 the difference in her mouth. Your chief reliance is upon imitation, 
 and if that fails you, where are you .? Now, these defects need not 
 arise at all when the sounds are taught by symbols; and when 
 they do occur, they are easily corrected, because we can express 
 the incorreft as well as the correct effect so as to exhibit the differ- 
 ence. 
 
 In teaching these sounds by means of visible speech we com- 
 mence in the same way as that just described — by teaching/), / and 
 k, as D>, D>, and a> but we employ two characters to express the 
 two positions instead of one. Then when the sound of s (y)is ac- 
 quired we combine as follows: 
 
 1. "Say D>" "That is right." 
 
 2. "Now sav U" "Right." 
 
 3. "Now say D>U " " Right again." 
 
 (Observe the difference in the attitude of the teacher towards 
 her pupil. Here is the very defective combination of ts alluded to 
 above. But the teacher of visible speech, having expressed the po- 
 sitions actually assumed by her pupil, can truthfully say, "That's 
 right," with an approving nod— where the Roman letter teacher 
 could only say, "No, no, that's wrong.") 
 
 4. " Now try first Q>U and then DO." 
 
 Here the pupil's attention is directed to the difference between 
 the two effefts, and his aim is to give the last combination without 
 the puff af air (>). Whatever he does, therefore, his aim is right — 
 which is not the case on the Roman letter plan. And whatever he 
 does, the teacher can give him a symbol for his sound and say, 
 "That's what you did, now do it again." 
 
 In a little time quite a number of variations upon the sound of 
 ts may be obtamed. Anxiety to avoid the puff of air often leads
 
 97 
 
 him to put it in — now in one place, now in another; for example, 
 the pupil may say, D>y or >DU or >D>y, etc. 
 
 It is not the teacher's objed to have him forget the incorreft 
 sounds but to remember them and contrast them one with the 
 other. The greater the number of slight variations that can be pro- 
 nounced at will by the pupil the more power does he obtain over 
 his vocal organs. A good marksman should be able to hit one 
 mark just as well as* another. 
 
 The sound of ch (tsh) presents exceptional difficulties to a dea. 
 child. Even when the / is properly combined with sh without any 
 puff of air between the two (thus OQ) your ear usually tells you 
 that there is something wrong when the combination is uttered in a 
 word. 
 
 1 think the fault lies in the undue prolongation of the sh position 
 (DQ»). If you observe your own utterance of such words as chair, 
 cheese, church, such, much, touch, watch, etc., you will notice that 
 the tongue does not remain for any length of time in the sh position. 
 Ihtsh indeed constitutes a non-vocal glide, a mere transitional effect, 
 between / and the succeeding element. The unnatural effed pro- 
 duced by prolongation is most marked when sh occurs finally, as in 
 much, touch, etc. (Q]OQ* D]DQi, etc.). I have rarely failed to ob- 
 tain the vernacular effeft from a deaf child by expressing the sound 
 oich final asDQ> instead of DQ. Indeed, as a general rule a non- 
 vocal consonant occurring as a final element is most naturally given 
 by a deaf child when the symbol for the sound is followed by >. 
 For example: 
 
 (cuff) a]3> (us) ]y> (both) B}-\U^ (wish) 3fQ> 
 
 (cup) a]D> (nut) CD]D> (sick) Ora> (maps) 91D^> 
 
 (cuffs) a]3y> (nuts) CC]DU> (deaths) W\UU> (books) 0iaO> 
 (watch) 3JDQ> 
 
 Of course, when these words occur in the middle of a phrase, 
 the puff of air must be omitted, for the phrase is pronounced as one 
 word and the consonant is then no longer final. 
 
 CORRECTION OF THE DEFECTS OF 'B, D, G, M AND N. 
 
 it is difficult to pronounce the sound of /' (B), without opening 
 the lips and when a deaf child attempts to do this a defedive sound 
 arises which, when combined in a word with other sounds, 
 produces the click form o( h [l+(D<)] alluded to in the table of 
 clicks. T) and g are subject to a similar defed. 
 
 1 would recommend combining these consonants with vowels 
 from the very first. 1 commence with an indefinite vowel (I), which
 
 98 
 
 may be er her, // in up, or any indefinite vowel sound that the 
 child can make. For example: B, d, g may be taught as 31 01 Ql 
 (ber, der, ger, or bu, du, gu, etc.) There is really no difficulty in 
 teaching b in combination, for it can be manipulated while the child 
 produces the vowel sound. Let the child prolong an indefinite 
 vowel sound with his lips pretty close together. Now place your 
 thumb and finger under his lower lip and move the lip rapidly up 
 and down so as to close and open the labial aperture. This results 
 in DI3IDI etc. (ber, ber, ber, etc.) Care should be taken to make 
 the movement an opening not a closing adion. The closure should 
 be only momentary. The under lip should instantly rebound from 
 the upper lip as a hammer rebounds from an anvil. Now teach the 
 child himself to move his lip up and down with his finger. His 
 attempt should be to pronounce the vowel (I) continuously and make 
 no muscular effort with the lips. After he can do this well let him 
 try to move his lip rapidly up and down in the same way without 
 the assistance of his hand. There should be no muscular tension, 
 but on the contrary the lips should feel soft and loose. 
 
 You cannot manipulate the point of the tongue in this way, but 
 when the pupil can pronounce DIDIBI, etc., analogy leads him to give 
 OlOl0letc., (der der der, etc.) and QIQIQI etc., {gu gu gil etc.) 
 
 The clicks of ;// and ;/ are more difficult of correflion. 
 After the child can give DIBIDI etc., or 010101 etc.. it is well to try 
 whether analogy will not lead him to give QI9I9I or CCICCiCCl etc. 
 This often succeeds, but in difficult cases the back of the tongue 
 remains closed against the soft palate, thus converting the sound 
 into 6+(D<D<D<) or e+(D<D<D< etc.) a continuous sound of w^ 
 accompanied by a succession of clicks. The defeft is due to the 
 retention of the back-shut position (Q). I think, therefore, the best 
 way to deal with a difficult case is to aim at control of the back of the 
 tongue, so that the pupil shall acquire the power of elevating and 
 depressing it at will. 1 would suggest taking the concealed posi- 
 tion 6 by itself and combining this with an indefinite vowel sound 
 thus: SieiSi etc. 
 
 When he can do this well contrast it with SiSiBl etc., thus 
 Gieiei ai9iSI etc., eiSieiSi etc. The analogy of the symbols 
 will probably lead him to give the correal effeft. 
 
 Wh and w (D and J3) are very defed;ively given by deaf child- 
 ren, but I have already spoken of the mode of corredion in a former 
 ledure.
 
 99 
 
 F, T, th, fh. (3 3 LS 65) present no difficulties of importance. 
 S, 5/;, yh, {U Q O) and their vocal forms Z, i^i, y (2i5 fi and (^), 
 are liable chiefly to faults of position. That is, the tongue may be 
 a little too far forward (>); or too far back (Ol or too high up. 
 There may be too much compression of the passage-way ( a ) ; or 
 too little compression (v). 
 
 My plan of correction is — to write what the pupil does, using 
 these modifiers according to the character of the defeft to be sym- 
 bolized. For example: Q< Q> Qv Qa etc. 
 
 I then get the pupil to vary the position slightly and contrast the 
 new position with the old making him pronounce both sounds alter- 
 nately so as to observe their difference. 
 
 In difficult cases it is well to manipulate the non-vocal forms s, 
 sh, and y/i, {U,Q, and O,) from t/i (15) in the manner I have 
 already described in answer to a question. When these are well 
 tlxed the vocal forms follow as a matter of course by the addition 
 of voice.
 
 VOWELS, GLIDES, AND COMBINATIONS. 
 
 Dr. a. Graham Bell : 1 have advocated the very general use 
 of an indefinite, in place of a definite vowel sound in unaccented 
 syllables. You must not, however, suppose from this that I under- 
 value vowels, or deem accuracy of vowel quality of no practical 
 importance in our work. Far from it. I only mean to insist that 
 vowels are of secondary importance to consonants. 
 
 Consonants give intelligibility to.speech, but vowels give beauty 
 of utterance. Consonants constitute the back-bone of spoken 
 language— vowels the flesh and blood. You cannot do without 
 them. 
 
 We want our pupils to acquire, not merely an intelligible articu- 
 lation, but also, if possible, a natural and pleasant quality of speech. 
 We must, therefore, attend to the vowels. It is neither necessary, 
 however, nor advisable, that every vowel in a phrase should be given 
 its full value. Unaccented syllables should be toned down like the 
 shaded portions of a pidure thus bringing out by contrast, the 
 accented parts of words. 
 
 Beauty of utterance depends as much upon shading as upon 
 form — as much upon the due subordination of the unaccented syl- 
 lables as upon accuracy of vowel quality. 
 
 It is a very difficult thing, even for hearing persons, to give un- 
 accented vowels their proper sounds without bringing them out too 
 prominently, so as to produce that pedantic style of pronunciation 
 which is often mistaken for elocution. 
 
 The really good speaker gives the proper value to unaccented 
 vowels without italicizing them to the ear. The mass of the deaf, 
 however, are no more able to do this, than the mass of the hearing. 
 Indeed, the attempt results in a much more unnatural effect than 
 
 iOO
 
 101 
 
 the utterance of the pedantic speaker — because the vowel quality it- 
 self is usually defedive. Under such circumstances indefiniteness is 
 of importance. It produces, not a worse, but a better efifecl;. You 
 must not suppose, when I advocate a careless utterance of un- 
 accented syllables, that I am_ urging you to teach worse speech than 
 your pupils now possess; on the contrary, I believe that the result 
 will be recognized as a great improvement. You will tone down 
 sounds that are usually defedive so that they will not come out so 
 prominently to the ear; and accent, which is now conspicuous 
 chiefly by its absence, will be produced by the subordination of the 
 unimportant parts of words. 
 
 Give as definite vowels as possible in the accented syllables, but 
 don't be too precise about the others. 
 
 Vowels are the most difficult elements we are called upon to 
 teach. Why is this so ? The discovery of the cause may perhaps 
 enable us to devise a remedy. Let us examine into the matter. 
 
 How do vowel positions differ from the positions that yield con- 
 sonant sounds ? They result from larger apertures. Can this have 
 anything to do with the difficulty of the acquirement ? It seems so, 
 for wide-aperture vowels are more difficult to obtain in perfection 
 than the others. 
 
 Get a pupil to prolong a small-aperture vowel. The sound, 
 even when defedive, has a definite quality of its own. A lower po- 
 sition of the tongue, however, yields an effect of indefinite kind. 
 The oral aperture is usually too large and the sound, when prolonged, 
 is unstable and variable in quality, showing that the pupil finds diffi- 
 culty in retaining the position unchanged. 
 
 Sometimes the attempt results in a visible trembling of the 
 tongue. 
 
 In forming consonants and small-aperture vowels, the tongue 
 makes actual contact with the upper part of the mouth at one or 
 more points; but in lower positions it is hung in the air, so to speak, 
 without anything against which to press. Extend your arm and 
 you can easily retain it in a fixed position if you press your hand 
 against the under surface of a table or shelf, but extend it in the air 
 and I fancy you will find more difficulty in keeping it still. Your 
 hand — if you do not watch it — is apt to waver like the unsteady 
 tongue of the deaf child, and a constant tendency exists to a lower 
 position. 
 
 Your ear aids you in the retention of a /owel position, because 
 any change affects the qi'jlily of the sound.
 
 102 
 
 Try to keep your tongue still, without making any noise, and 
 you will appreciate the ditficulty experienced by your pupil. He 
 lacks a guide. 
 
 Give him a mirror and at once he becomes conscious of the 
 movement of his tongue. 
 
 The sound is your guide, and if you merely think of the sound, 
 that helps you to retain the position. His guide must be sight, and 
 by seeing he will learn control. The thought of the image he has 
 seen in the mirror will help him, as the thought of the sound helps 
 you. 
 
 I cannot overestimate the value and importance of a mirror in 
 articulation work. It is not enough that a pupil should watch his 
 teacher's mouth, he must see his own. Accuracy and defmiteness 
 of sound depend upon the ability to retain a position unchanged. 
 Indeed, as I said in my second lecture, control over the vocal organs 
 is gained not so much by moving them as by keeping them still. 
 Clearness of pronunciation depends upon the ability to enunciate 
 every element in a word with clearness and deliberation. A poor 
 speaker finds difficulty in uttering a word slowly or separating it 
 into its component parts. 
 
 Give your pupil a mirror and let him learn to keep his tongue 
 still. It doesn't much matter what he does, so long as he assumes 
 different positions of the tongue, retaining each for some time with- 
 out motion. 
 
 Little children delight to puzzle one another by assuming 
 unusual positions of the tongue which the others cannot imitate. 
 They should be encouraged in this, for all exercises of that kind are 
 of value as a preparation for speech. By such exercises they 
 unconsciously gain control over their vocal organs and become 
 better able to imitate positions of the mouth. They feel a certain 
 muscular exertion and see the effect in the mirror, and this constant 
 association of seeing and feeling ultimately enables them to realize 
 by muscular sensation alone exactly what the tongue is doing. 
 
 I consider a looking glass as a necessity in the schoolroom. 
 You can no more expeft to teach a deaf child good speech without 
 a mirror than you can hope to teach a hearing child to paint well 
 without letting him see the result of his efforts. 
 
 I would especially recommend as an exercise before a mirror 
 alternately narrowing and broadening the tongue. The attempt to 
 narrow the tongue causes it to become stiff and hard to the touch, 
 with a rounded surface. When it is broadened the surface becomes
 
 103 
 
 flat and soft. The most common fault, I think, among deaf children 
 is an exaggerated muscular adion leading to a stiffening and narrow- 
 ing of the tongue. For example: Many children in trying to form 
 X (ee) make so much muscular effort that the tongue feels hard. 
 The tongue is narrowed and fits up into the arch of the palate, 
 making contacl with the top of the hard palate, and though a centre 
 aperture exists over the front of the tongue the effeft of the vowel I 
 is not produced. 1 have found C (the German cii) to be a very 
 important position. It forms, indeed, the key note to the vowels. 
 As a general rule if a child can pronounce C you can teach him to 
 glide the tongue forward (C>) and backward (C<), retaining the small 
 centre aperture, and by getting him to do this as far back and as far 
 forward as possible, the latter position generally gives O, which, 
 by the addition of voice, becomes a good I (ee). In difficult cases I 
 have found it a good plan to give the pupil the idea that O is C 
 modified by expansion of the tongue so as to press sideways against 
 the molar teeth on each side instead of pressing up into the arch of 
 the palate. In some cases the simple direction to broaden the 
 tongue will corred the defective f (ee). 
 
 Pupils sometimes give X ov even 1 for f (ee), and sometimes f 
 T or 1. These defeats arise from the attempt to say ee with the teeth 
 too far apart. You cannot pronounce (ee) properly with the mouth 
 wide open. A teacher, however, is apt to separate the teeth as 
 much as possible in order to show her pupil the position of the 
 tongue. The pupil imitates the opening of the jaws, and this is apt 
 to result in a position of the tongue too far back (I or 1) or in a 
 position having too wide an aperture (IT or 1). 
 
 1 give below a tabulated list of the elementary vowels used in 
 the English language arranged in such a manner as to show their 
 place in the complete vowel scheme elaborated by my father. 
 
 Melville Bell's Vowel Scheme. 
 
 1 T f 1 T f 
 
 ] I C 3 I C 
 
 J 1 I J I I 
 
 i T f Iff 
 
 } \ i } I i 
 
 ill ill
 
 104 
 
 
 English 
 
 Vowels. 
 
 
 I 
 
 I 
 
 3 
 
 C 
 
 3 I - 
 
 i 
 
 — I 
 
 J I I 
 i 
 
 J 
 
 
 J 
 
 Some one has observed that you may at any time produce a 
 Scotch air by striking at random the black notes of the piano. The 
 musical scale of the Celts is defective. Examine the gamut of vowel 
 sounds as given by my father, and you will notice that the English 
 ear seems to be as defective for vowel sounds as the Scotch ear is for 
 musical notes. Only three of the back series of vowels are used in 
 the English language; two of the mixed series, and five of the front 
 series. Out of the eighteen round vowels, only five in all are em- 
 ployed in our language. 
 
 You will observe a curious likeness between the peculiarities of the 
 "front " and "back-round " series of vowels. The first one of each 
 series (I and 1) is always of long duration in English, and the seconn 
 (I and i) always short. The third (C and }) never occurs by itself, 
 but is used simply as the initial part of a diphthong. The sound 
 ends with the gliding of the tongue to the high vowel of its series. 
 Thus C is pronounced as f (a), finishing off with the gliding of the 
 tongue towards the position for I (ee) ; and }; oh is given as H fin- 
 ishing off with a glide towards the position for i (oo.) The fourth 
 vowel of each series is wanting in English. The fifth vowel of the 
 front series I {e in pet) is sometimes long and sometimes short. The 
 fifth of the back-round series, however, J {aw) is always long. In 
 both series the sixth vowel (J or X) is always of short duration. 
 
 It may also be noted that the mid-back vowel ] (// in 7tp) also is 
 always short. 
 
 These peculiarities of duration do not necessarily pertain to the 
 vowels, but are mere matters of English usage. Many persons have 
 the mistaken idea that the vowels in the words eat, and /// pool and 
 pull; caught and cot; calm and come, are the long and short forms 
 of the same vowels, but if you sing these words you will recognize 
 that the vowels remain distin6t to the ear when equally prolonged; 
 and you can shorten the vowels in the words eat, pool, caught and 
 calm, without producing //, pull, cot and come. In faft, this is the 
 pronunciation given for these words by French speakers of English.
 
 I05 
 
 I am inclined to think that there is some natural cause for the 
 analogous peculiarities appearing in the front and back-round series 
 of vowels, because I notice in both dialectic and individual utterance 
 that variations from the standard, appearing in one series, have their 
 analogues in the other. For example : where [r (a) is pronounced C 
 without any gliding of the tongue toward r as in Scotch and in Con- 
 tinental pronunciation, you find also that }^ (o) is pronounced 3" 
 without the ^ glide. 
 
 So, too, where individual speakers give [f or ir for [f (J) they 
 also usually say H or J^ for }'\ (o), etc. 
 
 Examine the table of English vowels and you will see that the 
 front and back-round series are nearly complete, and you will recog- 
 nize at once the importance of I (ee) and ^ (oo), for from them the 
 other vowels of their series can be developed by simply enlarging the 
 aperture. 
 
 The mixed vowels (land I) and the bacl^ vowels (3 3 and J) pre- 
 sent no difficulties, for the following reason : any sort of an indefi- 
 nite sound will pass for I (er in her) and 1 (the sound of the indefi- 
 nite article a in a sentence) differs so slightly from this that there is 
 no need of distinguishing between them. In unaccented syllables I 
 would express these two sounds indiscriminately by (I) the voice 
 sign. 
 
 3 {a in ask, path, etc.) also differs so slightly from J {<z in father, 
 calm, etc.) that there is no need to bother a deaf child with the dis- 
 tinction. 1 teach and write them both as 3 and few deaf children 
 have any difficulty in giving the sound. 
 
 Then, again, this may be considered as identical with 3. For 
 example: though the vowels in calf and cuff are really different 
 vowels in teaching the deaf, we may consider them as the long and 
 short form of the same vowel, because, as a matter of fact, a 
 deaf child gives 3 when he attempts to shorten 3. It should be 
 noted that this vowel 3 (// in up) like ^ (// in pull), and J (o in oji) 
 never occurs by itself or as a final element. I would not, therelore, 
 teach these sounds elementarily, but always in combination with a 
 succeeding consonant. The short effect should not be produced by 
 a sudden impulse of voice, but by cutting off the sound by the 
 assumption and prolongation of the succeeding consonant. For 
 example: let a child prolong (♦)the vowel 3 and wind up with a 
 softly uttered 3>, and you get the effect of "calf" (033 = 03*3 •). 
 \A'hereas, let the child attempt to give the same vowel sound, 
 but prolong the 3 position, jumping as quickly as possible
 
 io6 
 
 from the Q to the 3 position and you get the effeft of "cuff" 
 (a]3 = a33*>.) So with calm and come. If he tries to give the 
 same vowel sound to both, prolonging the 3 in the latter word, 
 passing quickly from the G to the B position, you will, as a matter 
 of fad get a]9 although he tried to say Cl39 with a shortened 
 vowel and a prolonged consonant. 1 find in the case of the other 
 short vowels i J and f that deaf children generally produce the proper 
 effect by attempting to pronounce them as i J and I, passing quickly 
 over the vowel to the succeeding consonant and prolonging it. 
 Thus if a child pronounces DiCO {pool) corredly, let him prolong 
 the W passing as quickly as possible from the D to the CO and he 
 produces the effect of DlOO {pull). So with the word foot. Let him 
 try to pronounce it as 3iD>, prolonging the shut position D of the 
 /, and passing as quickly as possible from the 3 to the D>, and you 
 get the proper vernacular effect 3iO>. So with J. You can con- 
 vert caught into cot by prolonging the shut position (D) of the /, 
 (D>) or gaud mto god, by prolonging the d (CD.) 
 
 The vov^el I is the most frequent vowel in the English language. 
 It is rarely necessary, however, to explain to a deaf child that it 
 differs from I in any other respect than length. As a general 
 rule, if the deaf child prolongs the shut position (D) of the / (D>) in 
 the word eat, passing quickly over the vowel position, the vernacu- 
 lar effed of it is produced (lD> = r04>). Indeed, in all short vowels 
 the succeeding consonant is prolonged. 
 
 Mr. Lyon: How would you distinguish the final ^j' in words; 
 would you use that same sound .? 
 
 Dr. Bell: Yes. 
 
 Mr. Lyon : In that case you would not have a consonant fol- 
 lowing. 
 
 Dr. Bell: That's true. The other short vowels i J and ] are 
 always succeeded by consonants. I think this is also true of I and 
 perhaps of I except when followed by I {er). I do not think any 
 of the short vowels excepting f occur as final elements. The correct 
 sound of final v is T (/ in it). As a general rule the pupil will give 
 the effect correctly if he tries to make the sound X {ee) carelessly, 
 with little muscular exertion. I, therefore, begin by getting him to 
 try to say X {ee) softly. If the effeft of T results, well and good. I 
 leave it alone. If, however, we obtain too pronounced an X (ee) I 
 then explain that the aperture is too small. The great trouble is 
 that if you tell a deaf child that X has a larger aperture than X he is 
 apt to exaggerate the difference and give too large an aperture.
 
 I07 
 
 Whereas, if you don't say anything at all about the aperture the 
 attempt to say I with little muscular exertion usually results in a 
 satisfadory I. 
 
 I give below a table of the English vowels, as I teach them to 
 the deaf, arranging them so as to show their place in my father's 
 complete vowel system already given. 
 
 - - I - I — 
 
 I — - - - — 
 
 J - — — — — 
 
 The short vowels f 1 J and 3, as 1 have already explained, may 
 be considered as identical with I i J and 3, save in exceptional 
 cases, where the organic difference must be explained. 
 
 The medium aperture vowels [ and } are only used in English 
 as the initial parts of the diphthongs Cr (a) and 3-^ (d). They are 
 usually so difficult of acquirement that we are generally forced to 
 accept If and J^ for Cf and }}. The distindion of sound is so slight, 
 however, that the error is surely immaterial. 
 
 £ (ee) and 1 {oo) in unaccented syllables become f (i) 
 and ^ (66) in ordinary speech. For example: the word tlw by 
 itself is 2i5r (thee), but when unaccented becomes 65l; 2*5lDjr 
 2»5fO[f3C0, ' thi boy, thi table," etc. The word to by itself is Di, 
 but in unaccented positions becomes Di, as "1 gave a book 
 Di^ASrDjr [loo thi boyj." 
 
 We are so accustomed to give indefinite vowels in unaccented 
 syllables that any sort of indefinite vowel effect is more acceptable 
 than a precise pronunciation such as would be given if the syllables 
 were accented. For example: DI65I Djr (tu thCi boy) rapidly and 
 indefinitely uttered would be more acceptable to ordinary ears than 
 Di2»5r0jr (too thee boy) precisely uttered. In fad, in all unaccented 
 syllables I is better than a precise vowel. I and i should be given 
 or r and 1, but I is better than an incorred I or i. 
 
 Mr. Lyon: That does not differ materially from the sound of 
 u in up as we usuiiily hear it. 
 
 Dr. Bell: It is an indefinite sound, somewhat like that, but 
 more like er in her. 
 
 Mr. Crouter: Dont we get that effed in the word "carpet" 
 by dropping out the vowels ?
 
 io8 
 
 Dr. Bell: Yes, we can use this indefinite sound (I) in place of 
 the r. 
 
 Mr. Crouter : I was not speaking of r, but take any case, the 
 consonants may be held together in a word and you pass over from 
 one consonant to the other and this indefinite vowel sound is 
 produced. 
 
 Dr. Bell: Yes, it usually occurs as a transitional effeft. For 
 example: In the last syllable of " carpet " carelessly uttered. 
 
 Miss Black: What sound do you give the final y ? 
 
 Dr. Bell: Give me a word, Miss Black. 
 
 Miss Black: Well, "Mary." 
 
 Dr. Bell: T (/in it). 1 would write the word Blld)!, al- 
 though many Americans say SCfCi)!. When the letter r occurs be- 
 tween two vowels, and the first one is long, English usage de- 
 mands the insertion of voice-glide (I) between the long vowel and 
 the medial r, thus fairy (3IICi)f) weary (^Iiul) fiery (33riCi)f) fury 
 (3011(1)1), etc. 
 
 When the letter r occurs finally or before a consonant, for ex- 
 ample — ear, poor, farm, warm, etc. ; the r (<i>) is not pronounced. 
 Even elocutionists demand only a gliding of the tongue toward the 
 position for r (i). Thus L'i Oh 3hB ^hB etc. 
 
 When deaf children, however, attempt to give glide r («), they 
 exaggerate the effed and produce a consonant sound. ((i>), thus 
 XCi) DlCi) 33Ci)9 m(j)B, etc. 
 
 The effect is at once recognized as peculiar. Defe(5live combina- 
 tions usually result in such words as 33Ci)IQ (farum) 9JCi)l3 (warum), 
 etc. When the consonant r itself is defective, as it often is, for 
 example: Ci)" or Ci)*=, the effed; is so unnatural that it would be 
 better to omit the r altogether. Many hearing people fail to give 
 an r of any sort in such words as the above, and 1 would recommend 
 the omission of the glide r (y.) in teaching the deaf. The voice 
 glide (I) answers every purpose and is easily uttered. When a deaf 
 child gives that in place of (!i!) no one but an elocutionist could 
 tell the difference. For example: Let the child give II (ear) Dil (poor) 
 33l3 (farm) J3JIQ (warm). Indeed, in the last cases even the voice 
 glide itself may be omitted without any very marked peculiarity. 
 Thus: 339 (fahm) ^}B (wawm). This surely would be more 
 acceptable than the pronunciation usually given by deaf children, 
 for these words. 
 
 When r occurs as a final and the next word begins with a 
 vowel, a consonant r is usually required. IjSy would be to ordinary
 
 109 
 
 ears a satisfadory pronunciation of the word "there," but if you 
 put that into a sentence where the next word commences with a 
 vowel (for example, "there is," etc.), then 2»5tl Ii5 would not prove 
 acceptable, and we must introduce a consonant r thus 63i;iCi)f2i5. 
 
 Now it is a difficult thing for a teacher to get a deaf child 
 to say D]DXi! without an exaggeration of the glide r element that 
 produces an unnatural effect, but any child can give 3]Oi which is 
 perfe(5tly satisfactory to most ears. If you didn't know there was 
 no r there, you would never find it out. I would have the deaf 
 child give simply the indefinite vowel (I) for the whole syllable er 
 in such words. 
 
 In English utterance two other glides (« and i) are employed in 
 the diphthongal vowels. 
 
 In forming s the tongue glides towards the position for H) (j^*) 
 or I ((v) ; and in forming i the glide is towar.ds the position for 53 
 (w)ori(oo. ) Diphthongal sounds present great difficulties to the 
 deaf and are rarely given corredly. Both the initial and final posi- 
 tions are apt to be wrong, and a strong tendency is manifested to 
 prolong the final instead of the initial part of the diphthong. 
 
 My plan of correction is to write what the pupils do, so that 
 they may see the difference between the sounds they actually utter 
 and those we wish them to give. 
 
 My father has not provided a sufficient number of glide symbols 
 to enable us to represent the incorreft sounds uttered by our pupils, 
 and I have, therefore, found it advisable to express glides by vowel 
 symbols upon a small scale. This gives us a sufficient number of 
 forms without introducing new symbols. For the deaf I write the 
 diphthongs U 3^ Js }i 3i as Cf 3r Jf }> 3V 
 
 The common defeat of prolongation ot the final element can be 
 expressed as Cf 31 if }l3i. and when the combination is dis-syllabic 
 this becomes Cr3rjl}i3i. 
 
 In teaching such a diphthong as 3r, I commence in the following 
 manner: 
 
 Say 3. Now say I. Now 313131, etc. 
 Now DlDlDl, etc. Now 3r3r3r, etc. 
 
 Whatever variations occur during the course of the lesson are 
 represented symbolically, and the pupil is requested to repeat them 
 in contrast with the corred sound. Thus: " You said 3l. Doit 
 again 31 31 31 etc." " Now give 3l3r 3l3r etc." •• Now you said If. 
 Try it again HirXr etc. Now give 3r 3Llr etc." 
 
 The principle of corredion is: ixritc icliat the pupil docs, and
 
 I lO 
 
 then get him to repeat the incorrect sound in contrast with the sound 
 voit wish hi 111 to utter. 
 
 I do not think it is possible to obtain great accuracy of vowel 
 quality without the use of symbols of some sort for incorred sounds, 
 and the adoption of the principle referred to above. The symbols 
 of visible speech are of great utility for this purpose. Indeed, I 
 believe them to be essential. Without them vou can only hope for 
 approximations to the corred vowel positions. Without them 
 your great, and indeed your only reliance must be upon imitation. 
 In any case the power of imitation should be developed by con- 
 stant pra(5lice before a mirror. 
 
 I always teach [r {a) and 3r (7) in contrast with one another, so 
 as to make the pupil familiar with the difference. These diphthongs 
 are apt to be pronounced alike as ir or •£. For example: A pupil 
 will say iGiIajQiI. instead of i33fCi3Q[r (fine day). For the same 
 reason 1 teach 3-^ and 3> in contrast. 
 
 In giving Jr, pupils generally give too small an aperture between 
 the lips for the initial position. I think this results chiefly from 
 the spelling (o/). They try to give an 6 followed by short /. 
 Indeed, very often the o is followed by glide oo, thus making a dis- 
 syllabic compound something likeJ^X. For example: D}^! (bo-ee) 
 for Djr (boy). 
 
 Sounds that differ only slightly from one another should, I 
 think, be taught together, in contrast, as the best means of securing 
 a distinction. Thus : teach [r 3r and jr as one group, and 3-1 and 3^ 
 as another. 
 
 I would also pradice such compounds as [fi 3ri JFi (vowels in 
 layer, liar and lawyer written COjri or Cl)J(T)|). 3-^i y\\ (vowels in 
 sower and sour). In such words as sore and more there is no 
 glide. Many persons say 233-I SJi. My father would write ^Ji! 
 93-!i:, but I should recommend a still larger aperture in teaching the 
 deaf. I would write UJi and 9Ji (saw-er maw-er) for "sore " and 
 "more," and D3->l and SJ^I (so-er and mo-er) for " sower " (one who 
 sows) and "mower" (one who mows). 
 
 It is very difficult to get a deaf child to distinguish I from I {e 
 in pet, from a in pat). It is important, however, that the attempt 
 should be made, as the slight difference of sound often makes a 
 great difference in the sense. For example: Met, mat, bet, bat, 
 etc. The distindion is best obtained, I think, by pradice before a 
 mirror.
 
 Ill 
 
 Vowels are so difficult of acquirement by the deaf with accu- 
 racy and precision that we may consider it fortunate that usage 
 tolerates considerable latitude in the pronunciation of these elements. 
 
 The precise shade of vowel quality given in one part of the 
 country is not heard in another. Travelers in England are startled 
 by the cry QID ^l DID ^1 (" Keb, sir, keb, sir,") from the cabmen 
 in London. The Irishman says "ai" for 1. Many Americans say 
 311 ya> for 3IiUD> (first) and everywhere we hear I^CCI> for 
 3-^CD} (oh, no). 
 
 Certain defects are recognized as individual, or family 
 peculiarities of speech; others are characteristic of whole commu- 
 nities and constitute a provincial utterance or dialed; and still 
 others reveal the nationality of the foreigner. 
 
 1 think I am pretty safe in saying that the "standard pronun- 
 ciation," like the "average school boy," nowhere exists! We all 
 depart from it in a greater or less degree. Study the character and 
 extent of the variations that exist among educated people and don't 
 be too critical of defects of your pupils if they fall within those 
 limits. 
 
 The pronunciation of the consonant elements of speech is so 
 uniform in all English-speaking countries that very slight variations 
 are received as foreign sounds, while greater departures from the 
 standard convey the idea that the vocal organs are themselves 
 defective. People speak of "curing" such defects, as though they 
 were diseases, or the result of malformations, requiring the surgeon's 
 care. Vowel peculiarities, on the other hand, fail to convey this 
 idea and are more suggestive of provincial or foreign utterance. 
 
 Small-aperture vowels, like i or I, are given everywhere with 
 substantial uniformity, and any marked deviation from the standard 
 is suggestive of foreign birth. Vowels of larger aperture are con- 
 stantly mispronounced by the best educated people. Even culti- 
 vated Bostonians, for example, sometimescall their own city 33DDiCC 
 (Bahston). Unusual variations from the standard if of slight extent, 
 are suggestive of provincialism, and where the departure is greater 
 the speaker is supposed to be a foreigner. 
 
 From this it will be seen that exaditude of pronunciation is 
 more necessary with certain sounds than with others. Consonants 
 and small-aperture vowels in accented syllables must be accurately 
 given; whereas considerable latitude may be allowed in the pro- 
 nunciation of medium and large-aperture vowels, and of diphthongal 
 sounds wherever they occur. This is fortunate, for these are just the
 
 I 12 
 
 sounds that are most difficult of acquirement with definiteness by 
 our pupils. 
 
 I would direft your attention to the very great importance of 
 training a child to retain a position unchanged, until dired:ed to re- 
 linquish it. The common practice of pronouncing an element of 
 speech and then immediately relinquishing the position leads the 
 child to consider the relinquishment as an essential feature of the 
 sound. In speaking he relinquishes one position before he assumes 
 the next, thus producing a transitional effe6l or glide sound between 
 the elements. This sound appears either in a vocal (i) or non-vocal 
 (>) form, according as the elements themselves are vocal or non- 
 vocal. 
 
 Ask the average deaf child to say BLdLbL (bee-bee-bee) and 
 you will obtain DiLBiLDiIi. The tongue is raised for I and de- 
 pressed for 3 so that it moves up and down for every syllable. 
 Pronounce BLbLbL yourself and you will find that the tongue does 
 not move at all but remains continuously in the I position. The 
 movement is entirely labial. This defeft of combination runs through 
 all the elements. 
 
 The great principle to be kept in mind is that positions do not 
 merely succeed one another like the letters on a printed page, but 
 overlap. A position must be retained until the mouth is in position 
 for the next element. 
 
 In teaching the principle of combination to a deat child, I 
 would recommend you to commence with the vowel 1. Get your 
 pupil to prolong I while you open and close his lips with your fin- 
 gers. At once you get BLbIbL. Dired the child's attention to the 
 faift that the tongue remains in the I position all the time, and in 
 fa£t that it does not move at all. Then get him to manipulate his 
 own lips and then to produce the effeft without manipulation. 
 When he can do this well, try Ci5l0X, etc., and the analogy of the 
 symbols will help him to give the correct effed. Let him have the 
 idea ot retaining the f position continuously while he moves the point 
 of the tongue. Then try QLqL, etc., retaining the I position while he 
 moves the back of the tongue. The chief difficulties of articulation 
 teaching lie not so much with the elementary sounds as with their 
 combination into syllables. A deaf child may be perfectly able to 
 give every element and yet be unable to utter a sentence that is in- 
 telligible to ordinary people. 
 
 The most important point, I think, in the whole of articulation 
 teaching is the thorough comprehension by teachers and pupils of 
 the law of combination.
 
 SPEECH-READING. 
 
 The substance of this le6lure will be incorporated in a special 
 Circular of Inforniation, regarding Speech reading, which Dr. Bell 
 has been requested to prepare. 
 
 After his remarks upon Speech-reading Dr. Bell said : 
 I should like in conclusion to say a few words upon the general 
 subjedl of articulation teaching. We don't know yet how best to 
 teach speech to the deaf. If we did we wouldn't be here. We have 
 come here to learn from one another in the hope of improving our 
 methods of teaching. Now I am inclined to think that the more 
 nearly we can pattern our methods of teaching after the method 
 adopted by nature in teaching speech to hearing children, the better 
 should be our results. It is certainly the case that the methods 
 usually employed in schools for the deaf do not even approximate to 
 the nursery method of the hearing child. Not one of the little hear- 
 ing children whom you may have left at home commenced by learn- 
 ing elementary sounds. Mothers do not begin with elementary 
 sounds and then combine them into syllables and words. The 
 mother speaks whole sentences even to the infant in arms. The 
 child listens and listens, until a model is established in the mind. 
 Then the child commences to imitate, not elementary sounds, but 
 whole words. Indeed, people grow up to adult life without ever 
 having uttered elementary sounds, and when they do come to study 
 them, it is for the purpose of improving and perfeding their speech. 
 With hearing persons the elements come last, not first. They con- 
 stitute the final, not the initial, exercises of articulation. I would 
 commend this fad to the serious attention of the members of this 
 Association. The question is often in my mind whether we are not 
 making a radical mistake, and whether it would not be better to 
 commence with sentences and whole words, rather than with ele- 
 ments, and accept imperfed speech from little deaf children as we 
 do from hearing children. 
 
 113
 
 114 
 
 If you copy the natural process, what you want first is the use 
 of speech, and then perfect the articulation as the child grows up. 
 In this conne(5tion I would commend to your notice the paper of Dr. 
 Greenberger upon the Word Method, which constitutes the first 
 Circular of Information. This is an entire reversal of the position 
 I assumed when I entered upon the work of articulation teaching. 
 But the more I think of it, the more I am convinced that a great 
 principle is involved. Words first and elements afterwards. 1 re- 
 cognize, however, that there are real practical difficulties in the way 
 of its application to the deaf If you once allow a deaf child to speak 
 in a defective manner is there not danger that the defective pronun- 
 ciation will become habitual } This is a serious objection and should 
 be carefully considered. In the case of the hearing child, a correct- 
 ive element is always present, he hears. He hears the model pro- 
 nunciation constantly used by those about him, and also hears his 
 own imperfect babble. His ear forms a medium of comparison 
 whereby he perceives the relation of the sounds he utters to those he 
 desires to make. In the case of the deaf child we m.ight anticipate 
 that a corrective element would also be present if he could see 
 speech as others hear it. 
 
 (i) He must see the model pronunciation constantly and clearly 
 repeated so as to fix it in his mind, to take the place of the conver- 
 sation that goes on in the presence of the hearing child ; and 
 
 (2) He must see his own imperfed speech so that he may per- 
 ceive the relation of the sounds he utters to the correct pronun- 
 ciation. 
 
 It may be well to consider how far it may be possible for us to 
 bring about these conditions. 
 
 I. Speech-reading fulfills the first condition only in part. The 
 visible movements of the mouth may recall the model pronunci- 
 ation to the mind of the deaf child, when once acquired, but it 
 does not exhibit the pronunciation with clearness and definiteness to 
 the eye. If we could supplement speech-reading by books and 
 periodicals printed in phonetical type, a great advantage would be 
 gained. In Germany and Italy, where oral methods are most suc- 
 cessful, spelling corresponds to pronunciation, and this first condi- 
 tion is, therefore, fulfilled by the ordinary literature of those countries. 
 In English-speaking countries, however, ordinary literature is of 
 comparatively little use in impressing upon the memory of the deaf 
 child the corred pronunciation of the language. Our spelling is so 
 irregular and unphonetical that even hearing people often have to
 
 115 
 
 resort to a diftionary to ascertain how a word should be pro- 
 nounced. What we most need is reading matter for our pupils in 
 which the words are spelt as they are pronounced. In two hours 
 a deaf child can read as many words as a hearing child hears in 
 the course of a day, and if the spelling only corresponded to the 
 pronunciation, reading would fix the model in his mind, and speech- 
 reading would constantly recall it. Any kind of phonetical alpha- 
 bet would do for this purpose ; but Visible Speech would be especi- 
 ally advantageous because it would be possible through this agency 
 to fulfill the second condition also. 
 
 2. The deaf child must see the relation between the sound he 
 utters and the correcft sounds of speech. 1 know of no other 
 means of accomplishing this end than Visible Speech or the Lyon 
 Manual, — but they will do it. 
 
 1 think with these agencies we would have a corre(ftive element 
 that would lead to improvement of speech as the child grows up 
 and permit of the adoption of a more natural method of teaching 
 than now exists. 
 
 I throw out these thoughts as suggestions merely, for I recog- 
 nize, of course, the great difficulty of carrying them pradically into 
 execution, I would have you, however, appreciate the importance 
 of the principle involved, and consider whether in the face of the 
 difficulties that present themselves, it would be better to abandon the 
 principle, or study the difficulties and attempt to remove them. I 
 think that that method which conforms most nearly to the method 
 whereby hearing children require speech, will be most worthy of 
 adoption by teachers of the deaf.
 
 APPENDIX. 
 
 VOWEL THEORIES. 
 
 BY ALEXANDER GRAHAM BELL. 
 
 Read before the National Academy of Arts and Sciences, April 15, 1879, and reprinted 
 from the American Journal of Otology, Vol. 1., July, 1879. 
 
 Hemholtz has shown that an educated ear perceives a combina- 
 tion of musical tones where an uneducated ear supposes a single 
 sound ; and his theory, that the feeble, usually unheard, musical 
 tones are the cause of the peculiar sensation we term the "quality" 
 of a sound, seems now to be universally accepted as correct. 
 
 According to this theory, a vowel is a musical compound, con- 
 sisting of a mixture of musical tones of different pitches and vari- 
 ous intensities. The lowest, or fundamental tone, gives the pitch 
 to the whole, and is determined by the rate of vibration of the 
 vocal cords. 
 
 It is certainly the case that an attentive ear can perceive, in 
 every vowel uttered, a number of distinct musical sounds ; but the 
 hypothesis that the ear perceives them 07ily, and that it is unable 
 to appreciate the qualit}^ of the vowel directh', should be received 
 with caution. According to Helmholtz, the human ear is incapable 
 of perceiving any other than simple pendular vibrations, and it is 
 therefore under the necessity of splitting up a vowel into its con- 
 stituent musical elements before it can perceive its quality. 
 
 Thus Helmholz holds that vowels are itif erred from the pres- 
 ence of certain musical tones, and that they do not give rise in the 
 ear to distinct sensations of their own. 
 
 The rods of Corti are supposed by him to analyze the vibrations 
 imparted to the liquid of the internal ear, so as to split them up 
 into the pendular motions of which they are theoretically com- 
 posed; but it cannot be received as proven that the simultaneous 
 vibration of certain of these rods gives rise to the perception of 
 the quality of a sound ; for Pritchard and other comparative anat- 
 omists have .shown that the rods of Corti are entirely wanting in 
 parrots and other birds that imitate and therefore perceive the 
 sounds of speech. 
 
 Whether or not, however, there exi.sts in the human ear an apparatus 
 for taking direct cognizance of the quality of a sound, Helmholtz 
 has proved, by the synthesis of vowel sounds, that there exists the 
 most intimate relation between certain combinations of musical tones 
 and the qtiality of a sound. Not only arc such tones audible whenever 
 a vowel is produced, but the converse is e(|ually true ; and we may 
 therefore assume, at least as a working hypothesis, that vowels are 
 
 117
 
 ii8 
 
 compound musical tones. I shall follow the example of Ellis' in 
 terming the nuisical constituents of vowel sounds, "partial" tones, 
 to designate their subordination to the compound as a whole. 
 
 The illustrations of vowel-s^-nthesis given b}' Helmholtz show vowels 
 composed of tones whose frequencies are multiples of the fundamental 
 and the question arises: ''Are the upper partial tones, which are 
 characteristic of vowel sounds, invariably har7nonics of the fundamental 
 of the voice, or are they independent of it f 
 
 Again: "if they are harni07iics, do they iifiiformly bear the same 
 1-elation to the fundamental, whatever the pitch of the voice may be?''' 
 
 These questions at first sight seem to receive different answers, 
 accordingly as we attempt to solve the problem (i) from a con- 
 sideration of the organic formation of the vowel sounds, or from an 
 examination of the records, (2) of the phonautograph, and (3) of 
 the phonograph. The results common to all three methods of investi- 
 gation, however, point to a theorj- of vowel sounds in close accordance 
 with the ideas of Helmholtz, as expressed by Ellis in his "Earl}' 
 English Pronunciation," Part IV., p. 1277. 
 
 In examining these questions, it may be convenient to designate, by 
 distinct names, two varieties of Helmholtz's vowel theor}-, concerning 
 which there has been of late considerable discussion : 
 
 Ihe fixed pitch hypothesis. — The upper partial tones characteristic 
 of vowel sounds may be supposed to have fixed, invariable pitches, 
 and the element of pitch may be considered the distinguishing feature. 
 
 The harmonic hypothesis.- — According to this h3-pothesis, the upper 
 partial tones characteristic of vowels are always harmonics of the 
 fundamental, varying in pitch with it, the vowel characteristic lying 
 in the predominance of certain harmonics. 
 
 As an illustration of the difference between these hypotheses, let 
 us consider for a moment the musical composition of the vowel 6. 
 
 Helmholtz .states that he produced a ver}* fine b}- combining the 
 sounds of certain tuning-forks whose rates of vibration were multiples 
 of that of the lowest fork. The prime tone was B^; the 2d, 3d, and 
 5th forks were allowed to sound feebly, and the sound of the 4th fork 
 was brought out much more strongly. In this experiment, then, the 
 characteristic tone [b^f] was the double octave of the fundamental. 
 Now, if the fixed pitch hypothesis be correct, the vowel 6 should always 
 be distinguished by a partial tone of the [by') ; whereas, if the harmonic 
 ■hypothesis be correct, the predominant partial should vary in pitch 
 with the pitch of the voice, and be alwa^-s its double octave. 
 
 I. lozc'c/ Theories, considered in Relation to the Organic For- 
 mation of Wnvcl Sounds? 
 
 When we examine the vocal organs we find numerous cavities located 
 in the thorax, larynx, pharynx, nares, and in the mouth. The air in 
 
 1 The phraseology of Ellis will be used throughout this paper. For definitions, etc., 
 see Ellis's translation, " The Sensations of Tone," footnote to page 3fi. 
 
 - In all references to the organic formation of speech sounds, I adopt the phraseology 
 of Melville Bell. For definitions, etc., see "Visible Speech : The Science of Universal 
 Alphabetics."
 
 119 
 
 each cavity has a tendency towards a definite rate of vibration, and 
 when agitated in any way produces its resonance tone. 
 
 In the act of speech the air is set in vibration in all these cavities, 
 the resonance tones of the cavities mingle with the tones due to the 
 vibration of the vocal cords, and thus produce the complex sounds of 
 human speech. The movements of the tongue, lips, etc., modify the 
 shape and size of some of these resonance cavities, and thus enable us 
 to produce sounds the musical constituents of which are almost 
 infinitely variable at will. The constant cavities of the vocal organs, 
 the shapes of which are determined by nature, and are therefore inde- 
 pendent of will, probably give to the speaker's voice that individuality 
 of tone that enables us to pick out the voice of one speaker from a 
 multitude of others, while the variable cavities give prominence to 
 partials that characterize the elements of spoken language. 
 
 The cavities of the mouth are chiefly concerned in the production 
 of vowel quality. When a vowel position is assumed by the vocal 
 organs, the mouth-passage is slightly constricted at some particular 
 part (see Fig. i), and thus two resonance cavities, a and b, are estab- 
 
 FlG. I. 
 
 li.shed, the interior of the mouth somewhat resembling in shape the 
 interior of a chamber formed by placing two bottles neck to neck, the 
 two resonance chambers being represented by the bodies of the bottles, 
 and the con.stricted passage between them by the necks. 
 
 I have found that the resonance tones of these cavities can be 
 readily studied in the following manner: 
 
 To Study the Pitch of the Posterior Cavity : 
 
 Clo.se the glottis, assume the vowel position, and tap genth' again.st 
 the throat with the thumb-nail. (A .sound will be perceived .some- 
 what similar to that produced by tapping against the side of an empty 
 bottle). A double pitch will be noticed, but the tone due to the 
 jjosterior cavity, a, will be much more fully produced than that due 
 to the other. I have succeeded in making the sound audible to large 
 audiences by placing the forefinger of the left hand against the throat, 
 and tapping it very forcibly with the thumb-nail of the right hand.
 
 120 
 
 A loud sound can also be produced by striking a piece of wood or 
 cork held against the throat. 
 
 To Study the Pitch of the A^iterior Cavity : 
 
 Close the glottis, assume the vowel position, and strike gently a 
 piece of wood, or cork, held in front of the mouth or against the 
 cheek. I have found that an ordinary lead-pencil, held firmly against 
 one side of the mouth, readily yields the resonance tone of the mouth 
 cavity when struck with the thumb-nail. A double tone can be per- 
 ceived, but that due to the anterior cavity is much more prominent 
 than the other. 
 
 The tone due to the anterior cavity may be studied alone, by depress- 
 ing the soft palate until it touches the back of the tongue. (This is the 
 position for " ng " in the word " sing.") Under such circumstances the 
 soft palate cuts off all communication with the air in the posterior cavity, 
 and a single resonance accompanies each vowel position. When the 
 absolute pitch of the anterior cavity is sought, the former method is 
 preferable, as the depression of the soft palate alters the tone. 
 
 The tones of the cavities are best brought out by contrast. For 
 instance, assume successively the positions for certain vowels, and 
 observe the series of tones produced, first by the anterior cavit}^ and then 
 by the posterior cavity. The difference will be found to be very striking. 
 
 When the vowel positions are assumed in the order shown by Melville 
 Bell, in his " Visible Speech," the tones of the cavities are found to be 
 arranged in regular mvisical sequence. For instance, commencing with 
 the high-front vowel (ee), assume successively the positions for the other 
 vowels of the front group. A double series of resonances will be obtained, 
 so arranged that the tones of one series fall in pitch while the tones of 
 the other rise. The same rernark is true for all the organic groups of 
 unrounded vowels. The pitch of the anterior cavity falls iti pitch as the 
 vowel aperture is enlarged, and that of the posterior cavity rises. It is 
 different, however, with rounded vowels. For in.stance, take the high- 
 front-round vowel, (ii in German), and assume successively the positions 
 for all the vowels of that group down to the low-front-wide-round vowel. 
 The reso7iance tones of the anterior and posterior cavities both rise in pitch 
 as the vowel aperture is enlarged. All the groups of rotinded vowels behave 
 in a similar ynanner. 
 
 Again, compare unrounded vowels of similar aperture, but of different 
 organic formations. For instance, assume successively the high-front, 
 high-mixed, and high-back vowel positions : the pitch of the anterior 
 cavity falls and the pitch of the posterior rises. The same thing occurs 
 when we compare ihe mid-front, mid-mixed, and mid-back vowels, or 
 the low-front, low-mixed, and low-back vowels. 
 
 Comparing in like manner rounded vowels of similar aperture, but of 
 different organic formation, the same fact is noticed, namely, that the 
 further back in the mouth the point of constrictioyi is located, the lozver is t/ie 
 pitch of the anterior cavity and the higher the pitcli of the posterior. 
 
 Comparing rounded and unrounded vowels of the same organic forma- 
 tion and aperture, as, for instan^ie, the high-front vowel (ee) with the high-
 
 121 
 
 front-rounded vowel (ii in German) : the pitch of the posterior cavity is 
 the same for both vowels ; but the pitch of the anterior is lower for the 
 rounded (ii) than for the unrounded vowel (ee). 
 
 The changes of pitch produced in the anterior and posterior cavities 
 of the mouth become intelligible by reference to familiar facts of resonance. 
 Thus blow across the mouth of an empty bottle, and its resonance tone 
 can be perceived in the rustling sound caused by the breath. Pour water 
 into the bottle, and the pitch of the tone becomes higher. Place your 
 fingers over the mouth of the bottle, so as to reduce the size of the open- 
 ing, and the pitch falls. 
 
 It will thus be seen that the pitch of a cavity falls when its interior 
 capacity- is increased, and also when the exterior orifice is contracted. 
 
 The depression of the tongue (see Fig. i) should, therefore, cause an 
 elevation of the pitch of cavity a, and a lowering of that of cavity b ; for 
 the air space in cavity b l>ecomes larger when the tongue is depressed ; and 
 the width of the exterior opening [c] of cavitj' a is at the same time 
 increased. 
 
 Evety position assumed by the vocal orgatis deter7nines the shape and size 
 of the cavities of the mouth, and thus determines the absolute pitch of the 
 resonance tones proper to those cavities. 
 
 When air is passed through the mouth, as in the act of speech, a new 
 element enters into the inquiry : Is vowel quality due to the mouth 
 position assumed by the organs, or is it caused by the vibration of the vocal 
 cords ? If the former, we would expect that the characteristic upper 
 partials would bear some relation to the resonance tones proper to the 
 mouth cavities and be independent of the pitch of the voice. If the 
 latter, we would expect them to bear some relation to the pitch of the 
 voice and be independent of the pitch of the mouth cavities. 
 
 The mere passage of air through the mouth is sufiicient to bring oi:t 
 the characteristic tones of the mouth-cavities. Vowel quality is audible 
 in a whisper, and can even be produced hy forcing air into the mouth by 
 means of a pair of bellows. 
 
 It should be noted, in this connection, that vowels are general h' pre- 
 ceded in actual utterance by elements that approximate very closely to 
 consonants in their organic formation ; "initial vowels" being preceded 
 by the throat-shut consonant — an element belonging to the same general 
 class as p — T — k, but for which we have no letter in the English language. 
 
 Consonants result <"rom obstructive positions of the vocal organs. 
 Dtiring the utterance of speech the air in the thorax is continually com- 
 pressed by the action of the abdominal muscles, diaphragm, and muscles 
 of the thorax, so that when the emission of breath is momentarily checked 
 by the formation of some obstructive position, the cavities behind the 
 point of constriction become inflated. 
 
 Thus, when a vowel is preceded by a consonant — a sudden puff of air 
 accompanies the relinquishment of the consonantal position — and this 
 j)urf, passing through the vowel-configuration, is sufficient to bring out 
 the characteristic tones of one oi more of the vowel-cavities quite 
 independently of the vibration of the vocal cords. 
 
 An interesting case may be mentionetl which bears u])on this i)oint. 
 Dr. Moore, of Rochester, N. Y., had a patient whose glottis had become
 
 122 
 
 closed by disease. For twenty-five years the man had been dependent for 
 life upon air supplied through a tube inserted in the trachea. He could 
 speak, although no particle of air could be forced into the mouth from 
 the lungs. His speech was perfectly intelligible, distinct, and even loud, 
 but of course peculiar, on account of the absence of voice. He could not 
 pronounce vowel sounds by themselves, but they were distinctly audible 
 when preceded by consonants. By long practice the man had acquired the 
 power of contracting and expanding the cavity at the back part of the 
 mouth to a wonderful extent. The air which was moulded into speech 
 was alternately drawn into this cavity and expelled from it by the forcible 
 action of the muscles of the pharynx, soft palate, and back part of the 
 tongue. The cavity seemed to be continually in a state of change — alter- 
 nately expanding and contracting during the whole progress of his 
 articulation — so that the man could speak for any length of time without 
 pausing for breath ! There seemed, however, to be an instinctive remem- 
 brance of the connection between breathing and speech, for he was in the 
 habit of expelling air through the tracheal tube while he was speaking, 
 and of remaining silent when he drew air into the lungs. 
 
 Another curious case has been made public by Dr. Moore.^ 
 
 A patient had attempted suicide by cutting his throat. The epiglottis 
 was severed trom the larynx, and, when the man attempted to articulate 
 with his head thrown back, the air passed out through the opening in his 
 throat instead of through the mouth. Under these circumstances the 
 man could pronounce intelligibly certain vowel sounds. Dr. Moore 
 satisfied himself and other observers that there was no passage of air 
 through the mouth, by artificially closing the aperture between the 
 epiglottis and back of the pharynx. 
 
 Dr. Moore argued from this experiment that the vowels heard were 
 produced in the larynx instead of in the mouth. This position, however, 
 seems to be untenable ; for the mouth positions for these vowels might 
 have been assumed during the production of the sounds, and the tones 
 of the mouth cavities would be brought out by sympathetic resonance 
 without the necessity of passing air directly through the mouth. 
 
 In whistling, the resonance tone of the anterior cavity is brought out 
 so loudly as to constitute a clearly recognizable musical tone. A careful 
 observer will find that his tongue assumes a definite position for each 
 note whistled, and a person can be made to whistle an air unintentionally 
 by making him attempt to whistle certain vowels in succession. In 
 whistling, it seems necessary that the constricted passage in the mouth 
 should be much more narrowed than in actual articulation, and that the 
 anterior orifice should also be small. 
 
 Sing such a vowel as 6 and gradually contract the passage between 
 the back of the tongue and the soft palate. The resonance tone of 6 will 
 be observed to grow in intensity as the passage is contracted, until finally 
 the vowel is converted into a vocalized whistle. By labializing the 
 various vowel positions, the resonance tone of the anterior cavity can be 
 brought into great prominence, and a whistle produced of definite fixed 
 pitch for each vowel-position. 
 
 ISee Trans. N. Y. State Medical Society for 1872
 
 123 
 
 It is found that the pitch of the voice can be varied without appreciably 
 affecting the pitch of the vowel-whistle. It is certainly difficult to bring 
 out the whistle of 66 or o upon certain pitches of the voice, but the high- 
 front and high-mixed vowels labialized can be easily retained in a 
 whistlino- condition while the voice glides upwards or downwards. 
 
 In studving the double resonance of the niouth-cavitj', I have been 
 led to the belief that the fundamental of the anterior cavity is much more 
 essential to the production of vowel quality than that of the posterior. 
 
 If we prolong the sound of the voice, and study the effect of the 
 movement of the different vocal organs in modifying the quality of the 
 resultant sound, the attention will be arrested by the fact that movements 
 of the organs further forward than the back of the tongue produce 
 changes of vowel quality, but that motions of the parts behind the back 
 of the tongue do not. The motions of such parts produce quite as marked, 
 if not more marked, changes of quality than in the former case ; but the 
 resultant sounds would not ordinarily be designated as vowel variations. 
 We should rather speak of them as changes in the quality of the voice. 
 For instance, the depression of the soft palate produces a nasal effect, and 
 the movement of the base of the tongue towards the back of the pharynx 
 produces a " guttural " quality of voice. 
 
 If the passage between the base of the tongue and the back of the 
 pharj-nx be contracted lateralh', bj^ approximation of the posterior pillars 
 of the soft palate (shown b}- dotted lines in Fig. 2), a very curious change 
 
 Fig. 2. 
 
 of quality is produced. The voice acquires a metallic ring, somewhat like 
 the tone of a brass wind instrument. When the posterior pillars of the 
 soft palate approximate so closel}' as almost to touch, a very disagreeable 
 reedy quality of voice results, which can perhaps be best described as a 
 sort of " Punch-and-Judj- " effect. 
 
 When these various motions are produced while a vowel position is 
 assumed, the pitch of the posterior vowel cavity is affected, and the 
 quality of voice accompanying the vowel is changed, but not the vowel 
 itself. 
 
 From this it seems evident that the anterior cavity is more important 
 in determining the vowel qiiality than the posterior cavit}'. 
 
 An examination of the mechanism of speech leaves the mind decidedly 
 biased in favor of the fixed pitch theory of vowel sounds. 
 
 II. Voiucl Theories, considered in /he LigJit of Experiments ivith the 
 
 PJi onantograpli . 
 
 If the harmonic hypothesis be correct, and vowels are composed of 
 partial tones whose frequencies are multiples of that of the fundamental
 
 124 
 
 of the voice, we should expect, from the researches of Fourier, that the 
 tracings obtained from the phonautograph for vowel sounds should be 
 invariablj- periodic curv-es, whatever the pitch of the voice might be. 
 Whereas, if the fixed pitch hypothesis be correct, vowels should not yield 
 periodic curves when tones of voice are used which do not contain the 
 fundamentals of the mouth cavities amongst their harmonics. 
 
 Fvirtliermore, if the harmonic theory be correct, the predominant 
 partial tones, bearing always a fixed ratio to the fundamental in pitch and 
 loudness, should produce for each vowel sound a definite form or forms of 
 curve, which should be constant for the same vowels under different 
 pitches of the voice. 
 
 Early in 1874 I carried on a series of experiments with an improved 
 form of phonautograph devised by Mr. Charles A. More}^ of the Institute 
 of Technology, in Boston. Vowel sounds were sung to various pitches, 
 and their tracings preserved for study and comparison. The results were 
 briefly as follows : 
 
 1. Vowel sounds uniformly produced periodic curves, whatever pitch 
 of voice was employed. 
 
 2. The form of vibration was not a constant characteristic. 
 
 3. Different vowels sung to different pitches often seem to produce 
 similar curves. 
 
 4. Different vowels sung to the same pitch traced curves of different 
 shapes, but they were not sufiiciently marked to enable the vowels to be 
 certainly identified. 
 
 5. There seemed to be a relation between the complexity of the tracing 
 and the vowel aperture ; close aperture vowels yielding curves that 
 approximated very closelj'to simple pendular vibrations. 
 
 Thinking that the results obtained with Mr. Morey's phonautograph 
 might be influenced by the imperfection of the apparatus employed, I 
 tried the tympanic membrane of a human ear as a phonautograph . Dr. 
 Clarence J. Blake, of Boston, suggested this idea, and kindly prepared a 
 specimen for me, with which I carried on experiments. The tympanic 
 membrane and the ossicula were moistened with glycerine and water, and 
 a stylus of hay attached to the incus enabled me to obtain tracings of 
 vowel vibrations on sheets of smoked glass passed rapidly underneath. 
 The results obtained with this apparatus were similar to those obtained 
 with Mr. Morey's phonautograph, and I found it impossible to recognize 
 the various vowel sounds by their tracings. I do not know the full results 
 obtained by Prof. Eli W. Blake with his photographic phonautograph, but 
 all the vowel curves drawn by him, that I have seen, were periodic curves, 
 and seemed to support the conclusions noted above. The general indica- 
 tions of all forms of phonautograph seem to favor the harmonic hypothesis 
 much more than the other. The unstable character of the vibration -forms 
 might be explained by supposing the phases of the harmonic partials to 
 have varied at different times ; for Helmholtz has shown that the phases 
 of the upper partial tones are immaterial to the perception of vowel 
 quality.
 
 125 
 
 III. Vowel Theories considered in the Light of Recent Experiments 
 
 with the Phonograph. 
 
 Mr. Edison's phonograph furnishes us with an instrument which 
 fortunately can be utilized in the solution of acoustical problems that had 
 before seemed insoluble b}' experimental methods. I believe that this 
 instrument can be emploj-ed as a means of ascertaining the truth or 
 falsity of the harmonic hypothesis. 
 
 If the differences of vibration-forms obtained by the phonautograph 
 for the same vowel, at different pitches, merely indicated a difference of 
 phase of the upper partials — the same harmonics being predominant at 
 each pitch of the voice — then, if the relative phases of the harmonic 
 partials could be retained when the pitch of the voice was changed, the 
 same vowel at different pitches should be characterized by the same curve ; 
 and a vibration of uniform shape impressed upon the tinfoil of the 
 phonograph should produce the same quality of vowel, whatever the 
 speed of rotation of the cylinder might be. 
 
 If, on the other hand, my observations with the phonautograph were 
 correct, that different vowels could be found, which, when sung to different 
 pitches, produced the same tracing, then a vowel sung to the phonograph, 
 while the cylinder is turned at a certain rate of speed, should be repro- 
 duced by the instrument as a different vowel when the speed was 
 changed. 
 
 When Mr. Preece exhibited the phonograph before the Physical 
 Society of London, on the 2d of March, 1878, I suggested that this 
 experiment should be tried, and stated my belief that the quality as well 
 as the pitch of a vowel would be affected by the speed at which the 
 cylinder was turned. The experiment was at once made, and the results 
 were apparently as I had anticipated. (See Nature, Vol. XVII., p. 415.) 
 
 Prof. Fleeming Jenkin and Prof. J. A. Ewing about the same time 
 performed a similar experiment in Edinburgh, but arrived at quite 
 different results. They .stated, in a letter, dated ]\Iarch nth, 1878, that 
 " the pitch is, of course, altered, but the vowel sounds retain their quality 
 when the barrel of the phonograpli is turned at ver}- different rates. 
 "We have made this experiment at speeds var>nng from about three to one, 
 and we can detect no alteration in the qualit}' of the sounds .' ' (See Nature, 
 Vol. XVII., p. 384.) 
 
 Such a result determined me to repeat the experiment carefully and at 
 leisure. Mr. Stroh, the eminent mechanician of Hamp.stead Road, Lon- 
 don, kindly permitted me to use his automatic phonograph, the C3'linder 
 of which was moved by clock-work. Mr. Alexander J. Ellis was present, 
 and a.ssisted during the experiments. 
 
 We found it extremely difficult to use the phonograph in the observa- 
 tion of minute phonetical distinctions. While we differed in our appre- 
 ciation of some of the effects produced, we agreed in thinking that vowel 
 quality was affected to .some extent by the .speed of rotation; but we were 
 unable to determine either the amount or the nature of the change. 
 Among other results of these experiments, the vowels in the words 
 mC-an, mane, men were often reproduced to my ear as apj)roximately 
 the vowels in moon, moan, mOrn; the reproduced ee sounding to me as
 
 126 
 
 an extremely faint 65. Mr. Ellis, however, could not agree with me in this 
 conclusion, although he admitted that the qualit}^ of these vowels was 
 changed in the reproduction. 
 
 The results of our experiments were communicated to Nature by 
 Mr. Ellis, in a letter, dated the 3d of April, 1878. (See Vol. XVII., p. 485. ) 
 Dr. Clarence J. Blake, of Boston, and Prof. Cross, of the Institute of 
 Technology, in the same city, were as much struck as I had been by the 
 .statements made by Fleeming Jenkin and J. A. Ewing concerning the 
 fixity of vowel quality under varying speeds of rotation, and repeated 
 the experiments alluded to. Very striking differences of vowel quality 
 w^ere perceived by them. Prof Cross communicated the results to 
 Nature, in a letter, dated Boston, April 29th, 1878. (See Vol. XVIII. , 
 P-93) 
 
 This called forth a response from Edinburgh, dated May 29th, 1878. 
 Fleeming Jenkin and J. A. Ewing modified their former statements 
 concerning the fixity of vowel quality, but stated that "the five vowels 
 a, e, i, 0, u (Italian), pronounced in succession are, b}' contrast at least, 
 thoroughly distinguishable when the instrument is run at various speeds, 
 such as to reproduce the sounds at all the pitches within the compass of 
 the average human voice. That no marked change is produced in the 
 relative values of the vowels is confirmed by the fact that neither in 
 public nor private exhibitions do the hearers of sentences, alternately 
 run slow and fast, suggest that the vowels have changed with a change 
 of speed. . . . We do not, however, think that our instrument 
 speaks with sufficient distinctness to warrant our expressing an opinion 
 as to the constancy of quality of any single vowel when the instrument 
 is run at various speeds." (See Nature, Vol. XVIII., p. 167.) 
 
 Since the publication of this letter no notice seems to have been taken 
 of this most interesting subject until quite recently, when Mr. Preece 
 and Mr. Stroh revived the discussion in The Electrician, for March 29, 
 1879. Amongst other results they observed that the vowel ah is con- 
 verted by a slow rotation into 6, and that the converse is equally true. 
 This fact has also been independently discovered in America by Mr. Fran- 
 cis Blake and myself. 
 
 I have at various times, during the past few months, made experi- 
 ments with phonographs of different kinds to determine the question of 
 vowel change or fixity; and all the instruments have answered the ques- 
 tion of vowel fixity, under changing speeds of rotation, in the negative. 
 Some experiments made by Mr. Francis Blake and mj'self, on the 
 fifteenth day of March, 1879, "ot only demonstrated that vowel quality 
 does change under varying speeds of rotation of the cjdinder of the 
 phonograph, but also manifested the direction and nature of the change. 
 The ordinary mode of conducting the experiment previously, was as 
 follows : 
 
 A vowel vi'as sung to the phonograph while the cylinder was turned 
 at a uniform rate of speed. The sound was then reproduced from the 
 instrument, while the barrel was turned at a uniform, but different rate 
 of vSpeed. When the experiment was conducted in this way, the change 
 of vowel quality was not marked, and very uncertain results were 
 obtained.
 
 12/ 
 
 The experiment was now varied as follows: 
 
 A vowel was sung into the phonograh in a high-pitched voice, while 
 the cylinder was turned at a uniform but high rate of speed. When the 
 sound was reproduced, the cylinder was started at a high rate of speed 
 and allowed to come gradually to rest. At once the nature of the vowel 
 change became manifest. The vowel iih changed b}- insensible degrees 
 to awe, oh, and finally ob. (The same effect can be produced by gradually 
 contracting and "rounding" the orifice between the lips, while at the 
 same time the back of the tongue is slightly raised. ) 
 
 The vowel ee was gradually converted into the German vowel ii. I 
 am perfectly sure of the labial element of this change, but am uncertain 
 whether there was not also a change in the lingual element. The repro- 
 duction was ver}- faint, but to my ear the vowel finall)^ produced was 
 either the " high-front-round " or the " high-mixed-round " vowel. 
 
 The diptliongal vowel i long was reproduced approximately as "ow" 
 in " now." The exact reproduction seemed to be as follows: The initial 
 sound was the " low-back-wide-round " vowel, gliding finally to "high- 
 mixed-round." 
 
 A long series of experiments with various vowels satisfied us that the 
 reduction of the speed of rotation below the original rate at which the 
 cylinder was turned, occasions an effect analogous to that produced by 
 labializing the original sound. 
 
 It was also evident that there was a lingual element of change, espe- 
 cially when vowels of the " front " and "mixed " series were tried; but it 
 was extremely difficult to locate the resultant sounds. The fixed pitch 
 h^'pothesis offers a clear undenstanding of the natvire of the change. 
 
 With decreasing speed of rotation, the prime tone and the characteristic 
 upper partials fall simultaneously in pitch. Upon our theory the char- 
 acteristic partials correspond to the fundamentals of the cavities formed 
 in the mouth by assuming some position of the vocal organs. Hence we 
 might expect that as the speed of rotation decrea.sed, the sound produced 
 would correspond to a vowel having anterior and posterior cavities of 
 lower pitch than the original one. 
 
 The pitch of the anterior cavity- can be lowered by gradually approxi- 
 mating and rounding the lips, and the pitch of the posterior cavity can be 
 similarly changed b}' contracting the passage between the two cavities. 
 The change of vowel quality produced by decreasing the speed of rotation 
 of the cylinder of the phonograph thus seems to correspond to the change 
 pnxluced by gradually elevating the tongiie in the mouth and at the 
 same time contracting and rounding the orifice between the lips. 
 
 Effect of Iiicreasiiii^ the Speed of Rotation. 
 
 A vowel was sung into the phonograph, in a low-pitched voice, while 
 the cylinder was turned very slowly. When the .sound was reproduced, 
 the cylinder was started slowly, and the rate of rotation gradually 
 increased. 
 
 The vowel u/i changed gradually to fi in ask, and then to a (in the 
 word man). This change of vowel cfiect was accompanied by a remark- 
 able change in the quality of the reproduced voice. The sound was
 
 128 
 
 accompanied by a metallic twang. This disagreeable quality became 
 more and more marked as the speed of rotation was increased, until a sort 
 of " Punch and Judy" squeak was produced. 
 
 Upon the fixed pitch hypothesis this change also becomes intelligible. 
 
 Organically considered, the change corresponds to a gradual contrac- 
 tion of the posterior cavity, accomplished by approximation of the 
 posterior pillars of the soft palate (as in Fig. 2), accompanied by a gliding 
 forward of the lingual position. This is exactly what one would expect 
 upon the supposition that the posterior and anterior cavities were of 
 higher pitch than in the original vowel. 
 
 I was svirprised at first that I could detect no tendency in a (in aim) 
 or e (in men) to glide upwards towards ee (in see), for the anterior cavity 
 in ee is of smaller size and higher pitch than in the other vowels men- 
 tioned; but I now see that the elevation of the tongue would tend 
 to lower the pitch of the posterior cavity, which should theoretically be 
 raised simultaneousl}' with the elevation of the pitch of the anterior 
 cavity. 
 
 Results obtained by Fleeming Jenkin and J. A. Ewing. 
 
 Jenkin and Ewing have made a minute analysis of the records 
 impressed upon the tinfoil of the phonograph hy vowel sounds, and have 
 published the results of their researches in the columns of Nature, and 
 in the Transactions of the Edinburgh Royal Society for 1878. 
 
 Their analysis of the records of the vowel 6 brought out the fact that, 
 whatever pitch of voice was employed, the predominant partial was b\, 
 as fixed by Helmholtz, or within a few notes of that pitch. 
 
 As the final result of their researches, Jenkin and J. A. Ewing say 
 (see Nature, Vol. XVIII., p. 455): "We are thus brought back to our 
 original statement, that in distinguishing vowels the ear is aided by two 
 factors — one depending on the harmony or group of partials, and the 
 
 other on the absolute pitch of the constituents We are forced 
 
 to the conclusion, already adopted by Helmholtz and Bonders, that the 
 ear recognizes the kind of cavity by which the reinforcement is pro. 
 duced; that, although the sounds which issue differ so much that we fail, 
 when they are graphically represented and mathematically analyzed, 
 to grasp any one prominent common feature, nevetheless, by long 
 practice, the ear is able to distinguish between the different sorts of 
 cavities which are formed in pronouncing given vowels." 
 
 General Rcs7tlts. 
 
 Of the two hypotheses with which we started, it is certain that one 
 (the harmonic) is wrong and the other only partly right. The balance of 
 evidence inclines largely towards the fixed pitch hypothesis, the main 
 argument against it being found in the periodic curves of the phonauto- 
 
 graph. 
 
 The solution of the difficulty seems to be that suggested by Ellis, 
 namely, that, "what we call our vowels are not individuals, scarcely 
 species, but rather genera, existing roughly in the speaker's intention.
 
 129 
 
 but at present mainly artificially constituted by the habits of writing 
 and reading." (See " Early English Pronunciation," Part IV., p. 1279.) 
 
 I do not doubt that the distinguishing characteristic of the vowel- 
 individual, if we could examine it critically, would be found to consist 
 in the presence of partial tones of fixed pitch corresponding to the 
 resonance cavities of some definite fixed position of the vocal organs. 
 A resonance cavity, however, is found to be capable of reinforcing not 
 only a tone corresponding to its fundamental or proper tone, but other 
 tones that differ slightly in pitch from that. 
 
 The reinforcement is greater or less as the exciting tone is more or 
 less removed m pitch from the proper tone of the cavity. Thus^ when a 
 vowel is sung or spoken, those harmonics of the voice which are nearest 
 in pitch to the proper tones of the mouth cavities would be reinforced at 
 the expense of the proper tones themselves. And althongh the ear may 
 be guided in its recognition of vowel quality by a feeling of absolute 
 pitch, the vowels would be recognized from the presence of partials of 
 slightly different pitch— the ear locating, as it were, the distance of the 
 fixed pitch b}- the loudness of the reinforced harmonic. 
 
 In ordinary speech the voice is rarely on a level, but is constantly 
 gliding upwards or downwards. When a vowel is spoken, the pitch 
 of the voice is constantly changing. The reinforced partials must also 
 change in pitch, swelling and dying away in intensity, as they approach 
 or recede from the proper pitches of the mouth cavities. Thus, in the 
 rapid succession of reinforced partial tones, accompanying an inflection 
 of the voice, a point of maximum resonance should be perceived having 
 the absolute pitch characteristic of the vowel uttered.' 
 
 Treating vowels as we find them, as genera of sounds, instead of 
 individuals, the most plausible theory seems to be what we ma}- term 
 "the harmonic fixed pitch theory of vowel sounds," according to which 
 a vowel is a musical compound, of partial tones, whose frequencies are 
 multiples of the fundamental of the voice; the predominant partials 
 being alwajs those that are nearest in pitch to the resonance cavities 
 formed in the mouth by the position of the vocal organs assumed during 
 the utterance of the vowel. 
 
 ' It is well known that the duration of a v wel is an element in determining its quality. It is 
 extremely difTicult to detect the quality of short vowels and they are often spoken of as "obscure."
 
 3IOfVi£DlGAL LfBRARY 
 
 UNIVERSITY OF CAijrnRWiA SAN DIEGO 
 
 DATE DUE 
 
 
 
 
 
 OCT ?j 
 
 4 RECn 
 
 
 
 
 'IP. 
 
 
 
 APR 9.9. 
 
 RfdB 
 
 
 
 
 
 
 
 OCT 1 fi 
 
 FCTJ 
 
 
 
 ■ 'iM ^ ■■' 
 
 
 
 
 DEC 6' 
 
 
 
 
 M2 
 
 4 m 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CAYLORD 
 
 
 
 PHINTEOIN USA
 
 l^A 000 221316 3
 
 i I 
 
 
 ^