This y Le cc he ¥D rt ie aI Fe Set AUP ORG ROR EO a Es ‘i Ys ede ty Bh ae forts ay by Ne Wh Hi ae LAE Faved Fe hy ‘ Bi ‘i Oy iy 44 Vey F BYU i as ‘G Ys {eri fh ce i ae if a ‘ (2 ’ i on { ety ont yh ee iste ANE ai ci i on eh ae i ie f Mie HUsiae Rh Ree ovis Sate ‘ y ‘K f 3. npn a aye aM Bye Ny wf, Ha Wei HK { vi i ve si / i UE i) ig i uD ik i Ae it 4} oie ¥ Ran uy en ¢ iy oe : Oi Wart IP) sie wits i LAA BS at bia) of . Aravel DSN) s be ek i Mu HH NE ONY ats Aa a peargs WT} VW a iy a POOH i & ty <— — ee) vata SS rie es. not Smee. So ar cS Sof En Suess ms, ae os i ot es = Ot Og ee, nd ot, aeons ose ' i at tr ih t} ri ie ie i Ns N19) | ¥ is ‘Et A fat iy Ve nn Ry ah Ht) cele Ot ne on BN! As ia ih it H My te a i mi ol} i sit en Kye nt Nit ay ii is ai ue a vib Hit Ni ye a aS st in ‘y Wa $ aad i } _ a +f pay Pe i af, ‘ SN My vy , NY nay ’ SIN iN i! Madgton ate ‘ Heth a Rss My f Abe N RY Sf Seda te 7 Ria Hy Diath, Wine et Ys it hi “st Vasey! DATO EAY GUE INA TTL Tn vic Ht ANS i PIRATE NAY a) N nn Otek at 5 i a ty { ttt A ya AY an val wi { % a GR ay aR ¥ oe ATO, RO a iyaate WON RAC a any alae i « fay iss ae vd5 Be nas Athy ‘ wil ak ; Ay Hee 4} Meds Peoyeey ae ike) RuARi SUE CO ty ry Mae? oi aii ARAN aS yuan arlaye ah Me ae Naa Asia RN iat na A nue ii ae Ren sie i Sy A wear TS ' Wein Patent bit } ii fe tee 5) apy SPAS As reste f ny M Senet ae a a ae franans PRI LOREO UTTER Die Wa bite VAY Leys Haein ee WIA aN AGS uti) DEO POR Ks Ba ave oe AUIS BOR tN oe OPC HATO Havin) PEE iene ah +) eT as AO Tey Disiny i ( iN} HN Of ai iS a0 AIG i edie yey, nies mbye f iD Dre Men et iiest AN vei ABD Wat noe ANS , ' ae Ne, th yyy ay etn Ppa A Nari eet ynos BAW Det Ate ie) Os MY ays iY Rit! $ CSOT Ay) ah tA DHat at a NY b} WY svete yay mE PAS RLS i f Na Tienes , fies hs i gt AU ARS Sa ae aun ry a o 7 HAS Ne ist hy Nye ry ath | ' BiNye “One etl i 3 ae a Oy need Sie i, Apat WAIN Natt Wh ae a ms Hy Yi a Ai ay) ‘Nt ve rt cat wind * , TAAVAAY 4 Tey ; ve i i. "Tr ) My ’ << . MAY °1 1915 & Reo po1n55 ott Dsvision P22 | {S37 Sex tion e it) eae ae eres al et imate ae Be: hoon oe e's ah ae aaa | a) noah ale 5 eee i a UAE RR te c,h AML AE 7 Pit vy) Varo ia ” . opto, ‘ CLARENDON PRESS SERIES.<<) 4) |. ~ iw, . arf 637 1é \ ae se us »; ‘ A id} a' é FRI aw aestil® Ma A FEST RAP OS : et Y PRIMER OF PHONETICS | By HENRY SWEET, M.A. BALLIOL COLLEGE, OXFORD; HON. PH.D. HEIDELBERG; LL.D, GLASGOW CORRESPONDING MEMBER OF THE MUNICH ACADEMY OF SCIENCES AND OF THE ROYAL PRUSSIAN ACADEMY OF SCIENCES AT BERLIN ORDINARY MEMBER OF THE ROYAL DANISH ACADEMY OF SCIENCES AT COPENHAGEN ; UNIVERSITY READER IN PHONETICS AT OXFORD THIRD EDITION, REVISED OXFORD AT THE CLARENDON PRESS 1906 HENRY FROWDE, M.A. PUBLISHER TO THE UNIVERSITY OF OXFORD LONDON, EDINBURGH, NEW YORK TORONTO AND MELBOURNE PREFACE Tus book is intended to supply the double want of a new edition of my Handbook of Phonetics and of a concise intro- duction to phonetics, with especial reference to English and the four foreign languages most studied in this country— French, German, Latin, and Greek. The Handbook of Phonetics was published in 1877—at a time when the study of phonetics was confined to a few isolated specialists in different countries. Since then phonetics has made no progress in this country—has, indeed, rather gone back *—but has greatly developed on the Continent, where an extensive phonetic literature has sprung up in the last few years, especially in Germany—a literature which becomes more and more indigestible every year. Under these circumstances, aggravated by my own want of access to speakers of foreign languages, I find it impossible at present to comply with the demand for a new edition of the Handbook. Even the compromise I now offer has not been undertaken without hesitation. Knowing the unsatis- factory results of ‘paper phonetics ’—getting up phonetics by reading the statements of others and attempting to har- monize their conflicting views—I have been obliged to exercise great caution in dealing with the latest phonetic literature. It must not therefore be assumed that because I do not adopt a new view, I therefore reject it—it merely means that I have not had an opportunity of testing it by personal hearing. So also by the retention of the main features of Bell’s ‘Visible Speech’ terminology and notation—with the few modifications set forth in my Sound-Notation (Phil. Soc. Trans. 1880—1)—I by no means pledge myself to rigid conservatism. * Witness its total exclusion from the Modern Languages Tripos at Cambridge. A 2 iv PREFACE But I feel convinced that the path of progress lies through the Visible Speech analysis, and that the first duty of the very few who have a practical command of it is to do what they can to spread the knowledge of it. Leaving theory and controversy and the details of less familiar languages to a future new edition of the Handbook, I have tried to make the present Primer as concise, definite, and practical as possible, rigorously excluding all details that are not directly useful to the beginner. The only sound basis of theoretical phonetics is a practical mastery of a limited number of sounds. Some beginners start with an elaborate study of the physiology of the vocal organs —moving, perhaps, heaven and earth to get hold of a real glottis, or going in for the anatomy of the muscles of the tongue—so that by the time they come to real phonetic work they have no energy left. Next to the power of forming sounds correctly and easily, and recognizing them by ear, the most important requisite for the practical phonetician is facility in handling phonetic notation. Those who are inclined to grumble at the supposed difficulty of the ‘Organic’ (revised Visible Speech) notation here employed, may rest assured that such an alphabet can be learnt ten times over in the time it takes to get even a rudimentary knowledge of phonetics. The corresponding ‘Narrow Romic’ notation has been added only for the con- venience of those who are debarred from the use of the Organic symbols. I have, on the other hand, made a more extensive use of the less accurate ‘Broad Romic’ notation, because this really supplies a want—which will some day perhaps be better supplied by a rationally constructed system of shorthand. To make the book more generally useful, I have given some account of varieties of English and German pronuncia- tion. The phonetic texts will afford opportunities of practical training in pronunciation, I will not apologize for the errors and imperfections of which I have no doubt thereby been guilty, except by saying that they are inevitable. Want of space has obliged me to state my views of Latin PREFACE v and Greek pronunciation in a purely dogmatic form. The pronunciations given are those I habitually employ myself. As I consider it quite hopeless to attempt to restore the in- tonation of any dead language, I have simply put stress- marks for the Latin and Greek tones. My greatest difficulty has been with final m in Latin. As I feel convinced that Seelman’s ‘implosive-plosive voiced dorsal reduced » with simultaneous loose lip-closure ’ could not possibly have existed as an independent sound in Latin or any other language, I have been obliged to return to my own theory (Phil. Soe. Proc, 1882-4, xv), although it is not supported by any definite statements of the Roman phoneticians. But it gives a workable pronunciation, in harmony with the development. of the language and the structure of its verse. As regards the objects and utility of phonetics, I will quote the beginning of the Preface to the Handbook (1877): ‘The importance of phonetics as the indispensable foundation of all study of language—whether that study be purely theoretical, or practical as well—is now generally admitted. Without a knowledge of the laws of sound-change, scientific philology — whether comparative or historical—is impossible, and without phonetics their study degenerates into a mere mechanical enumeration of letter-changes. And now that philologists are directing their attention more and more to the study of living dialects and savage languages, many of which have to be written down for the first time, the absolute necessity of a thorough practical as well as a theoretical mastery of phonetics becomes more and more evident. . . . Again, if our present wretched system of studying modern languages is ever to be reformed, it must be on the basis of a preliminary training in general phonetics, which would at the same time lay the foundation of a thorough practical study of the pro- nunciation and elocution of our own language—subjects which are totally ignored in our present scheme of education.’ HENRY SWEET. 1890. PREFACE TO THIRD EDITION In this third edition I have done my best to bring the book up to date as far as possible without wholly recasting it and the system of which it is the exponent. The most important innovation is the extension of the vowels, an account of which will be found in Le Maitre phonétique, December, 1901, under the title of (mikst vauelz). I take the opportunity of advising all who are interested in phonetics or the rational study of Modern Languages to become members of L’ Association phonétique internationale, of which the above-mentioned journal is the organ; it is edited by Dr. Paul Passy (address Monetik, Bourg-la- Reine, France). For diagrams of the organs of speech and their positions the student may be referred to W. Vietor’s Hlemente der Phonelik (Leipzig, 1904) and C. H. Grandgent’s German and English Sounds (Boston, U.S. A., 1892), together with C. L. Merkel’s Physiologie der menschlichen Sprache (Leipzig, 1866). Those who wish to go deeper into the theoretical, acoustic, and philological study of phonetics should proceed at once to E. Sievers’ Grundziige der Phonetik (Leipzig, 1893), and O. Jespersen’s Lehrbuch der Phonetik (Leipzig, 1904). Bos OxFoRD, June, 1906. CONTENTS INTRODUCTION Emancipation from Spelling . ° Differences of Pronunciation Isolation . 2 Analysis . . ° ° : : . ANALYSIS The Organs of Speech . : 5 H F Throat-Sounds Breath, Voice, Whisper . : Other Throat-Sounds a Vowels Tongue Positions Rounding . : Narrow and Wide Nasal Vowels Other Modifications of Voveia! Breathed and Whispered Vowels Tables of Vowels and General Remarks Acoustic Qualities of Vowels Consonants Form : 2 : Place : - Non-oral 7 eee WE ‘ Table of Consonants and General Ealeerts 9 Acoustic Qualities of Consonants Relations of Consonants to Vowels . Sounds formed without Expiration . SYNTHESIS 3 : , Quantity . . . s >. . OSCR AMSTONB) 6 Deus uct ek Es ; Vill Glides Syllable Division Intonation Connexion between Quantity, Force, and Pitch Voice- Quality Organic Basis EncGLisH SounDs Vowels Consonants Synthesis . Specimens FRENCH SOUNDS Vowels Consonants Synthesis . Specimens GERMAN SOUNDS Vowels Consonants Synthesis . Specimens Latin SouNDS Vowels Consonants Specimen . GREEK SouNDs Vowels Consonants Synthesis . Specimen . List or SYMBOLS ; CONTENTS Vowel-Glides Consonant-Glides Glideless Combinations Glide-Consonants 71 102 104 105 107 108 IIo 112 113 114 114 117 INTRODUCTION 1. PHONETICS is the science of speech-sounds. From a practical point of view it is the art of producing speech-sounds and recognizing them by ear. 2. It describes the actions and positions of the vocal organs—throat, tongue, lips, ete.—by which speech- sounds are produced, and classifies sounds according to their organic formation. This is the organic side of phonetics. The acoustic study of speech-sounds classi- fies them according to their likeness to the ear, and explains how the acoustic effect of each sound is the necessary result of its organic formation. 8. We see then that the word ‘sound’ has two mean- ings. When we talk of the sound s we mean (1) the shape of the throat and the position of the tongue by which it is produced, and (2) the hiss which is the result of sending the breath through the passage thus formed. 4. It is indispensable for the practical phonetician to cultivate both the organic and the acoustic sense: he must learn both to recognize each sound by ear, and to recognize the corresponding organic position by the muscular sensations which accompany it. 5. We carry out these processes every day of our lives in ordinary conversation, All therefore that the 2 INTRODUCTION phonetician has to do in the case of familiar sounds is to develop this unconscious organic and acoustic sense into a conscious and analytic sense. In acquiring unfamiliar sounds he has, on the other hand, to begin at the beginning. Before attempting to imitate the mechanism of a foreign sound, he must familiarize himself with its acoustic effect by careful and patient listening, so that the acoustic impression may correct and control his attempts to hit on the exact position for producing the sound. It is evident that the more exact his practical knowledge of the organic positions is, the more easy it will be for him both to find out the right position, and to fix it when once found. Again, the more familiar a sound is, the easier it is to gain insight into its mechanism. Hence the more careful our study of the familiar sounds of our own language, the easier it will be for us to acquire un- familiar ones. EMANCIPATION FROM SPELLING. 6. The first step is, to emancipate ourselves from the influence of our traditional spelling. Every one realizes that rite, write, right, wright all express exactly the same sound, and that ow in bow-window and to make a bow expresses two distinct sounds; but many have a difficulty in realizing that father and farther, savour and save her have exactly the same sound in educated Southern English speech. INTRODUCTION 3 DIFFERENCES OF PRONUNCIATION. 7. Learn not only to recognize and tolerate differ- ences of pronunciation, but to capect them. Remember that pronunciation is incessantly changing, and that differences of pronunciation between the older and the younger generation are not only possible but inevitable. 8. Remember that language exists only in the in- dividual, and that such a phrase as ‘ standard English pronunciation’ expresses only an abstraction. Reflect that it is absurd to set up a standard of how English people ought to speak, before we know how they actually do speak—a knowledge which is still in its infancy, and can only be gained by careful observa- tion of the speech of individuals, the only absolutely reliable observations being those made by a trained individual on himself. 9. Avoid, therefore, all dogmatism and hasty gene- ralizations: be cautious in asserting that ‘everybody speaks in such a way,’ or that ‘no educated man pro- nounces so.’ Do not appeal to the authority of an imaginary ‘correct’ or ‘careful’ speaker. 10. Confine yourself to plain statements of facts. If people tell you that spelling reform is ‘a pestilent heresy,’ or that your London, Edinburgh, or Dublin pronunciation is ‘abominable, do not argue with them. ISOLATION. 11. Learn to pronounce every sound apart from its context. If you want to find out your natural pro- 4 INTRODUCTION nunciation of any sound, utter itin some familiar sen- tence, several times over; then try to pronounce it separately without altering it in the slightest degree. Then test your command of it by pronouncing it in unfamiliar combinations, transposing for instance king into zyik. So also pronounce hear in hear them exactly as in hear 2t—that is, with a consonantal 7 instead of a mere voice-murmur. If it is a vowel, learn to lengthen or shorten, and to emphasize it without modifying it in any way. Thus, learn to lengthen the vowel in di¢ without making Jit into beat. Again, lengthen both vowels in pity, and observe the difference (if any) between them. Isolate and lengthen the first elements of the diphthongs in high and ow, and observe the difference (if any) between them. ANALYSIS. 12. The next stage is, to learn to analyse the forma- tion of these familiar sounds. This analysis must be practical. Theoretical knowledge is not enough. It is, for instance, no use being able to explain the differ- ence between a breath consonant such as /, and the corresponding voice consonant v, unless we are able to feet the difference. Again, it is not enough to hear the difference: we must have a sense of the difference of articulation. Let the beginner learn to isolate and lengthen such a breath consonant as / in ife—fff, and the corresponding voice consonant v in fiver—vev. He will soon find that while fis articu- INTRODUCTION F lated in one place only—being the result of the friction of the outgoing air between the lower lip and upper teeth—v is articulated in two places—between lip and teeth, and in the throat. If he presses his first two fingers on the larynx or‘ Adam’s apple,’ he will feel the vibration which produces the effect of voice in v, but not inf. If in pronouncing v he removes the lip from the teeth, he will hear the unmodified voice-murmur. If he does the same with he will hear the unmodified breath in the form of a scarcely audible sigh. He can then reverse these processes. If he breathes in the ordinary way, and brings the lower lip and the upper teeth together, he will produce an /; if he makes a voice-murmur, as in err, and brings lip and teeth together, he will produce a ». Let him do the same with other familiar pairs, such as s and z, hard and soft ¢/ in thin and then, first separately, and then in one breath—sszz, zzss—till he feels the distinction, and has it thoroughly under command. 13. The great test of this command is the power of forming unfamiliar voiced sounds from familiar breath sounds, and vice versa. Repeat v, / several times in succession, and then try to carry out a similar change with 7; the result will be a sound which is essentially the Welsh “ in Llangollen. Carry out the same pro- cess with 7, 7, m, and ng in king. 14. Having thus mastered the fundamental dis- tinction of breath and voice, the beginner must learn to feel the movements of the tongue and lips. If he 6 INTRODUCTION passes from ce in see to ah, or aw in saw, he will feel that the tongue is moved backwards and downwards ; if he reverses the process, he will find that the move- ment is upwards and forwards. Pronouncing ee by itself, he will then find that in its formation the fore part of the tongue is brought very close to the fore part of the palate—or, in phonetic terminology, that ee is a‘ high-front’ vowel. In the same way he will find that aw is a ‘low-back’ vowel, the root of the tongue being pressed down and brought near the back of the mouth, a4 being a ‘ mid-back’ vowel. Again, in comparing aa with oo in too, he will find that while in aa the mouth is open, in oo the lips are brought together so as almost to close it—that 00 is in phone- tic language a‘ rounded ’ vowel—a ‘ high-back-round ’ vowel. In time he will learn to measure more accur- ately the degrees of raising, lowering, retracting and advancing the tongue, and narrowing the lip-opening by which the different sounds are formed. Thus he will find that ce and a@ in name and man are all three front vowels, but that in the second vowel the tongue is lowered from the ‘high’ ¢e-position to the ‘mid, while in the a of man it is further lowered to the ‘low’ position. So also he will find that the lip- opening of o in wo is greater than in 00, while that of aw is greater than in 0, the lip-opening being only slightly narrowed in aw, ANALYSIS 15. The foundation of speech is breath expelled by the lungs and variously modified in the throat and mouth. 16. Speech-sounds are generally formed with ozt- breathing or expiration (>), rarely with in-breathing or inspiration (<). The sounds known as clicks (go) are formed by the air in the mouth without either out- or in-breathing. The Organs of Speech. 17. The breath passes from the lungs through the bronchial tubes into the trachea or windpipe, and thence into the larynx. Across the interior of the larynx are stretched two elastic ligaments, the ‘ vocal chords.’ They are firmly inserted in the front of the larynx at one end, while at the other they are fixed to two movable cartilaginous bodies, the ‘ arytenoids,’ so that the space between them, the ‘ glottis,’ can be narrowed or closed at pleasure. The glottis is, as we see, twofold, consisting of the chord glottis, or glottis proper, and the cartilage glottis. The two glottises can be narrowed or closed independently. The chords can also be lengthened or shortened, tightened or relaxed in various degrees by means of the muscles they contain. 18. Above the true glottis, and still forming part of 8 ANALYSIS the larynx, comes the ‘upper’ or ‘ false’ glottis, by which the passage can be narrowed or partially closed. On the top of the larynx is fixed a sort of valve, the ‘ epiglottis,’ which in swallowing and in the formation of certain sounds is pressed down so as to cover the opening of the larynx. 19. The cavity between the larynx and the mouth is called the ‘pharynx.’ It can be expanded and con- tracted in various ways. 20. The roof of the mouth consists of two parts, the ‘soft’ and. the ‘hard palate.’ The boundary between them may easily be found by pressing the forefinger on the palate and sliding it back till the palate yields to the pressure. The lower pendulous extremity of the soft palate, the ‘uvula,’ can be pressed backwards or forwards. It is pressed back so as to close the passage into the nose in the formation of all non-nasal sounds, such as a, d. When the pressure is relaxed, the breath flows through the nose, as in ordinary breathing and in the formation of nasal sounds, such as 2, or French ain sang. 21. The other extremity of the palate is bounded by the teeth, of which we must distinguish the ‘edges’ and the ‘rim’—the place where they join the gums. The gums extend from the teeth-rim to the ‘ arch-rim,’ behind which comes the ‘arch,’ whose front wall is formed by the ‘ teeth-roots’ or alveolars +. The middle part of the palate from the arch-rim to the beginning of the soft palate is called ‘front.’ The soft palate 1 Strictly speaking the alveolars are the sockets of the teeth. THROAT-SOUNDS 9 and the wall of the pharynx behind it constitute the ‘back’ of the mouth. 22. Of the tongue we distinguish the ‘ back’ or root, the middle or ‘ front,’ and the tip or ‘ point,’ together with the ‘rim’ or edge of the tongue on both sides of the tip, and the ‘blade,’ which includes the upper surface of the tongue immediately behind the point. ‘Lower blade’ implies, of course, the lower instead of the upper surface of the tongue. Front, blade, and point are included under the common term ‘ fore.’ 23. Besides the main positions indicated by these names, an indefinite number of intermediate ones 1s possible. The chief varieties are designated by the terms ‘inner’ and ‘outer, inner implying nearer the back of the mouth, outer nearer the teeth. Thus the ‘outer front’ of the tongue is a place nearer the point than simple front, and is therefore an approximation to the blade position. 24. Sounds are also modified by the movements of the lips. Throat-Sounds. BREATH, Voice, WHISPER. 25. When the glottis is wide open, no sound is pro- duced by the outgoing breath, except that caused by the friction of the air in the throat, mouth, ete. This passive state of the glottis is called ‘breath,’ and is symbolized by o, pictorial of the open glottis, whence is formed the ‘breath-modifier’ : (§ 48). The most important active states of the glottis are those which produce ‘ voice ’ and ‘ whisper.’ SWEET B Io ANALYSIS a6. Voice is produced by the action of the breath on the vocal chords in two ways: (a) If the glottis is entirely closed by the chords so that the air can only pass through in a series of extremely rapid puffs, we have that most sonorous form of voice known as the ‘chest’ voice or ‘ thick register’ of the voice. (4) If the chords are only brought close together, without com- plete closure of the glottis, we get that thinner quality of voice known as the ‘head’ voice or ‘ thin register, which in its thinnest and shrillest form is called ‘falsetto.’ The symbol of voice is 1, pictorial of the glottal chink. The ‘ voice-modifier ’ symbol is}, 27. If the glottis is narrowed without vibration, ‘whisper’ is produced, which is symbolized by 0, as being intermediate between breath and voice. The ‘whisper-modifier’ pis a curtailed 9. There are two degrees of whisper, the ‘ weak’ and the ‘strong.’ (For the ‘wheezing’ whisper see § 30.) In the weak whis- per there is slight narrowing of the whole glottis: in the strong, which is the ordinary form, the chord glottis is entirely closed, so that the breath passes only through the cartilage glottis. 28. The popular and the phonetic use of the term whisper do not quite agree. Whisper in popular lan- guage simply means speech without voice. Phonetically speaking whisper implies not merely absence of voice, but a definite contraction of the glottis. In whispering as opposed to speaking aloud what happens is this : Breathed sounds, being already without voice, remain unchanged. Voiced sounds substitute whisper (in the THROAT-SOUNDS II phonetic sense) for voice. If we pronounce ‘ vee’ and ‘fee’ first aloud, and then in a whisper, we shall find that in ‘vee’ both consonant and vowel are altered, while in ‘ fee’ only the vowel is altered, the consonant remaining breathed as in loud speech. It must, there- fore, be understood in phonetic discussions, that when- ever we talk ofa whispered sound, we mean one uttered with a definite contraction of the glottis. Whether we talk of a‘ whispered 7’ or a‘ whispered v’ is indifferent: both names signify the ‘ lip-teeth whisper ’ consonant, the latter implying, however, a substitution of whisper for voice. The acoustic distinction between breath and whis- per is not very marked, but if we compare breathed and whispered 4, we perceive clearly that the latter is, like the voiced v, a composite sound, being formed partly in the throat. Whispered sounds are also feebler than breathed ones, the force of the outgoing air being diminished by the glottis contraction. OTHER THROAT Sounps, : 29. Glottal Stop (x). When the glottis is suddenly opened or closed on a passage of breath or voice, a per- cussive effect is produced, analogous to that of & or any other ‘stopped’ consonant. The most familiar example of this ‘ glottal stop’ is an ordinary cough. The student should carefully practise the glottal stop in combina- tion with vowels till he is able to produce xJ+ and Jax as easily as (kaa) and (aak), taking care not to let any B2 12 ANALYSIS breath escape after the x in x}, as is the case in coughing. He should then learn to shut and open the glottis silently, and to know by the muscular sensation alone whether it is open or shut. It is easy to test the closure of the glottis by pressing the fore-finger on the throat above the larynx and tapping on its nail: when the glottis is open, this tapping produces a dull sound; when it is shut, the tapping makes a clear and hollow sound, like the gurgling of water poured into a bottle, and the pitch of this sound can be raised or lowered at pleasure by retracting or advancing the tongue. x forms an essential element of some languages, such as Danish, where, for instance, und etx3 ‘dog’ is distinguished from hun 2t3 ‘she’ solely by the ‘stddtone’ or x. 30. Wheeze. If we strongly exaggerate an ordinary whisper, we get that hoarse, wheezy sound known as the ‘ wheezing’ or ‘stage whisper’ (0s). In the forma- tion of this sound there is not only the glottis narrowing of the ordinary strong whisper, but there is also contraction of the upper glottis, and the opening may be further narrowed by depression of the epiglottis. 31. Deep Throat-Sounds. If we narrow the passage below the glottis—apparently in the bronchial region ’, we get a clear hiss not unlike that of a goose, and closely resembling the back-open-breath consonant c in Scotch och. This sound (o-), the deep-open-breath consonant, is one of the ‘ Arabic gutturals ’—that one which is transcribed /, as in o-]+ 4a (the name of the 1 See Le Maitre Phonétique, 1904, p. 36. VOWELS 13 letter) sJororfo wahhid ‘proclaim the unity of God!’ osltor 77h § wind.’ Being formed independently of the glottis, this sound can be voiced and whispered. When voiced, it makes the vocal chords vibrate slowly and intermit- tently, giving that gruff, creaky quality of voice with which many speakers in all countries—especially Ger- many—exaggerate the effect of a deep bass voice. If we soften the creak by diminishing the throat-con- traction, we get the 7-like sound of the weak deep- open-voice consonant IV, which is the other Arabic guttural in 1-\[t3 ‘é~ ‘eye’ (also the name of the letter), p):1-vf[+0 ba‘td < distant.’ Vowels. 82. A vowel may be defined as voice (voiced breath) modified by some definite configuration of the super- glottal passages, but without audible friction (which would make it into a consonant, § 64). The symbols of the different vowels are all formed by modifying the voice-symbol 1. TONGUE PosITIons. As each new position of the tongue produces a new vowel, and as the positions are infinite, it follows that the number of possible vowel-sounds is infinite. It becomes necessary, therefore, to select certain definite positions as fixed points whence to measure the inter- mediate positions. 33. The movements of the tongue may be distin- 14 ANALYSIS guished generally as Horizontal and vertical—backwards and forwards, upwards and downwards. The horizontal movements produce two well-marked classes,(1) ‘back’ (guttural) vowels, formed by the root, and (2) ‘front’ (palatal) vowels, formed by the fore part of the tongue. In the formation of back vowels, such as J in father, } in fall, the back or root of the tongue is brought into prominence partly by retraction of the whole body of the tongue, partly by pressing down the fore part of the tongue. In the formation of front vowels, such as f[ in z¢ and [{ in man, the front of the tongue is raised towards the front of the palate, so that the main body of the tongue slopes down from the front of the mouth backwards. There is a third class of ‘mixed’ (gutturo-palatal) vowels such as the j in err where the whole tongue is allowed to sink into its neutral flattened shape, in which neither back nor front articulation predominates. 34. The vertical movements of the tongue, which are generally accompanied by lowering and raising of the jaw, produce various degrees of ‘height,’ or distance of the tongue from the palate. Thus in f the front of the tongue is raised as high and as close to the palate as possible without causing audible friction, or buzz. In {, as in men, it is somewhat lowered, and in [ it is lowered as much as possible. From among the infinite degrees of height three are selected : (1) ‘high,’ as in Lf, (2) ‘mid,’ as in [, (3) ‘low,’ asin [. These distinctions apply equally to back and mixed vowels, so we have altogether nine cardinal vowel-positions : VOWELS “1s 1 high-back I high-mixed [ high-front J mid-back 1] mid-mixed [ mid-front J low-back ylow-mixed [{ low-front. It will be observed that we place the back vowels on the left side of the table, because the direction of the stream of breath by which sounds are formed is supposed to move in the same direction as in our ordinary writing, viz. from left to right, so that ‘ left’ corresponds to ‘back’ in all phonetic diagrams and symbolization. Hence the ‘ vowel-modifier’ is turned to the left in the symbols of back vowels (J), and to the right in those of front vowels (I), mixed vowels combining the back and front modifiers (I). 35. In passing from [ to [ and { there is not only lowering of the tongue, but the point of greatest nar- rowing is also shifted back progressively, the size of the resonance-chamber (60) in the front of the mouth being thus increased in both directions. 36. Intermediate positions between the nine cardinal ones are marked by diacritics: + ‘ raised,’ ~ ‘lowered,’ 4‘inner, + ‘outer.’ Thus f- and [+ both denote a vowel intermediate between the 7 of 7# and the e of men, viz. the second vowel in pity plol-. 87. Besides the nine cardinal positions there are nine other ‘shifted’ positions. We have seen that normal mixed vowels are characterized not only by ‘mixed’ position—that is, intermediate between back and front—but also by flatness of the tongue. These shifted positions are obtained by combining flatness of the tongue with back position, and the slopes of 16 ANALYSIS back and front vowels respectively with mixed posi- tion, giving the three series : Tai in-mixed li out-back {44 in-front 14 jr [44 qi yt TH The in-mixed or ‘ back-flat ’ vowels are obtained by retracting the positions of the mixed vowels into the corresponding back positions. Thus if we form J, the low-mixed-narrow vowel in English err, si, and then retract the tongue into the full back position, we get the low-in-mixed-narrow J+, which may be heard in the Irish pronunciation of sir. J, is, therefore, simply an exaggeration of J1, inner-low-mixed. So also j+ and [+1 are exaggerations respectively of y+ outer-low- back and { inner-low-front. The vowels of the last two series are mixed in position, but retain the slope of the corresponding back and front vowels. It is not easy to distinguish between inner and in-front, but the in-mixed and the out-back vowels have a distinct acoustic character of their own by which they are sharply marked off from the corresponding back and mixed vowels. ROUNDING. 88. Rounding (labialization) is a narrowing of the mouth-opening by approximation of the lips. It can, of course, be combined with all the tongue-positions described above. It is symbolized by a line drawn across the vowel stem—f = French u in /ume. There are three principal degrees of lip-narrowing, correspond- VOWELS 17 ing to the height of the tongue, high vowels having the narrowest, low the widest lip aperture. This is easily seen by comparing the vowels of such a series as the high-back-round t in good, the mid-back-round $ in no and the low-back-round j in not. It will be seen that in t the lips are contracted to a narrow chink, while in } the opening is wider and broader, and in ¥ only the corners of the mouth are con- tracted. 89. There are two kinds of rounding, inner and outer. In outer rounding—with which front vowels are rounded—the lips are brought together vertically, If the lips are separated by a finger and thumb up- wards and downwards, it will be found impossible to form a front-round vowel such as £—the result will be {—the French «7. Back and mixed vowels, on the other hand, are rounded by lateral compression of the corners of the mouth and, apparently, of the cheeks as well. If a finger and thumb are put in the corners of the mouth so as to bear upon the cheeks about an inch inwards, and then expanded sideways, it will be found impossible to pronounce a back-round yowel such as }—the result will be a muffled form of the ] in father. If, on the other hand, the lips are spread upwards and downwards during the utterance of such a vowel as }, it will still retain much of its distinctive rounded character. The distinction between inner and outer rounding is taken for granted in the ordinary vowel-symbols, as in f,t. Where necessary they are expressed by the modifiers » for outer, > for inner 18 ANALYSIS rounding. Thus }) is a muffled form of a in father, distinct from =p. 40. The effect of rounding may be increased by projecting or ‘pouting’ the lips so as to form an additional resonance-chamber beyond the teeth. This action is generally avoided in English, but may be observed in Scotch, and generally in continental pro- nunciation, as in French and German. It is symbolized by adding the ‘ protruder ’ (2), thus pt»a=Scotch book. 41. The influence of the lips may also be observed in the unrounded vowels. In the formation of the high-front vowel f in it, the sound is made clearer by spreading out the corners of the lips. So also to a less degree with the mid- and low-front vowels. But in a back vowel such as the ] in father the lips tend to the neutral position of rest, although these vowels may also be made clearer by lip-spreading, which is symbolized by the addition of the ‘ spreader’ 1, as in fi = French « In English the lips are less spread than in many other languages, such as French and German. 42. Although there is a natural connexion between the heightof the tongue and the degree of lip-narrowing —tfor it would be a waste of sound to narrow the back of the mouth and then allow the sound to diffuse itself in the fore part of the mouth ; or to widen the back part of the voice-channel, and then mufile the sound by over- narrowing of the mouth-passage—there are many cases of abnormal degrees of rounding in language. We must distinguish between ‘under-rounding’ and ‘over-round- VOWELS 19 ing.’ Under-rounding implies a /ess degree of rounding than is normally associated with the vowel’s height, as when a high vowel is formed with the rounding of a mid or low vowel ; itis symbolized by adding the inner rounder > to the symbol of a back, the outer rounder ) to the symbol of a front vowel. Thus 19 is a combination of the tongue position of the t in good with the lip- position of the } in xo—a sound which seems to occur in the dialects of the North of England. Over-round- ing implies a greater degree of rounding than normally belongs to the vowel’s height; it is symbolized by adding the rounder to the symbol of the normal rounded vowel. Thus f), as in German déer, is the mid-front vowel with the rounding of the high front £; it is a compromise between the French vowels f, as in pew, and f, as in pur. Degrees of abnormal rounding may be discriminated by the addition of the raiser and lowerer (36); thus 1+ implying mid, and 1- low rounding. NARROW AND WIDE. 43. These are important general modifications of all vowels, The narrow vowels are symbolized by a‘ dot- definer’ ({), the wide by a ‘ hook-definer’ (f). The ‘narrow-modifier’ and the ‘wide-modifier’ ¥ are formed from \andy(70). The distinction depends mainly on the shape of the tongue. In forming narrow vowels there is a feeling of tenseness in that part of the tongue where the sound is formed, the surface of the tongue being made more convex than in its natural ‘ wide’ shape, in 20 ANALYSIS which it is relaxed and flattened. This convexity of the tongue naturally narrows the passage—whence the name. The narrowing is the result not of raising the whole body of the tongue (with the help of the jaws), but of ‘bunching up’ lengthways that part of it with which the sound is formed. Hence if we take a low-wide vowel, such as the low front-wide J] in man, we can raise it through the f in men to the high position of the f in 2 without its ever running into the mid-front-narrow[in French éé. So also {[ may be raised till it becomes a consonantal buzz without ever passing into the high- front-narrow [ in French s:—that is, as long as the tongue retains the laxity and comparative flatness of a wide vowel. The distinction between narrow and wide is not so clear in the back vowels, where the convexity of the tongue seems to be accompanied by tension of the uvula and soft palate. 44, It is, of course, possible for a vowel to be exactly half-way between narrow and wide, which is symbol- ized by adding the wide modifier to the narrow vowel symbol, thus fy = half wide z, as in Norwegian /isé. Nasa VOWELS. 45. In the formation of nasal vowels voiced breath flows through the nose as well as the mouth (20). If the nose passage is kept only slightly open, we get the ‘nasal twang’ of many American and some English speakers, which modifies all the vowels impartially, although, of course, it is more audible in the opener VOWELS a1 vowels than in the close high ones. In languages which distinguish between nasal and un-nasal vowels the nose-opening of the former is necessarily more marked than in a mere nasal twang. It is especially marked in French, where the nasal vowels in such words aS SjS sang, >[S vin are very sonorous, The nasality-modifier is supposed to be pictorial of the pendent uvula. OTHER MODIFICATIONS OF VOWELS. 46. Vowels may be uttered simultaneously with several of the consonants. Thus, if we put the tongue in the /-position (69 4), we shall find little difficulty in articulating almost any vowel, although, of course, the back vowels are the easiest. Most of these consonant- modified vowels are of little practical importance. But there are classes of point-modified vowels which deserve notice. They are symbolized by the addition of the ‘point-modifier’\. The English point-consonant © in starry is in itself almost a vowel, and if we carry the raised tongue-point position by which it is formed into the preceding vowel, we make sojtol+ into sojiuoly. | If the point is turned back or ‘inverted’ (71 ¢), we get jt\c with a peculiar ‘snarling’ effect, common to most of the dialects of the south-west of England, and heard in such words as dard, sir. 47. Another well-marked class of consonant-modi- fied vowels are the duzzed vowels, formed with the point of the tongue in the outer point position of E. Vv (p) or in the blade position of s (s): Iv or [5, etc. 22 ANALYSIS BREATHED AND WHISPERED. VOWELS. 48. If an open vowel such as ] is uttered with gentle breath instead of voice, we get a scarcely audible sigh, in which, however, the characteristic effect of the vowel is still audible. If we take a high vowel such as f in dit, the friction of the breath is clearly audible, and still more so in the narrow [. So also with t and 1, which when devocalized sound like a weak wh (72). In French this devocalization of high vowels is frequent, as in [S{s ainsi. Voiceless vowels are indicated by the addition of the breath-modifier s. 49. Whispered vowels, which are indicated by the addition of the whisper-modifier ), are common in some languages, such as Portuguese, where Oyorto is pro- nounced t'p'}wso'b. They may sometimes be heard in unstressed syllables in English, as in opafw\, together. TABLES OF VOWELS AND GENERAL REMARKS. 50. The seventy-two elementary vowels are given in the annexed tables, with key-words from English (E.), Scotch-English (Se.), American-English (Am.), French (Fr.), German (G.), and other languages. 51. In naming the vowels ‘height’ comes first, and ‘rounding’ last— high-back-narrow-round,’ ete. 52. In the ‘narrow Romie’ notation wide vowels are expressed by italics, mixed by dots. 53. A practical mastery of the vowels can only be acquired very gradually, and by beginning with those that are already familiar (5). Foreign sounds are best VOWELS 23 learnt in connexion with a general mastery of the lan- guage in which they occur. When the learner finds that intelligibility depends on what seem to him almost inappreciable distinctions, he is forced to learn to re- cognize andimitatethem. A practical mastery ofsounds implies the power of uttering them accurately and without hesitation or effort—above all, without any facial contortions—till at last the correct pronunciation becomes a matter of pure habit, and is preserved even in the most rapid speech. Few learners can really master half-a-dozen new vowels in less than six months, so that it is quite worth while making a general study of the grammar and vocabulary of the language in which they occur. 54, Whispering the vowelsis a great help in analysing their formation. After a time the student will be able to recognize each vowel solely by the muscular sensations associated with its formation : he will be able to say to himself, ‘now my tongue is in the position for {, now I have changed it into the I-position, ete., while not uttering the slightest sound, confident that if whispered or voiced breath is allowed to pass through the mouth, the required sound will be produced. 55. The tongue-positions may be compared in various ways. It is very instructive to run through a whole series vertically or (what is more difficult) horizontally, shifting the tongue with uniform speed from one extreme to the other, thus: f, f+, [, [-, [, [ and back again ; L J; JJ, Js and back again. Itis also important to acquire the power of rounding and unrounding at will. ANALYSIS 2.4 LS 10303 “9 (T *- *I-"M-FUOTJ-MOT gt ‘ed “I-"M-YWOIJ-pTur ps ‘poms (f ‘ddn "pemg °c f 40m °C f ‘Inad ‘1g “oO t c f ‘ND *C f “I-"M-POXTUL-MOT *I-" M-OBQ-MOT *I-'U-JUOTJ-MOT *I-"U-POXTUI-MOT *I-'W-YOBVG-MOT of be gi zl 9 *[OS “poemMg cf 0S") ‘0 f{ *I-"U-yOeq-prur eqn “1D g ‘ned "iq *e } *I-"U-JUOLJ-prur “(eft on ‘£oq ‘0 f *I-"M-OVQ-prur 2} “I-"U-poxTul-prur ‘xajoq ug ‘9 } *I-"MA-pOXTUL-pruLr sé 6z fz Lt It S$ "EFO 004 "goog og 4] ft F cae 8 ‘qnd nF ‘ind ‘aq °£ F| ‘snq‘pems 7 z qn3 "Hp on { punoxr punoz punoxr punoz punox *punox ~A-JUOT-YSty | ~"A-poxTU-YStTg | ~"a-Woeq-qoTy ~U-FLOrZ-G ITY ~"U~poxTur-YSTy ~"U-youq-YS1y be gz wz gl OI v : ‘MUIDD “41OT “ywnd Lewyo0N ‘upur -w 1 ‘moq ‘po T |'your “AION “a f ‘aye ) “8 °Y T ‘a f "M-7WO1J-MOT *AA-POXTUI-MOT *AA-YOBG-MOT *U-} WO1J-MOT *U-poxrul-MOT "U-ORG-MOT €€ Li Iz Si 6 € a7]s fips ‘es “UB +] ‘IL Yuva “ULy “200 ‘uaa ‘a |} wayeq 2, ‘zoqny *p [|| Avs ‘og 20s “H | ‘aqes "QD 2 a ‘nq “8 [ “M-FWO1J-prud *M-poxtud-prur "M-YOBq-prur *U-} WOIF-PrUad “U-poxrul-prur *U-yoVq-prur ze gz Oz vr 8 z *arpay ‘Aqid at ‘oys 90s ‘90q 2 J oprm-juory-q sty 1€ Gz 61 €1 ‘nufs ‘ssuyy tT 4 tt a) T “18 ‘IA a J "aun stom. “ia I] "Vv \é OPIM-poxtur-ysty | oprd-yoeq-ySrq || MorvU-yWOIF-GSTY |MOITVU-poxtUs-qsry MOLIBU-YOVqG-Y Sy L I 25 VOWELS a rt "10Q040Q ‘gC aif ort || «193 ‘pemg ‘eo tet "Zc aif Zo HE ‘I-"ah-"F-UT-MOT | “-"AA-310'BQ-9N0-MOT TV MNUI-UP-MOT || “E-"U-FUOTJ-UI-MOT | “I-"U-Yo’q-7N0-Mo] ‘I-"U-pOXIUI-UT-MO] zl 99 09 $s gb zh “Fe ted "MOTTOF *z0 uf *bQ re} ‘pins og ve re} "0 aif %Q re} “-"sA-"F-UI-pTur “I-"AM-YOVG-9N0-prur ‘I-"M-"U-UT-prug || ‘*I-"U-7Wory-uI-prur ‘I-"U-YOVG-9NO-prug | 'I-"W-pox) w-ur-prur 1 S9 6$ €S Ly Iv A bey ‘anTBa “mn 44 ETET “FA bey | smy ‘MIONN *,1 WT Hey punor-oprm punor-oprm punoir-opra punol-Mo1reu punol-Mo1reU punor-MorIeU ~yUorj-Ul-Ysry “yowq-yno-ystq =| -pexrur-ur-q3rq ~yUOrj-Ul-YSty “yovq-jno-qsry =| -pextw-ur-qstq of to gs zs ob ov ‘a te] | -ouour ‘ueg a +f Sp te) ‘IPT “pemg “Sez +r) a af ‘say fy “M-FUOTJ-UT-MOT | *A-YOR-9N0-MOT “M-PORTUU-UI-MOT "U-J WOIJ-UI-MOT "U-YOwq-9N0-MOT ‘U-PeXxTUL-UI-MOT 69 £9 . Lg 1S or 6€ ‘att | oka Sdn ‘yp +f bg rr], vq tr] 8 +f Bg rt] ‘A-PUOTF-UI-prur | “A\-YOVq-3n0-prur ‘M-POXTUL-Ul-prur}| *U-4001J-UT-prur ‘U-HOCG-yno-prur | *a-pexrut-ul-pruz 89 zQ 9S os tr gt "Ayjaad *% HT "ge AT Th HET ‘tr HY ‘SVT | Op joey -% rT Opt OptTa pra AOLIBU AMOLIVU MOLE “HMors-UEgsry | -yowq-yno-ySrq | ~pexyor-wi-ySry ~yuorj-UI-Y Sty “youq-no-ystq =| -pexrur-ur-y sry lo 19 ss 6b ee Lg & co iQ = m= 26 | ANALYSIS It is very easy to learn to round the front vowels on the pattern of the already familiar back-round vowels, thus J :}$::f:4, the substitution of outer for inner rounding being made instinctively. Unrounding should be practised first on the front-round vowels, then on familiar pairs of back vowels, such as }:]. The lips should at first be spread by the finger and thumb (39). The most difficult vowels to unround are the high-back t and f. 56. The use of these different processes in acquiring unfamiliar sounds is self-evident. They may be used to check one another. Thus, if the student has learnt to form j from J by lowering the tongue, he can also unround the f in xot, andif both operations are correctly performed, they will yield the same, or very nearly the same sound. 57. We say ‘ very nearly the same sound,’ because it is only occasionally that a key-word gives exactly the sound indicated by an unmodified symbol such as J or f. Not only are key-words ambiguous through the various divergencies of pronunciation classed as provincial- isms, vulgarisms, individual defects, and the inevitable change from generation to generation, etc., but also because every language tends to shift its sounds a little, often in order to make them more distinct—to avoid confusion with some other sound. Even in the same language pairs of rounded and unrounded vowels are often formed with slightly divergent tongue-positions, as we see exaggerated in G. i and a (296). The student must therefore learn to form the seventy-two vowels VOWELS 27 independently of the associations of their pronunciation in special languages. To him [+ is that vowel which is the result of bringing the convex front of the tongue as near the palate as possible without producing conso- nantal friction, and [ is the result of a position exactly half-way between [+ and {-, the latter being the result of lowering the tongue as much as possible. In practice such symbols as f and [ stand fora group of infinitesi- mally different sounds, to be differentiated, if necessary and practicable, by diacritics—I., [1, [4+, ete. ACOUSTIC QUALITIES oF VOWELS. 58. We have hitherto ignored the acoustic effects of the vowels. This has been done designedly. The first and indispensable qualification of the phonetician is a thorough practical knowledge of the formation of the vowels. Those who try to learn new sounds by ear alone, without any systematic training in the use of their vocal organs, generally succeed but partially. 59. The test of ‘a good ear’ is the power of dis- criminating and recognizing sounds. This is an indis- pensable qualification for those who wish to write down sounds from hearing, and should be carefully cultivated by all students of phonetics. In learning foreign sounds the habit of patient listening should be cultivated before everything. No attempt at imitation should be made till the acoustic impression has been fixed. Otherwise the student hears his own imperfect imitation as much as the correct sounds. It is important to hear the same sound pronounced by different voices. Hence the ad- 02 28 ANALYSIS vantage of learning sounds in the country where the language they occur in is spoken. In his daily practice of portions of the vowel-table, the student should sometimes whisper the vowels, sometimes utter them in a loud voice, and compare the sounds of those that are most like, till he can distinguish them. 60. It will soon be observed that vowels whose formation is distinct are often very similar in sound. This will be better understood if we consider that a vowel is, acoustically speaking, voice modified by are- sonance-chamber, viz. the mouth. Every time we move the tongue and lips we create a new resonance-chamber which moulds the voice into a different vowel. Every vowel can have its pitch raised or lowered by varying the vibration of the vocal chords, as when a scale or a chord is sung on one vowel. But each vowel has besides an inherent pitch of its own, due to the shape and size of its resonance-chamber. Thus, if [,], andt are all sung to the same note, we hear how much deeper tis than ], while J is much deeper than [. The best way of hearing the inherent pitches of the vowels is to whisper them, for the pitch of whisper being invariable, the differences caused by the resonance are clearly heard. The connexion between the shape and size of the resonance-chamber and the pitch is self-evident. Thus £ owes its high pitch to its being formed by a narrow channel in the front of the mouth, while the pitch of J is lowered by the greater size of its resonance- chamber, and that of t by the narrowing of the lip- aperture, both J and t being formed in the back of the VOWELS 29 mouth, which, of course, gives them a deeper and more hollow sound independently of other influences. The wide forms of the front and mixed vowels are lower in pitch than the narrow ones, because of the greater width of the mouth cavity in the case of the wide vowels: but in the case of the back vowels the narrows have the lower pitch, because they retract the tongue more. 61. The exact pitches of the vowels have not yet been determined ; we have not even got so far as to arrange them in the order of their pitches. The following is Mr. Bell’s order of the pitches of some of the chief vowels, beginning with the lowest, vowels having the same pitch being bracketed together: teytear da Tee trees The pitches of the back vowels offer the greatest diffi- culties, especially in the back-rounded vowels. Thus an increase in the force of the whisper will raise the pitch of t from one to two tones, by throwing the sound forward to the lips. But if we whisper the last six front vowels, we can hear that they rise uniformly by a semitone, making a chromatic scale. 62. It is evident from the table that the same pitch may be produced by different modifications of the same resonance-chamber. ‘Thus, if we start from [, we can lower the pitch either by retraction of the tongue, giving I, or by rounding, which gives £, and con- sequently I and f have the same pitch, the retraction of one vowel being equivalent to the rounding of 30 ANALYSIS the other, and they are so like in sound that those who hear I for the first time often take it for a round vowel. Hence also the tendency to confuse English J in err with an open sound of German ¢ or French ew. 63. It is also important to observe that such pairs as f and [, t and }, are as near in sound as fand [,t and ft, the order of pitches being (, f, [ and t, t,}. The ex- planation is precisely analogous to that of the similarity of I and f, namely that the pitch of [is deepened by increasing the distance between tongue and palate, and so enlarging the resonance-chamber, and that this can be effected either by flattening the surface of the tongue, as in f, or by lowering the whole body of the tongue, as in{. Land [are very similar in sound, and some foreign phoneticians still ignore the difference, denoting f and [ etc., by the same sign. Raised [+ has the same pitch as f, from which it can hardly be distinguished when short. We have then the following pairs of vowels very like in sound : 64. Consonants are the result of audible friction or stopping of the breath in some part of the mouth or T and f T and f | and f |, and J and { J and ¢¥ f and [ f and { f and f f and { t and } 4 and 5 Consonants. CONSONANTS 31 throat. The main distinction between vowels and con- sonants is that while in vowels the mouth-configuration merely modifies the vocalized breath—which is there- fore an essential element of them—in consonants the narrowing or stopping of the mouth-passage is the foundation of the sound, and the state of the glottis is something secondary. Consonants can therefore be breathed as well as voiced, the mouth-configuration alone being enough to produce a distinction without the help of voice. All consonants can be whispered. 65. The friction of an ‘open’ consonant depends on the width of the organic passage. Consonants whose friction is distinctly audible both with breath and voice, such as ¢ and z, f and », are called ‘hisses’ when breathed or whispered, ‘ buzzes’ when voiced. Some consonants are formed with so wide a passage as to be almost vowels when voiced, and consequently almost inaudible when breathed. Such ‘vowel-like’ consonants are /, the weak English 7, and the nasals x, m, whose want of friction is due to the unimpeded flow of the breath through the nose. Breath ‘stop’ consonants,such as ¢, have nosound whatever in themselves, and are only rendered audible by the puff of breath or explosion that accompanies them (124). Voice-stops, such as d, are practically vowels in themselves, and their consonantal character is only shown atthe beginning or end of the stop. Zand thenasals betray their consonantal character in the same way: when they are final, the consonantal flap is distinctly audible, as in ¢e//, ten. Breath in con- sonants is left unsymbolized, voice being indicated by 32 ANALYSIS the incorporation of the voice-stroke, thus 0=/, o=d, w=voiceless, w=voice 7. Whisper is shown by the whisper-modifier ; as in c= whispered /. 66. The distinction of narrow and wide applies to consonants as well as vowels, though in consonants it is less noticeable, and can generally be ignored. Itis symbolized by the addition of the modifiers ¢ for narrow, : for wide consonants (43). If we consonantize t and t by narrowing the mouth-opening till a buzz is produced we shall find that the resulting 9; and 2 are still quite distinct, the latter being the English , the former the French sound in owi. Soalso narrow s may be heard in energetic hissing, wide si in gentle hushing—s# 24. Consonants seem to be generally wide in English, narrow in French. 67. Consonants admit of a twofold division, (1) ac- cording to form, (2) according to place. 68. The foundation of the consonant-symbols is a segment—c—of the open-throat symbol o. This fundamental symbol is modified to express the different forms of consonants-—thus a=the ‘stop’ 4—and is turned in different directions to show the place of the consonant—a=the ‘back’ 4, o=the ‘ point’ #, ete. Form. 69. By form there are five classes of consonants : (2) Open consonants are those in which the passage is simply narrowed without any contact. They are ex- pressed by the unmodified consonant-symbol or some modification of it: ¢ = ch in German and Scotch /och, CONSONANTS 33 >=f,S=s. The restriction as to contact applies only to the actual friction-channel, and even then there may be slight contact, provided the current of breath is not impeded. Thus, in forming c the uvula often touches the back of the tongue, but without modifying the sound perceptibly, and even in $ the tongue often comes in contact with the ridges of the gums without influencing the sound. In such a conso- nant as >, on the other hand, the contact of the lips and teeth has the effect of forcing the breath to seek a channel elsewhere, namely through the interstices of the teeth, which form the real friction-channel. It is, however, also possible to form an f between the lips and the teeth without any contact. (4) Divided consonants are formed by stopping the middle of the passage, leaving it open at the sides— often only on one side, giving a ‘unilateral’ consonant. The commonest type of this class is the ‘point-divided- voice ’ consonant 7. These consonants are expressed by indenting the consonant-symbol, sometimes by the ‘divided-modifier’ 4. The ‘ unilateral modifier’ is », (c) Stopped (or shut) consonants are formed by com- plete closure, as ina #,o7. They are expressed by a bar across the consonant-symbol, or by the addition of the ‘ stop-modifier ’ |. (2) Nasal consonants are formed by complete closure of themouth-passage while thenose-passageis left open. If we take any stop, such as o d, and allow the air to pass through the nose by lowering the soft palate, we obtain the corresponding nasal, in this case3 2, The 34 ANALYSIS symbols of the nasals are formed from those of the corresponding stops by joining on the nasal modifier to the bar, and omitting the consonant-segment. When an unstopped (open or divided) consonant is formed with the nose-passage open, it is said to be ‘nasalized,’ which is denoted by the addition of the nasal modifier 5; thus os is a nasalized 7. (ce) Trills (or rolled consonants) are a special variety of unstopped consonants—generally of open consonants. They result from the vibration of the flexible parts of the mouth, either against one another, as when the lips are trilled, or against some firm surface, as when the point of the tongue trills against the gums in forming the strong Scotch 7, and the uvulaagainst the back of the tongue inthe Northumbrian burred7. Their com- mon character is due to the rapid periodic interruption of the breath by the contact of the trilling body with that against which it is trilled, its elasticity—or, in the case of the uvula, its weight—causing it to resume its former non-contact, to be again driven or to fall back. Trills are therefore intermediate between open and stopped consonants. Trilling is indicated by the ‘trill modifier’ s, thus os = Scotch *. 70. All consonants may be pronounced with tenseness (\) or Zooseness (v),according to the degree of approxima- tion of the organs. Thus loose vis practically equiva- lent to the vowel [or f. Even stops may be pronounced loosely, so that, for instance, it may be difficult to distinguish between 0dandw4@, Tenseness and loose- ness are most nearly allied to the distinctions of height CONSONANTS 35 in the vowels,and must not be confounded with narrow- ness and wideness, which, in consonants as in vowels, depend on the shape of the tongue. PLACE. 71. By place there are six main classes: (2) Back (guttural), formed by the root of the tongue and the soft palate, expressed by turning the consonant- curve (representing the root of the tongue) backwards (34), the symbols of nasal consonants following the direction of the corresponding stop symbols. The ‘back modifier’ is (—a curtailed c. Examples are the back- stop G as in come, the back-nasal-voice 4 as in hing. (4) Front (palatal), formed by the middle part of the tongue and the hard palate, and symbolized by turning the consonant-symbo!l so as to be pictorial of the arched tongue, as in the front-open-voice @ inyow. The ‘front modifier ’ is \. (c) Point, formed by the point of the tongue and the gums or teeth. This class is commonly called ‘dental,’ but the point of the tongue is not necessarily brought against the teeth. Point consonants are symbolized by turning the points of the consonant-segment upwards in the same way as the point of the tongue is directed upwards in such a consonant as the point-open-voice © in red. The point-modifier is \. The point-teeth consonant in thinkis symbolized by sharpening © into V so as to be pictorial of the teeth- edges. There are two special modifications of point-position 36 ANALYSIS that require notice, inversion and protrusion. In in- version the point of the tongue is turned back—whence the ‘inversion-modifier’ (c)—towards the soft palate, so that the narrowing, dividing, or stopping is formed between the lower edge of the tongue-point (or blade) and the arch. c is the West-of-England inverted r. In protrusion the point of the tongue is extended to the lips—whence the ‘ protrusion-modifier’ (2)— as in blowing a small object from the tip of the tongue : 029. (d) Blade, formed by the blade of the tongue. The representative blade consonant is s=s, whose symbol is a combination of those of front-open and point-open— 9 and 0, blade being a compromise between these two. The blade-modifier is §. Blade-point is blade modified by raising the point of the tongue. Itis symbolized by reversing theunmodified bladesymbol,and its represen- tative is the 2 in she. (ec) Fan (spread) consonants, symbolized by the addition of the spreader 1 ($ 41), are modifications of point and blade consonants, of which the Arabic ‘emphatic’ consonants are the best-known examples, as in si[s> séf ‘sword, ovules tin ‘clay compared with séf ‘summer, ¢in ‘fig.’ In their formation the side edges of the tongue are spread out so that the hiss of si is formed not only between blade and gum, but also between the sides of the tongue and the back teeth; this lowers the pitch of the consonant, and gives it a peculiar ‘ cuttural’ effect. So also in on the sides of the tongue are strongly pressed against the back CONSONANTS 37 teeth, so that when the stop is released the off-glide has the same peculiar dull sound as in si. Ovi, Ol may also be heard as Irish substitutes for the English th-sounds v and v. (f) Lip (labial) consonants are symbolized by turn- ing the consonant-curve forwards, nasals following the direction of the corresponding stops, as in p p, F m. The ‘ lip-modifier ’ or ‘ rounder’ is ). Lip-teeth consonants, such as /, are symbolized ana- logously with the point-teeth consonants, by sharpening the curve of 9 into >, which is the symbol of f 72. Besides the simple positions, there are compound consonants formed by narrowing etc.the mouth-channel in various places at once. The most important are the front-modified and lip-modified consonants, formed re- spectively by raising the front of the tongue and narrowing the lip-opening simultaneously with some other action. These effects are symbolized by the addition of the proper modifiers. Thus w\ is a com- bination of w and @ (front-modified or palatalized 2), w) is a rounded 7. c) and o(—lip-modified back-open and back-modified lip-open, the back action predomi- nating in the former, the lip action in the latter—are for convenience made into c, as in German auch, and 9, asin what. The lip-back modifier is>. Thus a=/+w. There are many other combinations, expressed either by modifiers or by + between the symbols of the two elements, as in f[rastatowl- exactly. ato could also be expressed by au, where \ = ‘ point-stop modifier.’ 73. We have hitherto considered only the main 38 ANALYSIS ‘gardinal’ consonant-positions. Such a symbol as a really includes an infinite variety of positions, defined more accurately by the inner and outer modifiers, the combination + being used to emphasize medium position. The deep inner ax is not an English sound. 4: is the English sound before back vowels, as in caw, outer G+ being the English sound before front vowels, as in hey. The considerable difference between these varieties is clearly brought out by transposing the elements of the two words—a:fo, aye. It will be found on trial that a has a different position before each vowel, being less advanced before low front vowels, as in can ay3, than before high front vowels, and so on. 74, The distinction between inner, medium, and outer is peculiarly important in the front consonants, where the medium position is represented by English @ in you. Outer front comes near the blade position, but is dis- tinguished from it by absence of point-articulation, the point of the tongue being kept clear of the palate. It is still more difficult to distinguish between blade and the combination o\—outer front + point. 75. The medium point position is on the gums just behind the teeth, where English ¢ and d are formed, in which, however, the tongue is generally flattened into the blade-shape. In English and most other languages point 7» tends to the inner position—o. Hence the blade-point 2 in she is naturally more retracted than the unmodified blade-consonant $, but lessso thane. But even © can be advanced almost to the teeth, although it ultimately merges inv. If the point is retracted to CONSONANTS 39 the arch-rim it practically assumes the outer inverted position—@c. Outer pointimplies articulation against the teeth, or with the point half on the teeth, half on the gums. When the tongue is put between the teeth, the consonant is called ‘ interdental,’ which may be indicated by +. 76. We have hitherto seemed to take for granted that the tongue always articulates against the nearest part of the palate, that, for instance, in a: the inner front of ; the tongue is brought against the inner front of the palate. But it is possible to produce a compromise between a and a in quite a different way, namely by bringing the medium front of the tongue against the outer back ofthe palate. The neatest way of expressing this would be to denote the palate-positions by the addition of diacritical consonants ; and it is probable that as our analysis becomes more minute, some such device will be adopted. Meanwhile we will content our- selves with expressing the palate-positions by ordinary consonant-symbols in parentheses. Thus the combina- tion we have just been considering can be written Qu+(a+) or more briefly a(ar). NoN-ORAL CONSONANTS. 77. The throat-consonants—the glottal stop and the various wheezes—have been described above (29, 31). 78. There is also a uvula-stop 51, but it occurs only in such combinations as du in didden. Here the mouth- stop is maintained from the beginning of the d to the end of the », the sole change from the d to the x being 40 ANALYSIS the opening of the nose-passage required to form a nasal consonant. As this action is necessarily implied by writing 05 together, it is superfluous to indicate it further. If we devocalize the x of such a word as eaten, making it into [03 with a strongly snorted 3, we can hear and feel the uvula-explosion distinctly. The action 03 is a not uncommon and very disagreeable form of sniffing. TABLE OF CoNSONANTS AND GENERAL REMARKS. 79. The chief consonants are shown in the annexed table, with key-words. 80. In naming the consonants place comes first, then form, and breath, etc. last. In the stops and hisses ‘breath’ may be omitted. Thus a & back-stop, > v lip- teeth-open-voice, os 7 point-trill-voice. 81. The consonants are generally easier to recognize by ear than the vowels, the chief acoustic difficulties lying in the various transitions between them and the adjacent sounds, which will be treated of under Syn- thesis. s2. But their articulation often offers great diffi- culties especially in the case of the trills, which require long practice. ACOUSTIC QUALITIES OF CONSONANTS, 83. The following table shows the order of the pitches of the chief open consonants, beginning with the lowest : D9GcIa>rOV2S5O,. CONSONANTS Se eg a i epee = 2s ue eg ee SN Us "72DI7 burs um A 416, Ug u 5 & c aoe 1D8DN' | Se RS aes Pe a eee es eA HT ae wag l< | qq IZ;IS SQ} ww] po fO 25) 26% he Se cre ae a ee a EAD Neat Ne ieee, | Seve oat ee | foe A ORES = | —— ane ‘13 ‘7077 | BYyed ‘ssnz S (aa) rats) ae) 13 “a papuruy MY) YyNON abnor me8Yy? nofh uabes +9 UIs, “ AE wa 6 g€ 3 z$ QM I” CO 29 r+] uado Y}08-"'T | Yoo_-"T dyT qupod-7g | epnrg | YI90I~"T | quL0T JUOLT YT PDOLY,T, OI 6 8 Z 9 s 14 £ z 1 te “SSH'TIHOIOA Scag enEERnnenmnemeeneemeeneee nee ee SWEET 42 ANALYSIS 84. Observe the close analogy with the vowel-pitches, the lowest-pitched consonant 2 being a lip-narrowed unvoiced form of the lowest-pitched vowel t, and © corresponding to [. Within each consonant there are lesser gradations of pitch from inner to outer, thus 04, 0+, oF; which, again, agrees with the relations of con- sonants to vowels, or being the exact consonantal equivalent of [ (86). RELATIONS OF CONSONANTS TO VOWELS. 85. The various positions of the open voiced con- sonants must necessarily yield more or less distinct vowel-sounds when expanded enough to remove audible friction. The relations between the consonant and vowel positions are very important, and should be carefully studied, Thus, starting from buzzed medium ex the student should carefully increase the distance between the back of the tongue and the soft palate till he obtains a pure vowel-sound—which will be J or ]. The following are the more important of these relations: E1 € E+ M1M M Gi G& Et, M)4 M) Mt Pd Ayt Lb Fide ty eat at 8G. In comparing 1, , o with t, [, [ ete., we see that the retraction of the tongue-narrowing from @ to m: corresponds exactly to a similar progression in the vowels (36). It would, indeed, be possible to substitute some such symbolization as [+ [ [4 for [ [ [; but this would be ambiguous and would ignore the distinctive peculiarity of vowels as opposed to con- CONSONANTS 43 sonants, namely their power of indefinitely expanding the voice-channel from which result the distinctions of height. 87. The point-consonant © may be weakened into a vowel, the result being practically a low mixed point- modified vowel—ju. 88. wv, 8, and 2 may be weakened in the same way, with similar results. SOUNDS FORMED WITHOUT EXPIRATION. 89. All the sounds hitherto described imply. out- breathing or expiration (>). It is also possible to form sounds with in-breathing or inspiration (<). Ibis a com- mon habit of speech to pronounce such words as o in this way, to express emphatic denial. 94 isthe natural symbol of drinking, and 9) is an ejaculation of pain. 90. Some sounds are produced without either out- or in-breathing, but solely with the air in the throat or mouth. The ‘implosives’ (130) are formed in the former, the suction-stops or ‘clicks’ in the latter way. In the clicks the tongue or lips are placed in the position for a stop, and the air is sucked out from between thestop-forming organs; they are thus pressed strongly together by the air in the mouth, so that their separation produces a sharp smacking sound. This action is regarded as a kind of stopped inspiration and is accordingly expressed by adding a stop to the in- breather. Thus p< is an ordinary kiss, 0+ is the ex- pression of impatience written tué / In many savage languages these clicks are essential elements of speech. D2 SYNTHESIS 91. Wr have hitherto considered sounds from a purely analytical point of view, that is, each sound has been considered by itself, as if it were a fixed, isolated element. But in language sounds are com- bined together to form sentences, and many sounds occur only in certain fixed combinations. Hence the necessity of synthesis as well as analysis. Analysis regards each sound as a fixed, stationary point, syn- thesis as a momentary point in a stream of incessant change. Synthesis looks mainly at the beginning and end of each sound, as the point where it is linked on to other sounds; while analysis concerns itself only with the middle of the fully developed sound. Syn- thesis is thus the science of sownd-joints or ‘ glides.’ There is also a more general kind of synthesis which deals with the ve/ations of sounds to one another in sound-groups—their difference in length, loudness, pitch, ete. Synthesis, lastly, deals with the organic and acoustic grouping of sounds into syllables, ete., and the divisions between these groups. 92. The popular fourfold division of the elements of speech into letters (that is, sounds), syllables, words, and sentences, is not purely phonetic, but also partly graphic and logical. The first and most impor- SYNTHESIS 45 tant point to see is that our ordinary word-division is a logical and not a phonetic analysis. No amount of study of the sounds only of a sentence will enable us to recognize the individual words of which it consists. We may write down every sound, every shade of synthesis, but we shall never be able to analyse the sentence into separate words till we know its meaning, and even then we shall find that word-division postu- lates much thought and comparison of sentences one with another, 93. The only division actually made in language is that into ‘breath-groups.’ We are unable to utter more than a certain number of sounds in succession without renewing the stock of air in the lungs. These breath-groups correspond partially to the logical divi- sion into sentences: every sentence is necessarily a breath-group, but every breath-group need not be a complete sentence. 94. Within each breath-group there is no pause whatever, notwithstanding the popular idea that we make a pause between every word. Thus, in sucha sentence as put on your hat we hear clearly the ‘recoil’ or final breath-glide which follows the final ¢ of at, but the ¢ of put runs on to the following vowel without any recoil, exactly as in the single word putting. In put back there is no glide at all after the ¢ (143). 95. The only phonetic divisions in a breath-group are those into sounds and syllables and intervening glides. 96. The most important general factors of synthesis 46 SYNTHESIS are quantity and stress, which modify glides and also constitute relations between adjoining sounds. Quantity. 97. We may distinguish five degrees of quantity or length : very long di long qt half long or medium jé short de very short dee but for practical purposes the threefold distinction of long, half long, and short is generally enough. Often, indeed, it is not advisable to do more than denote the distinction of long and short, assigning ¢ to long, and leaving short unmarked. 9s. Full length may be heard in English stressed vowels when final, as in sea, and before voiced conson- ants, as in seize, half-length in stressed vowels before breath consonants, as in cease. Short final stressed vowels, as in French si, are difficult to English speakers. 99. The distinctions of quantity apply to consonants as wellas vowels. In English consonants are long after a stressed short vowel, as in Aid/, short after a stressed long vowel, as in /ee/. But in such cases the length is often distributed equally over vowel and consonant. It may also be observed that length in the case of a pbreath-stop like ¢ means length of pause or cessation of sound. Short consonants after short stressed vowels FORCE 47 offer great difficulties to English speakers; they may be heard in German words such as mann. 100. We can also distinguish degrees of rapidity of speech in different breath-groups or longer periods. Such differences of ‘tempo’ may be indicated by pre- fixing the quantity-marks + the symbol of breath: O+ :a]F Jo = come up uttered slowly. Force (Stress), 101. Force, like quantity, belongs essentially to synthesis, for it is always relative, always implying comparison either of two different groups of sounds or of two different portions of the same group. Physically it is synonymous with the effort by which the breath is expelled from the lungs. Every impulse of force is therefore attended by a distinct muscular sensation. Acoustically it produces the effect known as ‘loudness,’ which is dependent on the size of the vibration-waves which produce the sensation of sound. When we say, therefore, that one sound, or group of sounds, is uttered with more force than another, as in the first syllable of heavy compared with the second, we mean that in its utterance the air is expelled from the lungs with a greater muscular effort, and that in consequence the resulting sound-waves are larger, which produces an effect of greater loudness on the ear. Force in its synthetic sense must be carefully dis- tinguished from those gradations in the friction of unstopped consonants which are due to the varying width of the configurative passage (70), although, of 48 SYNTHESIS course, all articulation postulates a certain amount of force to be audible at all. 102. We have now to consider the changes of force in a single breath-impulse, as in pronouncing any long vowel, such as Jt. Here we have three degrees of force, level = increasing (crescendo) t< diminishing (diminuendo) ]#>. 108. In examining the force of any sound-group it is a great help to whisper it, which gets rid of any disturbing changes of pitch. 104, The general tendency of language is to pro- nounce with diminishing force. Thus in English the e of cat is pronounced with more force than the ¢. Hence also the end of a long is weaker than that of a short vowel, the force diminishing continuously throughout the long vowel. Thus the ¢ of cart is weaker than that of cat. In German the diminution of force is still more rapid than in English. In French, on the other hand, the force is nearly equal, the final t of toute Orlor, for instance, being pronounced with as much, or nearly as much force as the initial one. Increasing force may be heard in interjections, such as j#< denoting joyful surprise. Here it is accompanied with a marked rise in pitch, but if it is whispered, the < is unmistakable. 105. The influence of force on the synthesis of speech is very important, for the sense of unity and separation depends mainly on it. Continuity of force gives a FORCE 49 sense of wnity, as in }f<, J+>, ]t<>, discontinuity, as in <<, Jt>>, that of separation, the J+ being broken up into two syllables, even when there is not the slightest pause. | 106. The comparative force with which the syllables that make up a longer group are uttered is called ‘stress. In speaking of the stress of words in a sen- tence as opposed to that of syllables in a word the term ‘emphasis ’ is commonly used, but this distinction is not admitted in phonetics, which ignores word- division (92), and divides sentences straight off into syllables. 107. There are three main degrees of stress : strong (+), Aadf-strong or ‘medium’ (:), and weak (*). Very strong or ‘extra strong’ may be marked (;). (:) is assumed to be an abbreviation of (*), which is a modification of \(70). Weak stress is generally left unmarked. (*) may then be utilized to indicate a weak stress slightly stronger than another weak one. In connected texts where it is necessary to mark unstressed words, weak stress is denoted by (-). All stress-marks are put before the element on which the stress begins, so as to leave room for quantity-marks and other diacritics after it. Another advantage of this method is, that it marks the divisions into syllables (156). All three degrees of stress are shown in such a word as contradict ‘azIow,ofao. The sentence come at once! has exactly the same stress: :a]r0'9)3s or :a]r -0 ‘a]55. 108. The degrees of stress are really infinite, and in a single sound-group (word or sentence) every syllable 50 SYNTHESIS may have a different degree of stress. Thus such a word as wmpenetrability has, roughly speaking, two stresses, a strong one on the fifth, and a medium one on the first (or sometimes on the second)—:lrpf3]|00], plol-ol-. But if we pronounce -dzlity by itself, we shall find that all three syllables have a different stress, the third being stronger than the second and weaker than the first. In penetra- there is the same relation, but all the syllables are a shade weaker than the corre- sponding ones in -dz/ity. The order of the syllables in stress is therefore as follows, 1 being the highest: MEE BN ah FE ei eee | fe pf 3, col ol of ols. 109. The surest way of determining the relative force of any two syllables is to pronounce the other syllables mentally only, or in a whisper, pronouncing the special syllables aloud, and their relative force will then come out clearly. Thus, taking -dz/ity by itself, if we utter the first syllable mentally, the other two aloud, we shall find that the second of these two has the stronger stress. | 110. There is an important feature of stress generally which in most cases makes any minuter symbolization of stress unnecessary. This is its rhythmic character, or the tendency to alternate weak and strong stress. Per- fect uniformity of stress is as phonetically unnatural as level force in a single syllable, but the tendency of stress is not, like that of a single force-impulse, to decrease progressively, but rather to sway to and fro. Hence in FORCE 51 a group of three syllables, of which the first has the pre- dominant stress, we may generally infer that the second will be weaker than the third, as in relative ‘olwlol->. Of course, in very rapidspeech these minute distinctions become unrecognizable, the two last syllables of such a word as relative being apparently uttered with a single progressively diminishing force-impulse. 111. But stress in all languages is more or less governed by logical as well as phonetic laws, which, of course, often clash. Level stress is, accordingly, very common in English, as in thirteen when uttered by itself, while in ¢hzrteen years the stress on the second syllable is diminished: -vyrolos, ‘vy+:ofo3 ‘of|s. In French there is a general tendency to level stress, the strong syllables rising only a little above the general level. The discrimination of degrees of stress is no easy matter in any case, because of the counter-associations of quantity, intonation, and vowel-quality, which make us apt to fancy that long, high-toned, or clear-vowelled syllables have stronger stress than they really have. A long weak-stressed vowel may be heard in the drawling pronunciation of what a pity! :oyo -| ‘plol++. A stress- less clear vowel may be heard in such a word as zusect ‘[ssfao compared with edict ‘{mol-a0, whose second vowel is one that occurs only in unstressed syllables. 112. Difference of force in whole groups of sounds miay be indicated analogously with differences of speed (100), thus ‘o :a]r ‘Jo denotes come up! uttered forcibly. 52 SYNTHESIS Glides. 113. Synthesis introduces us to a special class of elements called ‘glides’ or transitional sounds, pro- duced during the transition from one sound to another. Thus in such a group as al+ (or English sey) we have not only the two sounds a and [, but also the sound produced in passing from the back position of a to the high-front one of [. This glide differs from the two extremes, a and [, in having no fixed configuration : it is, in fact, composed of all the intermediate positions between q and [, through which it passes without dwelling on any of them. 114, lt would clearly be impossible to symbolize all the infinitesimal intermediate positions of which a glide is made up; nor is it ever necessary, the general principle being that in all cases of transition from one position to another the shortest way is taken: given, therefore, the symbols of the fixed positions, the direc- tion of the glide follows as a matter of course. These ‘essential’ glides are implied simply by the juxta- position of the symbols of the fixed positions between which they lie, as in af. Vowel-glides are expressed in the same way as an ‘unsyllabic’ vowel (153), namely by writing the vowel symbol consonant size, thus r= glide-f. Consonant-glides are expressed by adding ), thus m)=glide . 115. Glides are distinguished, according as they pre- cede or follow a sound, as ‘ on-glides’ and ‘ off-glides.’ Thus in Jaf (or English echo) the on-glide of a is that from the J, and the off-glide of a is that which joins it GLIDES 53 to f. Initial glides, such as the on-glide of thea in af, being only preceded by a silence, are generally in- audible. Final, or ‘on-silence’ glides, such as the off- glide of a in fa (or English e#e) are generally audible. Although the direction of a glide is implied by the position of the two fixed points between which it lies, its character may be varied. 116. The acoustic effect of glides varies according to the force and rapidity with which they are uttered. If in such a group as Jo] the transition from the J to the is made rapidly and with slight force the glide is not noticed at all, although any break, however slight, would be at once heard. But if the transition is made slowly and with only a gradual diminution of force, we hear the glide from the ] up to the front position of the m as the second element of a diphthong, giving the effect of (aija). Insuch combinations as oJ, as in Spanish //ano, itis often difficult to know whether to write the glide or not, whether to write m]3} or cor} 3}, om)s}. VoOWEL-GLIDES. 117. Vowels may be begun and finished in various ways: (a) The glottis is gradually narrowed, passing through the various positions from breath and whisper till voice is produced. This gives the ‘gradual’ beginning °J, which is the usual one in English, and in most other languages. (6) The breath is kept back till the glottis is closed 54 SYNTHESIS for voice, which begins at once without any introductory breath. This is the ‘clear’ beginning '}+, well known to singers, who are taught to avoid the ‘ breathy’ °}}. It is the usual way of beginning a vowel in German. 118. In both these cases the stress begins on the vowel. If it is thrown on to the preceding glides, they are at once recognized as independent elements, *]+ (aa) becoming os (haa) with the ‘aspirate,’ while ']+ becomes x}+ with the glottal stop (29). °, which is simply a glide-o, is generally modified by the following vowel, whose mouth-configuration it partly anticipates, the organs moving from the beginning into the position for the vowel, although they do not fully reach it till the aspiration ceases. It is then partly a weak throat-consonant, partly the voiceless glide-vowel cor- responding to the vowel which follows, and is then placed on a shortened vowel-stem, being written ¢. ef, ot are thus almost equivalent to ref, 2st or ov[, vi. 119. But there is also a ‘strong aspirate, in whose formation the full vowel-position is assumed from the beginning of the aspiration, which is, therefore, literally a voiceless vowel: Ac = sf or ssf. The strong aspirate may be heard in Finnish, and occasion- ally in American-English, Swiss-German, and other languages. 120. In actual speech the acoustic effect of aspiration is often produced by substituting imperfect vocality for breath, giving the ‘ voice-aspirate ’ ¢: or ‘jerk of the voice, as Ellis calls it. Most languages pronounce i between voice-sounds in this way, as in E. a house, GLIDES 55 behold compared with hold ! this house, and the emphatic aha! It is easy to hear that in dehold the voice-murmur is continuous, and yet the contrast between the faint vocality of the 4 and the full vocality of the adjoining vowels produces much the same effect on the ear as that between full voice and full breath. Some languages, such as Czech, employ this ‘half-voice’ initially as well as between vowels. 121. Vowels are finished analogously : (2) By a gradual opening of the glottis, the final glide passing through whisper to breath, giving the ‘gradual’ ending Jj. (4) By a cessation of expiration while the glottis is still closed for voice, giving the clear ending ]', which is the usual ending in English. 122. If uttered with stress these endings become re- spectively Jo or jg, which is still the Sanskrit visarga, as in manah, and jx, the Danish ‘stéd-tone’ (29). If a high vowel-position is relaxed slowly before ° or g, the off-glide has a strongly consonantal character, even if there is very little stress, giving [ov, tov, ete., which is frequent in the Scandinavian languages. CONSONANT-GLIDES. 123. Stops: Off-glides. All consonants consist acoustically of three elements, the consonant itself, and its on- and off-glide. Each of these three elements may be breathed or voiced, and modified in other ways as well. 124. The glides of stops are peculiarly important. 56 SYNTHESIS Voiceless stops, indeed, are, acoustically speaking, pure glide-sounds, only audible at the moment of transition from or to some other sound. Voiced stops, on the other hand, can have a distinct sound of their own in addition to that of their glides, but as they can only be voiced by driving voiced breath into an air-tight chamber—forming the ‘blahlaut’ of the German phoneticians—they cannot be maintained for any length of time. 125. Confining our attention for the present to the off-glides, we may distinguish four chief kinds of breath—or rather voiceless—and voice-stops : (a) voice- less stop and breath-glide av], as in come; (6) voiceless stop and voice-glide a'], nearly as ingo when no vowel precedes ; (c) voice-stop and breath-glide a’, as in egg; and (d@) voice-stop and voice-glide 4', as in eager. 126. The following table gives all the combinations —initial, medial, and final—those that occur in English being marked *: INITIAL. MEDIAL. FINA. Qe] *a*] *Ja° a") Ja") Ja’ ar] jar) *'Jae a) *a') ja’. (a) In a’ the glottis is left open while the stop is being formed, and the chords are not brought into the voice-position till the moment of loosening the stop, so that before the glottis has time to form voice there is a slight escape of breath between the stop and the vowel —the glide from the stop to the vowel, or from the stop GLIDES 57 to silence, is breathed. In English the puff of breath varies in force according to emphasis, ete., and is always weaker medially than initially or finally, as may be seen by comparing the second a in cooking with the two as in cook, where, again, the initial a has the stronger glide of the two, because of the progressive diminution of stress (104), whence also the force of the breath-glide is still more diminished after a long vowel. as in eating. (2) a' seems to be formed in two ways. In initial English go at the beginning of a sentence the glottis is in the position for voice during the stop, but no air is driven in, and so the stop is inaudible as in &, but voice begins the moment the stop is loosened, and the off-glide is therefore voiced. In this kind of stop the voice is apt to break out a little before the end of the stop or at any rate to whisper part of the stop— ay]. Pure a'is formed by sounding voice simulta- neously with the loosening of the stop, so as to prevent any escape of breath. The French and South German (South European generally) &, ¢, », are formed in this way, which makes them sound like g, a, 6, to an un- accustomed ear. | These sounds offer great difficulties to English speakers, who, however, will find initial g 10 go a con- venient stepping-stone. But they must be practised carefully, for the breathy stops in English come, etc., are very offensive to a South European ear. The student must not be satisfied till he can explode a vowel loudly from the a, 0 or p position without the SWEET E 58 SYNTHESIS slightest escape of breath or voicing or whispering of the stop. (c) Jac}, and thence a], may be easily obtained from the familiar Jae by joining on a J. (d) @'] is obtained by pronouncing the familiar Ja‘), dwelling on the consonant, and then dropping the initial }. These ‘full’ initial voice-stops, which are common in the South European languages, suggest the corresponding nasals to an unaccustomed ear when sounded very fully. Final a may be obtained by shortening such a word as digger. The French final g, as in pyre! dague, often has this sound. 127. Stops: On-glides. The on-glide after a vowel is generally voiced: ]'a°,J'a°. Breath on-glides before voiceless stops occur in Icelandic and occasionally in Scotch, as in of}*o° what. 128. Stops: Stress-glides (Aspirates). All stops, especially when voiceless, postulate a certain com- pression of the breath behind the stop, so as to produce an audible explosion when the stop is loosened. On the force of this compression, which is due to the action of the lungs, the force of the glide and consequently the audibility of the stop mainly depend. The English k, ete. are generally uttered with but little force, but in the ordinary German /, as in a’n}3 4ann, there is a strong puff of breath. 129. But even in German the force of the breath- glide is something secondary, due only to the compres- sion with which the stop is formed. If, however, the initial force is maintained during the formation of GLIDES 59 the glide itself, the glide is heard as an independent element—ae]. In this way the Irish-English, Danish, and Sanskrit aspirates are formed, as in Irish /e//, Danish ¢a/e coy]. These sounds have nothing harsh about them, their characteristic feature being the distinctness of their glide, which has something of the character of the preceding consonant, so that ae], for instance, sounds very like ac], 0°] like Ov) or os). The analogies with the different vowel-beginnings (117) are obvious. 130. Implosive stops (Choke-stops). The implosion consists in closing the glottis simultaneously with the stop position, and then compressing the air between the glottis stoppage and the mouth one by raising the glottis like a plug by the action of its muscles. This action produces no sound while the stop is being held, but modifies the off-glide in a peculiar manner, giving it a ‘choky’ effect. These sounds occur in Saxon-German and Armenian. They are written a), with the throat-stop modifier ; the off-glide is voiced. 131. Unstopped Consonants. With unstopped con- sonants there is no difficulty in voicing the consonant itself, and there are many consonants that are only occasionally unvoiced, such as the nasals. The glides of these consonants are always voiced as well. 132. In the breath unstopped consonants, both the consonant itself and the off-glide are breathed as in the corresponding stops : s*], Js°], Js*. But the breath- glide of the unstopped consonants is always weaker E 2 60 SYNTHESIS than that of the stops, because the explosive effect of the latter is wanting. : 133. Hence also the aspirates of these consonants are weaker and less marked than those of the stops. But they may be distinctly heard in Irish-English in such words as s2. 134. The voiced buzzes admit of more variety than the voiced stops, because in them the different stages of glottis-narrowing that may precede voice are dis- tinctly audible, whereas in the voiced stop there is hardly anything between full vocality and absolute silence. 135. In medial z (]s]) there can be no doubt of the vocality of the consonant, but initial and final z admit of various degrees of vocality : (2) The glottis does not begin to put itself in the position for voice till the s-position is assumed, and consequently all the intermediate stages between full breath and full voice are heard in succession while the s-position is being maintained. This is the ‘ gradual’ initial z in English zead, ete.—s°]. (6) The z is fully vocal throughout—that is, the glottis is closed for voice simultaneously with the beginning of the z. This is the ‘clear’ initial z of French, Russian, etc.—s:). It will be observed that these varieties of initial buzzes are exactly analogous to the two ways of begin- ning vowels (117). (c) The glottis is open during the formation of the consonant, and is only brought together at the moment GLIDES 61 when the off-glide begins. This is the ‘ half-voice’ z (s'), corresponding to initial a', and seems to be the usual German initial s in so, etc. As it is not easy to make the beginning of the voice correspond exactly with the beginning of the glide, this last variety is often modified into a compromise between (a) and (4), formed by beginning to narrow the glottis during the end of the consonant itself, so that the transition from breath to voice is completed just defore beginning the glide. 136. Final z may also be fully vocalized throughout, or else gradually devocalized, passing from voice to whisper while the consonant position is still being maintained—jsi°, Js. Both may be heard in English as, etc., the latter ‘gradual’ ending being the most usual. The ‘clear’ ending with voiced off-glide—]s}— is the usual one in French. 137. A final buzz preceded by a buzz or voiced stop is completely whispered in English ; as in ¢hieves, rage Ulm>sr, ofr-D20. 138. In this last case the glottis is not fully opened till the consonant is finished, which therefore consists of voice passing into whisper, followed by a breath off-glide. If the transition from voice to breath is completed during the beginning of the consonant it- self, we have the ‘half-breath’ final z—]ss. 139. The vowel-like consonants when final occasion- ally end in a breath-glide. Thus in French fille=>[o3: in Icelandic vel=>ftows. 140. Whisper-glides. In the case of stops, whisper 62 SYNTHESIS is inaudible in the stop itself, and is only heard in the glide. In most cases a whisper-glide is a transition to or from voice, and has the effect of a weak breath- glide, from which it can generally hardly be distin- guished. Final Ja*p, however, is easily distinguishable from Jar. It is heard in Icelandic egg. 141. Modified Glides. We have hitherto considered consonant-glides as modified mainly by voice, breath, and force. But they are capable of various oral modifications as well, of which rounding is the most important. Rounded glides are heard in such Russian words as voti >for, komnata :ad}e3)}o), where the inner rounding affects the consonant as well as the glide. a>} sounds intermediate between (ko) and (kwo). In English cool the off-glide is only slightly rounded by the following vowel. GLIDELESS COMBINATIONS. 142. In speech the general principle is to take the shortest way between two sounds in immediate juxta- position. This often results in combinations which are effected without any glide at all. This is regularly the case in sequences of consonants having the same place, and differing only in form. Thus in passing from (n) to (d) or vice versa in 30, DF all that is done is to close or open the nose-passage, the absence of glide being as much implied by the juxtaposition of the two symbols as in the case of the uvular stop (78). Similarly in wo ow the transition is made by simply closing and opening the side apertures, the tip of the tongue GLIDELESS COMBINATIONS 63 retaining its position. Combinations in which a stop is followed by an open consonant formed in the same, or nearly the same place, are effected either with no glide at all, as in po, or a very slight one, as in p>. In such combinations as the latter one the glide is generally got rid of by assimilating the place of the first consonant to that of the second. Thus German pf in pfund and English m in nymph are both lip-teeth instead of the pure lip-consonants p, F, these words being pronounced >pt30, 3f>\>. Such changes may almost be considered as implied by the juxtapositions >a >e 143. Even when two consecutive consonants are formed in different parts of the mouth, it is possible to form them without any glide, although in such cases gliding combination must be regarded as normal. Absence of glide is marked by (.). Thus English act is [a.0, the tip of the tongue being brought into posi- tion before the q-contact is loosened, while in French active there is a slight glide—jyrarol>. In some East- Asiatic languages final stops are always glideless : Ja., ete. 144, Combinations of stops and vowel-like con- sonants, such as ¢r, dr, kl, kw, are glideless in English, the breath-glide after a voiceless stop being carried into the vowel-like consonant, the first half of which it unvoices, as in ¢ry 00)0)}rtr, clock aw)w)za. If the preceding stop is aspirated, its aspiration may be carried into the vowel-like consonant, so that the latter is completely unvoiced. Thus in Danish, where 64 SYNTHESIS initial voiceless stops are aspirated, initial Ar, £7, etc., become stop + breath / ete., as in awza'] kokke, aspx kne. GLIDE-CONSONANTS. 145. Most consonants, as compared with vowels, have more or less the character of glides. Breath stops are acoustically pure glides (124). In such combinations as 02 in chill, 0s in German Os] £-0 zeit, the hiss is acoustically a mere modification of the breath-glide in ¢i//: we may almost say that the 2 or S is the glide between the 0 and the next vowel. 146. In slovenly speech, when a stop follows a vowel, the breath impulse is often so feeble that nothing is heard but part of the glide on to the consonant, the actual closure being formed without any breath at all: pla) dig. With nasals man becomes rfIs), only a nasal glide being audible. Other consonants are weakened in a similar way. 147. But there is a class of Flap-consonants which are pure glides, organically as well as acoustically, there being absolutely no fixed point in their formation. The East Norwegian and Swedish ‘ thick 7’ is such a sound. It is an inverted 7 finished off with momentary contact of the tongue-tip against the inside of the arch-rim, the tongue moving forwards all the while from the moment of its being turned back to the single strong trill which finishes it. This sound ean be roughly symbolized by wc, as in Norwegian ‘}+y oc] Ola = Olaf. SYLLABLE DIVISION 65 Syllable Division. 148. Sounds differ much in sonority—the force with which they strike the sense of hearing. The most sonorous sounds are those formed with voice, and the less the voice is impeded, the more sonorous the sound. The two extremes are the opener vowels, such as J, 5, and the stops ; the high vowels, such as [, being about on a level with the vowel-like consonants, of which the nasals are the most sonorous. Of the voiceless consonants the high-pitched hisses are the most distinct. 149. The audibility of language depends then, rough- ly speaking, on its vowels. Acoustically, consonants are mere modifiers of the vowels, and the ideal of distinct- ness would be reached by a language in which each consonant was separated from the next by a vowel. 150. Hence the ear learns to divide a breath-group into groups of vowels (or vowel-equivalents), each flanked by consonants (or consonant-equivalents)—or, in other words, into syllable-formers or sy//abics, and non-syllabics, each of these groups constituting a syllable. Syllabics are marked by ], non-syllabies by ), when necessary. 151. The relation between syllabic and non-syllabic is evidently a purely relative one. In such a group as clay the sonority of the vowel completely overpowers that of the 7 and makes it non-syllabic, but the 7 in cattle ajOwW = aTow] is so much more sonorous than the 0 that the whole group is disyllabic to the ear, 66 SYNTHESIS as if the 7 were accompanied by a vowel. Even a voiceless hiss may be syllabic in such a combination as DStO or even NSO. 152. The same sound varies in audibility according to the length and force with which it is uttered. When two vowel-like consonants come together, the one that has the greater length and force is regarded as the vowel. Thus F'3+=F)53] suggests F)3, while :r5 =F|3 suggests Jr3 or rather ‘Jr. 153. So also a vowel ean lose its syllabicness in combination with another vowel, with which it then forms a diphthong. These diphthongic or glide vowels are written consonant size, being from a syllabic point of view consonantal vowels, as in jf a2, where the group is uttered with one impulse of diminishing force, and £] 7a, which implies increasing force, the latter diph- thong being equivalent to ov]. Want of stress is more essential than gliding quality, for Jr with the diph- thongic vowel lengthened is still mainly diphthongal to the ear if the £+ is kept stressless. 154. The unsyllabic element of a diphthong is gene- rally a closer vowel than that which constitutes the syllabic element. The most perfect types of diphthongs are, therefore, ai and au, which are also the commonest. When clearly formed—jr, J1—the second elements are almost consonantal in character—suggesting a7, aw to an English ear—because English and most other lan- guages content themselves with making the second element a mere approximation to the high position ; thus in English the nearest equivalents of the above SYLLABLE DIVISION 67 diphthongs are J+ and j%, the second elements being still more obscured in Cockney pronunciation. 155. Such combinations as 02 may be regarded as consonantal diphthongs (145). 156. The answer to the question, Where does the syllable begin? is, that if it has a distinct stress (strong or medium) its beginning corresponds with the beginning of the stress, as we see in comparing such pairs as ]O'J+w:0)}+£-F8 and | OftweT3, |O'[ao|3 and ora=at all times, a tali man ; at Acton, attack. 157. The difference between long and double con- sonants is a syllabic one. In Jo] the consonant posi- tions are simply held with uniformly diminishing force till the J is reached, when a new impulse may begin. In Joo] the consonant is held as long as in the preceding case, but the new force-impulse begins in the second half of the held consonant, which, of course, breaks the sense of continuity. This break is very distinct in such a group as Jookcase ‘pta:afrrs, because of the medium stress on the second syllable. 158. The distinction between ‘aJo‘zjto and |ozlm> cut short and achieve, ‘JtO'S)+£-0 outside and the German aL oslter geziemen is exactly analogous. 159. The distinction between close and open stress is also syllabic. In the close stress of English and gene- rally North German in such words as Jetter, vetter, the 0 is uttered with the same force-impulse as the preceding vowel—although this force-impulse has diminished by the time it reaches the 0 (more so in German than in English, § 105)—any new impulse beginning on the 68 SYNTHESIS following vowel: ‘pfov|, >fo’le. In the open stress of South German, and the South European languages generally, the fresh impulse of force begins on the 0 in such a word as vetter—>f'O| os. Open stress—which is also heard in Welsh—sounds less abrupt than close stress, and to an unaccustomed ear suggests doubling of the consonant. The otherwise superfluous (*) may be used to indicate open stress, as in the South German Sfo'Los, Welsh sarjo'k cadw. 160. When several syllables are uttered with one impulse of force, it is, of course, impossible to mark off the boundaries of the syllables by stress, and syllable- division becomes a subjective problem. It is, for in- stance, difficult to hear much syllable-division in such a word as zecessary, when uttered rapidly. Syllable- division is most clearly marked when it turns on stopped consonants, because of their greater force and abruptness. It is less clear when it turns on other consonants. Thus the difference between au aim and a name, between alla with long and with double / is not very marked. Intonation. 161. Intonation, or variations of tone (pitch), depend on the rapidity of the sound-vibrations, which again depend on the tension of the vocal chords (17). Changes of tone may proceed either by /eaps or by glides. In singing the voice generally dwells without change of pitch on each note, arid leaps upwards or downwards to the next note as quickly as possible, so INTONATION 69 that although there is no break, the intermediate glide is not noticed. In speech the voice only occasionally dwells on one note, and is constantly moving upwards or downwards from one. note to the other, so that the different notes are simply points between which the voice is constantly gliding. An absolutely level tone hardly ever occurs in speech, whose level tones are only relatively level, generally ending in a slight rise. There is often in speech a marked difference between a rapid rise or fall in which the ear is mainly impressed by the beginning and end of the voice-inflexion, and a slow glide which allows the intermediate tones to come out. We may distinguish these as voice-leaps and voice-glides, remembering that the distinction is only a relative one, which cannot always be made with certainty. The difference between voice-leap and voice-glide is analogous to that between Jo] and jim] (116). 162. There are three primary ‘forms’ or ‘inflections’ of intonation : level . rising ’r falling ‘a. 163. (’) and (*) are, strictly speaking, symbols of voice-glides only, though in practice they are used to denote voice-leaps also, whose proper symbols are (r) and (4). 164. The level tone—or an approach to it—may be heard in wel/ as an expression of musing or medita- tion; the rising in questions or doubtful hesitating 70 SYNTHESIS statements, as are you ready ?; the falling in answers, commands, or dogmatic assertions, as in yes, [ am. 165. Besides the simple tones there are compound ones, formed by uniting both in one syllable : compound rising (”) compound falling (*) 166. The compound rise may be heard in such a sentence as take care! when uttered warningly ; the compound fall in of/, oh really! when implying sarcasm. 167. It is also possible to combine three tones in one inflection. Thus we can have (*’), which has the effect of (’) being only more emphatic. 168. All these tones can be varied indefinitely ac- cording to the interval they pass through. As a general rule, the greater the interval, the more marked the character of the tone. For ordinary purposes it is enough to distinguish between a high rise (’) and a low rise (,), the former passing through a less interval than the latter. Conversely a high fall (*) passes through a greater interval than alow fall (,). A high rise may be heard in what? as an expression of mere inquiry (‘o30), a low rise in FO as an expression of surprise. In music semitones have a plaintive effect, and this is to some extent the case in speech also, where, however, plaintiveness is also expressed by modifications of the quality of the voice (177). 169. Besides the separate inflections of which it is made up, each sentence, or sentence-group, has a general pitch or key of its own. Key is marked by INTONATION m1 prefixing the voice-leap symbols in the same way as with the other group-modifiers, thus ro :a]rJo = come up! ina high key. For ordinary purposes it is enough to distinguish three keys: high r middle rt low at the middle being generally left unmarked. 170. The high key is the natural expression of energetic and joyful emotions, the low of sadness and solemnity. 171. Change of key has also a purely logical signifi- cance. Thus questions are naturally uttered in a higher key than answers, and parenthetic clauses in a lower key than those which state the main facts. In all natural speech there is incessant change of key. 172. Changes of key may proceed either by leaps or progressively, Progressive change of key may be expressed by using (’), ete., as group-modifiers. Thus ‘o is heard in all cases of passion rising to a climax. Connexion between Quantity, Force, and Pitch. 173. There is a natural connexion between force, length, and high pitch, and conversely between weak force, shortness, and low pitch. 174. The connexion between force and pitch is especially intimate. All energetic emotions naturally express themselves in high tones and forcible utterance, and increased vehemence of emotion is accompanied by a rise in force and pitch. 72 SYNTHESIS 175. The association of force and quantity is less intimate. There is, however, a natural tendency to pass over the less important unaccented elements of speech, and to dwell on and lengthen the more prominent ones. 176. It is, however, a mistake to suppose that these natural tendencies represent necessities, and that high tone and strong stress can be regarded as convertible terms. Just as on the piano the lowest note in the bass can be struck with the same force as the highest one in the treble, so in language it often happens that strong stress is combined with low pitch. Still less can length be identified with stress, Voice-Quality. 177. Besides the various modifications of stress, etc., the quality of the voice may be modified through whole sentences by various glottal, pharyngeal and oral influences. 178. The influence of the lips is seen in the two qualities of the voice known as ‘clear’ and ‘dull.’ The clear quality is the result of opening the mouth widely and spreading out its corners. When exaggerated it gives a harsh, screaming character to the voice. 179. The dull quality of the voice is the result of slight separation of the jaws and neutral lip-position. English speech generally tends to the dull quality. When exaggerated it gives a ‘muffled’ character to the voice, which, when accompanied by low pitch, results in what is called the ‘sepulchral’ tone. VOICE-QUALITY 73 180, The dull quality of the voice naturally leads to nasality, for the breath, being impeded in its passage through the mouth, seeks another passage through the nose. Slight nasality is almost universal in English speech. Its presence is at once made manifest in singing. 181. Narrowing of the bronchial passages gives a wheezy character to the voice, sometimes approaching to strangulation. This effect is familiarly known as ‘the pig’s whistle.’ It may be heard from Scotchmen, and combined with high key gives the pronunciation of the Saxon Germans its peculiarly harsh character. 182. These modifications—which are the result of controllable organic positions—must be carefully dis- tinguished from those which are due to peculiarities of the organs of speech themselves. Thus defects in the palate may cause permanent nasality (together with a peculiar hollowness of sound), an abnormally large tongue gutturality, ete. All these peculiarities are inseparable from the individual, while those first described may—and often do—characterize the speech of whole communities. 183. Voice-quality may be readily symbolized by prefixing modifiers : cvl = clear quality. oy = dull quality. os = nasality. 00+ = wheeziness. Oo = gutturality. SWEET F 74 SYNTHESIS Organic Basis. 184. Every language has certain general tendencies which control its organic movements and positions, constituting its organic basis or basis of articulation. A knowledge of the organic basis is a great help in acquiring the pronunciation of a language. 185. In English we flatten and lower the tongue, hollow the front of it, and draw it back from the teeth, keeping the lips as much as possible in a neutral position. The flattening of the tongue widens our vowels, its lowering makes the second elements of our diphthongs indistinct, front-hollowing gives a dull resonance which is particularly noticeable in our /, its retraction is unfavourable to the formation of teeth- sounds, and favours the development of mixed vowels, while the neutrality of the lips eliminates front-round vowels. Our neutral tongue-position is the low- mixed or mid-mixed one of the vowels in further >Tw,. 186. In French everything is reversed. The tongue is arched and raised and advanced as much as possible, and the lips articulate with energy. French therefore favours narrowness both in vowels and consonants, its point-consonants tend to dentality, and, compared with the English ones, have a front-modified character, Which is most noticeable in the 7, while the rounded vowels are very distinct. 187. The German basis is a compromise between the ORGANIC BASIS 75 English and French, standard North German approach- ing more to the French. 188. No language, however, carries out the tenden- cies of its basis with perfect consistency. Thus in English we have the point-teeth v; and mixed vowels occur in French and German, ENGLISH SOUNDS VOWELS. 189. The following is the English vowel-table, weak vowels (those which occur only unstressed) being marked by a preceding -, and the half-long quantity of the first elements of diphthongs not being marked. {(m), -f+ (1) We e(ee) | [(t-) l I* t 1/33) -Tr+ $2), Fee | -dee(3ee) 190. The following list shows the correspondence of the Broad Romic symbols, with examples: e = J+, ] ase come. aa = | » tather. Ai == ie oa ng au = Jt er MS @ I T » man. VOWELS a4 e =f as m men. el = fr i say. Geese tiene together. eo = ji 3 bird. eo = 71 ij care. Vt yd 1 ke Se fill, pity. Waid Fe sea. O = f, Fr 9 not. ol = jr J boy. ou = 4a) 5 SO. wes Tt ms fall, sen EAT. ey full, value. uw = ty 4 too. We will now consider the vowels more in detail, noting varieties of pronunciation. 191. J+, mid-back-wide-out; }+, mid-back-narrow- outer. Q)Hr Jo, a}r Jo come wp. The former pro- nunciation seems now to be the more usual. Full back ], and, apparently, low-back-narrow J may be heard in the dialects. In Irish- and American-English in- mixed vowels appear to be substituted for it. The first element of the diphthong in igh is some- times retracted towards ], especially in Cockney Eng- lish, where it is often lowered to j: refined o}re,, Cockney jr-, jrr. Before 7 the glide is obscured to t in Cockney English, so that ryww mile is hardly distinguishable from mar/e, 192. |, mid-back-wide. >Jw, father, farther, ola> half. In the formation of this vowel the tongue gradually relaxes into the neutral mixed position, so that it 78 ENGLISH might be written Jv. In Cockney English it is lowered to jt, which is often narrowed into jt, as in pforsl- pyj+a Battersea Park. 198. |, mid-mixed-wide. olafv|, together. It is difficult to define the formation of this vowel with precision, because it often degenerates into a mere voice-glide— 1—without fixed configuration, which is often whispered in protonic syllables. The most correct way of writing such a word as ¢ogether in rapid speech would be, therefore, onafwvi. The exact position of this vowel—as far as it is capable of being defined—is probably between mid and low: \{-. 194. J, low-mixed-wide. The first element of the diphthong in how e]%, which in Cockney English becomes [—2T3, [t- Scotch has J]—- 2}: —and American- English J—e]. 195. J+, low-mixed-narrow. jt evr, py Jird. In American-English this sound seems to pass into an in-mixed vowel. In some American pronunciations, especially that of New York, J+ becomes |i, as in p10. 196. f, high-front-wide. This is the sound of strong (stressed) 7 in it [0, fil/ >Lw, the corresponding weak sound f[- being intermediate between f and [: pfol- pity. After © preceded by a consonant f is often retracted to the in-front position, as in poluol+ pretty. The long vowel corresponding to f is fm or fr, as in Sf sea, s{ms cease; an [+ modified by gradual raising of the tongue, In Cockney English the first half of this diphthong is lowered, so that sf+r often sounds VOWELS 79 like sfer say. Before r, as in fear, fearing >fr, >fwl-a, there is no raising of the tongue, and the f itself is often lowered—>f-1. In Scotch, Irish and American- English, as well as in North English educated speech, sea, etc., keep the old long narrow undiphthongic f‘—sIs. Scotch also has it before 7, as in >f+os. In Scotch strong f[ is lowered, as in sf+d si/, sounding to an English ear like set, which in Scotch has { (197). 197. [, mid-front-wide. Ff men; sft- say, Oftra take. In say, etc., the tongue begins to rise as soon as the [- position is assumed, and goes on rising gradually till the f~-position is reached. American-English has the same sound, sometimes raising the tongue to the full high position—sfr. Scotch keeps the old monophthongic narrow vowel—s[t—which in Edinburgh is raised to- wards [— si{[+—sounding to an English ear almost like see. In the North of England the first element of the diphthong seems to be narrow—s[r-. In Cockney Eng- lish the first element is broadened into }, or J+, so that say is confounded with sigh, except when the latter is broadened into sjrx (191). In North English and Scotch the short vowel in men is the broad f. 198. 1, low-front-narrow. Only before 7, as in att care, where Scotch has (¢. 199. J, low-front-wide. pja ack, rt3 man. This vowel is often modified slightly in the direction either of j or of f, into which latter it often passes completely in Cockney, as in a[p, alt cad, 200. t, high-back-wide-round. pta dook, >tw full ; Ol two, too, Itz new. In North English, Scotch, and 80 ENGLISH in Irish and American-English the old narrow tis pre- served in ol, 3mt4. Scotch has also narrow short t in pta. In Cockney English tg often becomes ts with the out-back vowel, especially after m, as in mts you, the being dropped in Cockney English after a con- sonant, as in St», Thr. Unstressed t+ is regular in educated as well as vulgar speech, as in stowtrs Zulu (where the second 9 is very indistinct), >jwmt value. Monophthongic t is kept before 7, as in ptt poor, Amt. cure, where Scotch has t#—ptios. In Cock- ney English tv is broadened to 74(1), so that poor is levelled under pour. Even in educated speech there is a tendency to lower t before 7: the strong form of your, yours is sometimes m}+1, m}+18 with the out- back vowels which otherwise occur only in weak forms, but often my1, +8 with the full Cockney form. 201. ja), mid-back-wide-round. s}3) so, sow. The second element of this diphthong is formed by a gradual narrowing of the lip-opening to the t-position, which begins almost as soon as the } itself, the position of the tongue remaining unchanged. In weak syllables, as in the second one of s}s)w}13+) soo, the tongue is advanced to the mixed position, the diphthongic char- acter of the sound becoming almost imperceptible. This change often takes place in stressed syllables as well, where it is often used in polite or conciliatory address; thus } 3}s) and Jr++) 3}rd++) = 04 no may be heard from the same speaker, the former in more decided and dogmatic statements. The constant use of }+ gives a character of effeminacy or affectation to VOWELS 81 the pronunciation. In Cockney English the first element is unrounded and the second lowered and partially unrounded—jt Jt. Scotch keeps the old monophthong }+ —s}+. The first element of the diphthong }s- in doy is sometimes lowered to 7 as in wot. 202. 34, low-back-narrow-round. sy saw, 3710 naught. In the formation of this vowel the tongue is gradually relaxed to the mixed position, the rounding being maintained, so that it might be written Jt, aS in St saw distinct from sjv soar. In Scotch this vowel is generally made into }. 203. 3, low-back-wide-round. 3JO not. Scotch sub- stitutes }—7}o0. American-English has sometimes +— which seems often to verge on the out $+—sometimes the unrounded J of /ather—s]o not, >Jeo}r follow. Weak short 0 becomes J+, as in xra'o}ap), October. 204, Although weak vowels are generally levelled under ', or f+ or the rarer t+ and $+, the vowels and diphthongs J, J, [, 1, etc., can all occur in perfectly unstressed syllables, as in [3s]wo, azrpy+30, [3sfao, TOSOOTAS wmsult, compound, insect, abstract (adj.) (noun). But if we compare the strong and weak vowels in these words, we shall find that the weak vowels all show a slight obscuration of sound, though in most cases hardly enough to justify a change of symbol. But the weak diphthongs (ai) and (au) might almost be written |t-, lt. Weak f- itself in very rapid speech seems to tend to the mixed position, especially after ® preceded by a consonant. 82 ENGLISH CONSONANTS. 205. The English consonants are as follows: 206. The Broad Romic equivalents are: = asin bee. 3 day. »» » then, eee 1 ek, » house, sah sgt OUD 5 come. » Look. 2s OAT of 4 TOs » sing. » pay. » red. 9», SAY. »» ship. a a » thin. (dh) = rl ie 2 Eine ae aon Hid au u ud ued fCaxvnnéegurdadwAaeeraga re QvVvVE das — ™ =F — I SS ct &eemH to SoS l (th) CONSONANTS 83 Vv =). >. 23 12VIEW, Ww a A ae: we. wh S55 i ese. tn EL, Z Se Le zeal. 5 = @ i rouge. 207. The point-consonants w, 0, 0, 3 seem to be blade-consonants—ws, 05, Os, 35; Os being equivalent to SI, as we see in comparing the tongue position in so and toe. 208. The voiceless stops are all breath-glide stops —ar]r, ete. (126 a). The buzzes vy, 8, e,> are often whispered (137)—Ww, ete. We will now consider some of the consonants more in detail. 209. 2, aspirate. ets house, elm he, ote who. 2 is dropped—that is to say, its stress is shifted on to the vowel (118)—in Cockney English and in most of the English dialects, being always kept in Scotch, Irish, and American-English. In natural speech it is always dropped in weak syllables when not initial, as in -}r sj -[r 1 saw him, where, however, the dread of vul- garism often leads to its insertion, especially in ‘ ladies’ English.’ 210. o, front-open-voice. ots you, rLom],3 million, In such words as Aue, humor the aspirate precedes the m as if it were a vowel—eots, entor, but in Scotch the combination gm is made into 9—ol+. The nearest approach to this sound in Southern English is in such words as pure DO)m)11 (144). 84 ENGLISH 211. ©, point-open-voice. ofo ved, olwla rearing. This consonant is practically a vowel, there being no buzz in it, even when emphasized and lengthened. It never occurs except before a vowel, being elsewhere weakened into a voice-glide, as in oft, ofto rear, reared. It is sometimes rounded into ©). Trilling— rolling one’s 7's’—is a defect in pronunciation, which is, how- ever, often affected on the stage and in recitation. The substitution of the back open e, which is sometimes trilled—es—is a frequent individual peculiarity. esis the ‘Northumbrian burr.’ In Scotch 7 is a strong trill everywhere: osftws, wsfoso. So also often in Ireland, where also sounds similar to the Norwegian ‘thick ¢’ (147) may be heard. In American-English r before a pause or a consonant appears as a point- modification of the preceding vowel, as in >] far, being completely lost in other American pronunciations. @ is always formed in the inner position (75), and in the dialects of the West of England o: is exaggerated into Wc (71 ¢). 212. VU, wv, point-teeth-(voice). uls thin, vfs then. In these consonants it is enough that the breath is directed on to the teeth by the tip of the tongue, which need not itself be brought against the teeth. Certainly the most distinct form of these consonants is that produced by placing the tip of the tongue firmly on the back of the upper teeth and forcing the breath partly between the interstices of the teeth, partly between the sides of the tongue-tip and the surface of the teeth ; but they can be—and are often—formed by bringing the tongue CONSONANTS 85 against the gums in the s-place without touching the teeth. In the latter formation the contact is of course very slight. Weak v, vare formed without any contact. Trish-English substitutes its peculiar ‘fan-stops’ (71 ¢) for v and vy, in whose formation the point of the tongue is spread out like a fan, so the whole of its rim is brought against the teeth or gums together with the point, the back of the tongue being slightly raised at the same time: of, os. S, 8, blade-open. sl see, slow zeal. 213. 2, 2, blade-point-open. 2fp ship, wtse rouge. These consonants are formed more inwards than s (75), but after the point-blade consonants 05, 0 they are less retracted, as in a[d2 Dez catch John. If we drop the 0 in ajdz we can feel the difference between the resulting ajz+ and ajz2=cash: 2+ has a sound intermediate between 2 and s. 214. 9, 9, lip-back-open. ofr why, sfm we. These consonants are practically tsa, ti, being wide (66). In Southern English » generally becomes x, but it seems probable that G will be completely restored in a few generations. 215. >, >, lip-teeth-open. >jw fall, >mts view. 216. w, point-divided-voice. wtaJlook. In this con- sonant the English concavity of the fore part of the tongue (185) is especially noticeable. In Cockney English and in Scotch / has a still deeper tone, which may be sometimes the result of back modification—oc. 217. a, a, back-stop. ajr come, a}s) go. 218. 0, 0, point-stop (blade-stop). oft ten, ofe day. 86 ENGLISH 219. b, bp, lip-stop. pfr pay, plo bee. 220. a, back-nasal-voice. sla sing. 221. 5, point-nasal-voice. 3} xo. 222. F,lip-nasal-voice. FTI man. >I5, lip-teeth-nasal- voice in zymph (142). : SYNTHESIS. 223. For the English synthesis and organic basis see §§ 98, 99, 104, 185 and the section on Synthesis generally. SPECIMENS. Of the three following texts the first two are accom- panied by a Broad Romic transcription, the ordinary spelling of the first being also added. 224. :olmpw ‘ots -0}, :vlaa -vl+ ‘pr -2]s -], :a}ee30 -|> >wTo ‘apra’, -al-w -v, ‘slo -f+w wp1ID -[-0°; -pDd -of-e 3} 57% -v]o -L-0s ‘oliwl- 07350, :w}rra -} ‘PHW— IFO VAB}H+L0 :oTtIO’, -o0 -], :wlow ‘>wTOIO, ‘cojrra -|3 ‘yol-se. 225. :pijpl ‘juws -te :pink -S1 ‘oop -woez -o :kaind -ev fleet ‘keik*, -wi6 -So ‘sij -0l ‘raund -it‘; -boet -wij ‘nou nau -Sot -its ‘rioli ;raund’, :laik -a ‘bol‘—:not Y-kwait ‘raund’, bot -o':litl -fleetnd’, :laik -on ‘orin;.. 226. People used to think the earth was a kind of flat cake, with the sea all round it; but we know now that it is really round, like a ball—not quite round, but a little flattened, like an orange. 227. -|3 ‘Laawl-ze]3 -9]s ‘J5s ‘owyswl+s -L-5 ‘Depre3// -et+9 ‘atTIO :splod :o2)ur5los. -a)3 -ope-Lr- SPECIMENS 87 “9/3 ‘Ojeetf+a -lo :sjr ‘Lool-aeyis’, -Y5 -v] -ofro} ‘pozto -[+F -] ‘Flo ‘p}ee’. -]s -Ler -0]s dfol- -ejaawl-’, -2ftr [Oo -f+0 ‘Jn’, -|3 ‘vpo -L-0 ->fol- ‘ato, -pvo -Ly- ‘ALDI :F [LA *[tO" 070 -L- -2]3 “Ffro -gHS. -sJHa of -V), sof ro), :afer -0}, :awfw Yofe’, -2ft- p}rtol-0 -|o “Vi- -[rol- -pjrrolz, -|3 ‘astao -wlra -] oja. -v} ‘a[f0o -|o -s)3s :zta -[+s ‘2fo’, -|3 ‘pad -wlra -), oye. = -|3 ‘s}p -vl- :Laawl-2r]3 tots -Le-o -pfers Ojrer3lya -FH+I ;Opasp>wfe2’. 228. -on ‘inglifmon -wez ‘wens ‘treevlin -in ‘tfain’’ -huw ‘kudnt :spijk :tfai-nijz’. ‘wen ‘dei -ij -woz -dainin -ob :svm ‘ijtin:haus’, -on -Se ‘weito ‘brot -im -o -mijt ‘pai’, -0z -ij -woz ‘veri ‘hengri’, -hij ‘et -it ‘ep’, -on -pot -it’ verl ‘gud’, -bot -ij ‘kudnt :meik ‘aut’ :whot -it -woz ‘meid -ov*. -sou :when -%o :weita :keim to :klior o-wei’, -hij ‘pointid -ot -8i ‘emti ‘pai:dif’, -on ‘kweekt -laik -o ‘dek’. -Se ‘weiter -et ‘wens :fuk -iz ‘hed’, -on ‘baakt -laik -o ‘dog’. -on ‘sou -%Si :inglifmon ‘njuw -ijd -bijn ‘dainin -on ;dogzfle/*. ‘p)tso -|'3 ‘polsto.. 229. -|c ol-rlen}, -lc ol-e(ee -V], ‘2]%S -nTW -Le -2]s “pz, -V, ‘wlow ‘alto}aHy opr -v] ‘s]5 ‘aftr ‘plopol-a [3 -]o ey; -2f-r -5[5] :afre -] jeflsa :ots ‘ste’, -3J* ‘pop cots ‘wys -] ‘ole; “DO "St -]e Jos ‘ale -v] -3]ero -|O ‘py+3 -F le ‘poy |ofe. 88 230. 231. 232. ENGLISH -Ye ofsefeo], -le ols efee -v)|, ‘ofpsl+s, -ofo -|3 ‘oro, -v|, ‘>)jer10f-os -|5 -v], ‘wlol-ajos— “VJHS DWIS :F[LD -]> ‘w}rLo! -V|, ‘w}rro]la -op -v), ‘opel+5s ‘plwo, -|3 cop -F'Le ‘polv), ‘slo’ Vv, ol pyres le -ye3 -[+8 ‘pp evofe— -v), ‘ool -f+s wlsflra ‘ofo! -(e ol-rfen}, -le ol-efen} -D[W -|£ -2]8 :otes -ol ‘soly’, -|3 vp -vl- ‘qu -F|so -oje -|s >ofz -O|, ‘Sazojrans -p+3 -v] ols; -F et ‘sofol-0 Dwts -[+3 >fv]s :vfs, vo :fs -s}3) -ofaf+ +379, -|3 ‘s)F|, ‘pisws -ao -sapswl- ‘atow -V|, >Lor|o -fH3 -F]L poy. -Ye olvefeo], -Ye ol-rfee} -v], >proolr-s ‘o}ia -]|3 -o}ur; -[£ ‘otss -o], ‘visa -vp ‘swfs0], -ozps -2, awjns |:afsso -v), ‘sa}ue: -[-0 -3|8 -] -d2)rrool-2 -Laso]ss\, -prno ‘Sy% -Ol-s ‘wlow ‘oetr -O|, ‘S$ -]er Dpw]o jo ->0]r ‘eps v3 :of -[e -2]s -] ‘pf. FRENCH SOUNDS 233. VOWELS. 234, 4 = jr asm patte. an = js 4 sang. 2 hhes Fig » ©, pate. De hsaot ep ae Pere: e = [ 598, été, en = TT » vin. 9 were sot » peu. on = fF a un. A. sche k @ fini. 0) f= htehs. worse beat 0 thins son. SWEET G go FRENCH D Cea as eer: 02 Vite pe Ranh, Mh eter pti " sou. yi tier eee ee ine 235. For the French organic basis see § 186. 236. The quantity of French vowels is often undeter- mined, half-length predominating. Final vowels are short, as in fini, son; except in exclamations, such as jt ak. Vowels are lengthened before final 7 and open voiced consonants generally. Nasal vowels are short finally, long before another consonant, as in $s}5 son, >J+5Or fonte. 237. jr, low-back-wide-out. o'y+rOr patte, o'yHter page. This is a vowel intermediate both in formation and acoustic effect to the English ] in part and [ in pat. 238. J, low-back-wide. b'y pas, p'ytOr pdte. This sound is easily obtained by unrounding the English Fin pot, taking care not to muffle the sound. 239. ys, low-back-wide-nasal. sys sang, jue ange. For French nasality see § 45. 240. [, high-front-narrow. >I[sf fini. 241. [, mid-front-narrow. [of é¢é. 242, 1, low-front-narrow. pyre: pére. Insome pro- nunciations this vowel appears to be wide—f. 243. {5, low-front-narrow-nasal. >]5 vm. May be wide—j. Often retracted—t4, or even jr. 244. i, high-back-narrow-round. st sou. Often advanced tr. 245. }, mid-back-narrow-round. pp} deau. 246. $, mid-back-wide-round. je: or. Often ad- CONSONANTS gt vanced nearly to the out position $+. Sometimes pronounced as a lowered }. 247. js, mid-back-wide-round-nasal. $45 son. 248. f, high-front-narrow-round. pfte: pur. 249. {, mid-front-narrow-round. p'f peu. 250. {,low-front-narrow-round. p'qttei peur. Some- times wide f. The weak vowel in /e seems to be a slightly retracted and partially unrounded f = fw. 251. ~/, low-front-narrow-round-nasal. yun. May be wide—yp. 252. CONSONANTS, 2 |c1,€4] 0,0 St, St | Zr, 2+ 19\,9\] 9, 9/ >, > 253. b = ) asin bon. B = 9. & buis. ¢ Si: CHLOE pied. d — ML) a doux. f = o> 7 fin. g = @ “ gout. h = @ j =e ¥ briller, k = 4 “ qui. 92 FRENCH ] = wo asm belle. lh a J table. TO ier ob . mon. n = 5 : non. fi = & 9 agneau. p =) 4D bs Paris. ) Soi Oy ages puits. r wie if rare. Thuvesgae is quatre. s — yi 3 sou. if =e * chat. t to " tout. Vv = > 5 vin. Ww a ee Mr oul. AN =i) i; poids. Z aes a zele 3 ma Cit 4 jour. 254, 2,aspirate. This sound is formed only occasion- ally and involuntarily as a hiatus-filler, as in >w[2} fléau, wy+2} la-haut. 255. ci, €4, back-open-inner. eiyH+e17are. This, which is only an individual peculiarity in English, is the regular sound of 7 in French. It is generally slightly trilled. The point ws sometimes occurs, and is always trilled. At the end of a word after a consonant 7 is unvoiced, and often dropped entirely, as in QJ#OC:, AjHo. After a breath-consonant—especially a breath-stop— it is sometimes fully, sometimes partially unvoiced, as in pct, ac)e){ pres. These remarks apply also to a, 9\, and 2. CONSONANTS 93 256. 0,0, front-open. pelof driller, nol pied. Final ® unvoices its latter half, as in >L[o)o), >Los fille. In connexion with the vowel [, as in driller and fille, o is a consonantal r, but after more open vowels it appears to be lowered to the [-position and to be retracted somewhat. 257. St, S, blade-open-outer. st sou, stu zéle. These consonants sound clearer than in English, partly because of the greater convexity of the tongue in French, partly because they are formed with the tongue nearer the teeth than in English. 258. 2:, er, blade-point-outer. zy chat, elves jour. The same remarks apply to these consonants as to $ and 8. 259. 9\, 9), lip-front-open. pol duis, pol puits. In these words we have a consonantal f, the tongue being lowered before other than high vowels, as with o. 260. 9,9, lip-back-open. sf oui, ony poids. In the former of these words we have a consonantal t, in the latter rather a consonantal }), and so before other open vowels. 261. >, >, lip-teeth-open. > [5 fix, >[5 vin. 262. t+, w+, point-divided-teeth. pw delle, oynKL. The unvoicing of this consonant is parallel to that of €. The point consonants w, 0,0, 3 are all formed against the teeth, the fore part of the tongue not being hollowed as in English (185), so that the French / has a palatal effect to an English ear. 263. 0,4, back-stop. af gui,atgo#t. The voiceless stops Q, 0, pb are followed by voice-glides, there being 94 FRENCH no escape of breath asin English (126 a); hence gut often suggests (gi) toan English ear. Initial 4, 0, p are pronounced with full vocality (126d), which suggests (n), etc., to an English ear. Final voice-consonants often end in a voice-glide, as in pyrta' Jague. Even with voiceless consonants this is sometimes the case. Or, Dr, point-stop-teeth. o'k tout, ot dou. 264. b, b, lip-stop. p'yreif Paris, ps Con. 265. «,front-nasal-voice. +L} agneau,ptwjL Boulogne. Like » this consonant ends voicelessly, as in [xs vigne. In vulgar speech it is retracted nearly to the 4+-posi- tion of the English 2g in sizg. It is sometimes formed with imperfect stoppage, giving of. 266. 3+, point-teeth-nasal-voice. %3}s mon. 267. F, lip-nasal voice. Fjs mon. 268. Double consonants occur only in learned and foreign words such as immense [rrjiss, except when they are the result of contraction, as in mourrons. 269. Stops formed in different parts of the mouth are joined with glides, as in Jra’d acte. SYNTHESIS. 270. French synthesis is very rudimentary as re- gards quantity and stress. Frenchmen are unable without long training to distinguish vowel-quantity, stress, and syllable division in foreign languages. 271. In French every syllable is uttered with almost even stress, which is always open (159), so that such a word as capacité is divided ‘a'yrp'y'sfor. SYNTHESIS _ 9s 272. In French there is no such thing as word-stress or word-division. Sentences are cut up into syllables without any regard to the structure of the words they are made up of. Thus the sentence quel dye a-t-il ? is pronounced (kz lev Za ti). 273. Although stress is nearly even, there is a distinct tendency to weaken the stress of the last syllable of a syllable-group, whether it consists of one or more words. Thisis regularly the case when the last syllable is uttered with a falling tone, when it is often pro- nounced with breath instead of voice, as in 3f pys"s45 pots n'y pensons plus. So also the word-group ‘wywofs la lune takes the stress on the first syllable when pronounced with a falling tone, while in /e soleil wf's}wtm: it falls not on the article but on the first syllable of soled, because that syllable happens to be the last but one. In the French pronunciation of home, sweet home! as :}re1szf:0-'J+r we can observe the three main features of French synthesis—want of determinate quality, monotonous syllable-division, and illogical stress. 274. Frenchmen, in fact, generally have no idea of where they put the stress. But they have a tendency to stress intensitive words, as in ‘Oc{ pats trés-bien, ‘a'tw Sfowls quel supplice! and often (though not always) mark antithesis by stress, as in ce n’est pas lui, c’est moi, where /u2 and moz have an extra stress. 275. Stress in French is, as we see, greatly dependent on intonation, which is the most important factor in French synthesis. French intonation goes more in leaps 96 FRENCH than by glides, and the intervals are considerable, which together give it a peculiarly lively character. Isolated words are pronounced with a rise on the last syllable, as in ajroy+sf"o[. In a sentence they keep this high tone when emphatic. Doubt, question, etc, are ex- pressed by a rise, as in English, as also lively statement, a falling tone being frequent in statements, answers, commands, ete. High tone is often accompanied— though not necessarily—by increased stress. In French there is a tendency to alternate high and low tone in the same way as other languages alternate strong and weak stress, so that a low tone in a subordinate syllable often appears as a kind ef preparation for an emphatic high tone. Thus in the sentence vous voyez done, messieurs, que c'est ainsi, the highest tone is on the first syllable of ainsi, the lowest tone being on the one immediately before it,—c’est. The next prominent syllable before c’est is the last of messieurs, which ac- cordingly has a high tone—though not so high as that on the first syllable of ainsi. The first word in the sentence—vous—has a tone of about the same height as the second syllable of messtewrs, from which the tone descends by short leaps to done mes- whence it leaps up to -siewrs: ‘>t sayrol ‘ofs “e['smf af ‘sto ‘psls. SPECIMENS. 276. pyrewl >t >eysst? yok. eys sf efso pusl pte rf Ye apspeywoc. f wf pyrew atesrrys. Lo fael 02 DEJIST a}r sy pe}pef wyya. ato pe tol? fw [ Ocays pe [ SPECIMENS 97 Orl. syra[ >t sf alo pe 45 Ofejr S}5[? atwl[ sts Jroysoo PT [Sys ? | 277. parle vu fransee? on po. gan se Zyst ase pur moe feexer konpraandrh. i lce parl kuraman. il ekri ] fransee kom sa propre laang. keel cer eti? il x trwez cecere dmi. save vu si Pit cecer on deza sone ? vule vuz ataandr onn enstan ? 278. Parlez-vous francais? Un peu. J’en sais juste assez pour me faire comprendre. II] le parle couram- ment. I] écrit le frangais comme sa propre langue. Quelle heure est-i1? I] est trois heures et demie. Savez-vous si huit heures ont déji sonné? Voulez- vous attendre un instant ? 279. wf ryreal 3 [ot pteoys pys ps3 Fe Of e[5l. [w [OT SyH>J5, FT SyH>TS STS SO[SoTrolOl, Jr FATS a Js 5 STM jrow[ [sl £3 acy jrolwlol pte steops beye spss folwlo[ fafs, oF5 sis Je}ss Jus[ stays wmf O pjrew[ pwf ore, [ aL proto Jeosfen[ efsayr wy pytsmgs, efsayr cost FEStETHIL wl OL O[etToTes Jr3[ 0 shs5 [aslsoyess. 280. loe marki n etz purtay pez onn om dee zeni. il etze savan, me savan san spesjalite,a mweyn k on n vee} aple ensi yn graand abilite pur sertenz uvraaz saanz ytilite okyn, don nuz oronz ase suvan ljo d parle ply taar, e ki aveet apsorbe zyska la pevsjon, zyska la monomani le di dernjerz ane d sonn egzistaans. 281. Le marquis n’était pourtant pas un homme de génie. I] était savant, mais savant sans spécialité, a moins qu’on ne veuille appeler ainsi une grande habilité pour certains ouvrages sans utilité aucune, dont nous 98 FRENCH, aurons assez souvent lieu de parler plus tard, et qui avaient absorbé jusqu’a la passion, jusqu’a la monoma- nie, les dix derniéres années de son existence. wf pels}5of. 282. ayro°ol> } efspute of rye YU areal z2yreel st s[ ope Ost : ef >t esayrs ysra}e ayrs}s [srl o[s [stte, fefsotw af wrsp[eytss sof efsays s[ peftwys aowlry. sys Oto >t alof wy >dejtss: of Fs plol 3f & pyrewl >t py? Ofpo\l ocays js @& st afsefie of F jrofeol ps stosiee of >yrojfs t ry al Josafre Sf BIEST DO Ys ts jrasite. } o[otte & £3 } of earls Jt >w} pfte st 0 deT wloy, sts yra{ af 3}0 2}els: Of S dprops Sf F pjrewfl st py? 283. NORTH GERMAN VOWELS. ai 8 i sil s-sv Sn ht see. thrane. gerettet. schon. bin. biene. 100 GERMAN o = }$ asm _ sgonne. oo = } 5 sohn. Oy has » = hauser. Op mef » gotter. 3 ee ugeed chile os User kt oipuerute Vv) ciee ae » schtitzen. Vy aot wee erin, 285. ], mid-back-wide. ¢]3 mann, 3}ir| name. In some parts of North Germany (Hanover, etc.) }+, J+. In Middle and South Germany (Saxony, Bavaria, etc.) J.j¢ or J, Jt The low-back-wide j is usual in North German in the diphthong je, as in e738. Observe that the second elements of the diphthongs je, J3, $e are generally lowered even more than in English, being apparently sometimes narrow, sometimes wide. 286. |, mid-mixed-narrow. Only in unstressed sy1- lables, as in al‘e[olo. In rapid speech this vowel loses its definite configuration, and becomes a mere voice-glide, which is perhaps wide. It varies in differ- ent parts of Germany. In South German it becomes [. 287. |, mid-mixed-wide. Occurs only in the diph- thong |¢, as in F\¢3 mem, which has almost the same sound as in English mine. In Hanover and elsewhere it is broadened into jc. 288. [+, high-front-narrow. f+3] diene. 289. f, high-front-wide. pl dm. Tends to fin some parts of North Germany. 290. [+, mid-front-narrow. [+ see. Sometimes diphthongic—[e-. VOWELS 101 291. {, mid-front-wide. >[so fest, de[#3] thrdne. In many parts of North Germany long @ is narrowed to (+. 292. t+. high-back-narrow-round. alo gut, 298. t, high-back-wide-round. et30 hund. 294. }+, mid-back-narrow-round. $}+3 sohm, Some- times diphthongic—}s-. 295. }, mid-back-wide-round. s}5] soune. }in South Germany. The diphthong ew, du, is generally }+ with slight rounding of the second element, but often je; practically identical with the English ov. 296. f-, mid-front-narrow with high rounding. aef5 griin. Might be written f). In all the German round-vowels the rounding is a degree higher than the tongue-position. Pure f+ is heard in the Middle and South German artificial pronunciation of long #, which generally becomes [+ in Middle and South German. Even in North German the tongue lowering —which seems often to be partly retraction—is less in some pronunciations than in others. 297. f-, mid-front-wide with high rounding. 2f0s5 schiitzen. Becomes f in Middle and South German. 298. {+-, low-front-narrow with mid rounding. 2){+5 schén. Often wide in North German. Becomes [+ in Middle and South German. 299. f+, low-front-wide with mid rounding. dafole gotter. 300. Initial vowels have the clear beginning (117 4), as in -']e3] Jwo] "col eine alte erche. 102 301. qe Ae SB BS a ea oe ee ee CONSONANTS. GERMAN as im bin. ich. du. voll. gut. sage. hat. ja. komm. lang. mann. nun. lang lieb. retter. was. fisch, tun. was. CONSONANTS 103 Wu) (aseso as im quelle. Spy eI Cty COURS: ach, auch. Z = $ 9 sO. Se hye) » courage. 808. 2, aspirate. eo hat. 804. c,back-open. Jc ach. After back-round vowels ¢ is rounded into c, as in y3c auch. 305. €, back-open-voice. g between vowels is stopped in some pronunciations (Hanover, South Germany), open in others, especially in Middle Germany, s)a], sre]. The latter is considered the more correct in ordinary conversation. Final e = g becomes c, as in Otc zag, which in other pronunciations appears as oj}a. The Hanoverian pronunciation is o}sa], otc, which is the easiest compromise for English people. 306. €:, back-open-voice-inner. e:[o]e: retter. This is a very soft, vocalic sound, which is not generally trilled. The point os is provincial. 307. 9, front-open. [0 ch, 308. , front-open-voice. o'}+ ja. 309. S, 8, blade-open. >]s was, 3}+ so. Initial s is often formed with half-voice (135 c). 310. 2), 2), blade-point-lip. >f2) fisch, at-e}+e)] courage. Differ from the corresponding English sounds in being always rounded, which allows the tongue-position to be somewhat relaxed. 311. 9, lip-open-voice. asfw] quelle. Often pro- nounced as a weak >. 312. >, >, lip-teeth. >} voll, >]s was. These sounds are weak, and are formed with very little buzz. In 104 GERMAN Middle and South German > often becomes pure lip 9. 313. w+, point-divided-voice. wa Jang. All the point-consonants approximate to dentality, the point of the tongue being often partly on the gums, partly on the teeth, the fore part of the tongue being more convex than in English. 314. a, 4, back-stop. ate tomm, aio gut. Voiceless stops are generally followed by a strong breath-glide (128) in North German. In Middle and South Germany they are followed by a voice-glide—a', etc.—and the voice-stops are pronounced in the same way, so that the distinction between / and g, ¢ and d, p and 4 is completely lost. 315. Or, Or, point-stop-outer. o°t3 ¢hum, Olt du. 316. b, b, lip-stop. wltp Ziel, pfs bin. The lip-teeth stop occurs in the combination pf, as in pfund. 317. a, back-nasal-voice. wa dang. 318. 3+, point-nasal-voice-outer. Th3 nua. 319. F, lip-nasal-voice. FJ3 mann. SYNTHESIS. 320. In the combination Am there is often a breath- glide between, the consonants, as in ast Anze. 321. Final stops are always voiceless, as in ied, hund, In Middle and South Germany all voiced consonants— even 7 and /—are imperfectly voiced (whispered #), so that they are hardly audible except in the glide to or from a vowel. 322. Long vowels are not shortened before voiceless SPECIMENS 105 consonants, as in English (98). Long vowels are shortened to half-long or even short under weak stress. Final consonants are always short (99). 323. Vowel-like consonants are often lengthened before breath as well as voice-stops, as in 20 halt. 324. For stress see 104. 825. German intonation closely resembles English. SPECIMENS. 326. :'}+ ‘s[ssot, >fwles ‘e}3072)] 03! -ostr ;wlosos FJ} -J3> :F cI] ‘oles, :O[3 Lo :s}+ FIG] -eloles]co -]3 :oltsr ‘piwo eferjsalsjco! -0)s+ -fple: ‘pfio less -150 p) plies, :ceifto's[tom]e1 >e}eI0! fer2,le380 ote rafter. “Jc, afs0 -fo :o}c -ya> ;pferalsefes -L5 :o]es le ‘wltor -wloo] alts, -te ‘pJea]sefiw] -rlo -alesoless ‘2)ator, -j3> lisa -[5 -o]es]e -olrle: ->[tor, ->}5 “jolr plstsaajor (so-w}os -L5 -oles]|r ‘073 alst50 -Flo ‘p}to5. 327. :00 ‘zcestu, ‘folor ‘moondnfain! -tsum ;letstn :maal -auf :maino ‘pain, :deen :i¢ :zoo ‘mance ‘miternaxt -an :diizm ‘pult her‘ange:vaxt! ‘dan :yybor ‘byycern -unt pa'piir, :tryyp‘zeeljor ‘froynt! er‘fiinst :duu :miir. ‘ax, ‘koent -i¢ :dox -auf ;bergeshoon -in :dainom ‘liibm ‘ligte ‘geen, -um ‘bergoshoolo -mit ‘gaistorn ‘[veebm, -auf ;viizn -in :dainom ‘demor ‘veebm, -fon ‘alom ;visnskvalm ent'laadn -in ‘dainom ‘tau go‘zunt -mig ‘baadn. 328. O sahst du, voller mondenschein! zum letzten- mal auf meine pein, den ich so manche mitternacht an diesem pult herangewacht! dann iiber biichern und papier, triibsel’ger freund! erschienst du mir. Ach! SWEET H 106 GERMAN konnt’ ich doch auf bergeshéh’n in deinem lieben lichte gehn,um bergeshéhle mit geistern schweben, auf wiesen in deinem dimmer weben, von allem wissensqualm entladen, in deinem thau gesund mich baden! 329. -l95 “lesen s:oshte -ye> -Leat0 -[es LF ‘p}i32}> ‘sfoso] -sfo ‘ae-}o] -fr p] [tats -Jos ‘pwfoswlo -j3>F peta -ofte]so sFlcle, srlele: -2edt. s)-ofrewlo ‘2)ofaod +35 ‘otos3o ‘rele: -[re+] -afp>]| -J2s -0]3 Psso]ex3, tr -ost ‘s[+3 ->]s ;wHs ->}res. -D]e+ ‘2)€/s] e+ -wlts -0]5 ‘oshe ;etlo np] slte:]5, -150 -a}ip -0]3 -0LF ‘wfosos -o]e: 2]erjesatales -3] es[tawes[oo] ‘}reo| ca. ‘lise: Je: 3] oftewlo :s[te1 >[slo lerayzo -ftples -olts] pl >ediofota -s]es]e: “s}ealte], -t30 :2]0] -3foos “|ewlole:]s -ost ‘oles -Jws -slo -ja> -ole ‘sfosos 2ojosl}3 -phor aojosljs:>}re2o[tLes -ost pl 2r[te]3. olisler «freo] -Lto edelo 5, 450 dEx}1co] :2,ofrswlo :‘'efes3 -sf -of3 -lraxowlo ‘eleler?’ “sft, :s}ecol -DLe: -J30e+] a)3s |30'exfs0]0, ‘Js ‘fs(0) -]s -o} -ae}10], -Lo ‘2]es] -3foo -|erJo ‘elcles, -190 d[tew] -a}o > |ps ->fer ->[oolr ‘Flele al He>|coo!’ ‘-s}+ :a[to -sf -of :s]c] -o] -jea :3fo0s 45,’ ‘Flexo -Dle: > les] >Hesoftle. ‘-aJs -f[s(0) -jee D}es, :sfico] -Oler :J30e+], -t90 -ala pletion ->}3 ‘O}35. LATIN SOUNDS 330. VOWELS. fe | f (+ || dd [ a) ft t f } 4 331. a= J. ae 2m 4. pater. re i _ mater. re) ‘ caelum. eng | » aurum. Orde acai L 3s eeu celer: Si ose eta ALOR eta = aah ‘s heu. Lies ee D - difficilis. I = f a finis, rae eee 7 modo. Gray =n eee ” nolo. Ceri cma Sy 5 gi) by DORN Ee 108 LATIN u = t aswm _é *fluctus. st He nubés. Vo wistat r hymuus. Vie eee jj) yr us. 832. CoNSONANTS. BS3.00) oe aD as im ibd. Chee = calculus. Cal eae 5 pulcher. ER bs domus. Foecc i » fama. i joer ¢ ego. jeer * hora. Lore SS » jejanus. Po 2k Or, OF a, lidus, athléta. _ The ee 2 mé. PT eee icang st ie non, longus. p = dD »» pater. ph =~ pe » philosophia. q = @s i) Ne r1= @§ a rarus. CONSONANTS 109 rh = Of as im rhétor, barathrum. Ss tS e sic, GG H=0 De a! tu. thy =a Ore * theatrum. Viet awe + Vivus. <= GS ne axis. ale ak : zélus. 834. Final m was assimilated in place to a following nasal or stop : cum nobis - = ats 5}iplts, tam magnus = Ojr rjasts, cum quattuor = td adsjoottus, decem dies = ofafs olffts, tum pater = otF pjofos; was assimilated to 7 and /: cum regibus = atos osfralnts, tam levis = Odjw ofols, while before 4, 7, v, s, fit represented a nasalization of the preceding vowel—perhaps accompanied by lengthening : cum gudice = als otolaf, cum virtute ats sfosotof, cum silentio ats slwfsol}e, cum faleibus = ats >Jwalpts. So also before a pause, or before a vowel (or vowel preceded by 4), where the vowel was slurred in metre: caput magnum = Ajpto Fjasts, magnum opus =4 riers \ dats d g ~~ U eats 335. The m of com-, etc. in composition was treated in 110 LATIN the same way, as in concremare, consistere = AFAAUsp FI} HOsf, a}s sfsofosf. 336. The voiceless stops were followed by a voice- glide, as in French and Italian: capita = ‘a'}p'fo'}. Voice-stops had full voice initially, as in French. SPECIMEN. 837. aa}risas} ‘ojso[s Jotroftof, ajol-of+5}, pJof[sol} 3}so0}+? -as)s0ft+ Lol]s >to}o -fsof ‘otts ‘SHs [ewholo? -as[s Jo ‘>fesps -s[es[+ f>>o[+-3]10] ajaojplo jrojal}? sfeosf -ofs sfaotosstr pojesfoltr pj w)rolt, ‘slo ‘topls of-alwl}c, “sft ‘Olrfo ‘pfotols, ‘slim afacatwsts pf 5} ots -feolts, ‘Slr ‘efra rheslrolsslets of -pfsols sf-s}iotts ‘w}ats, ‘Slo -e}ots Ho) stooisasl rH s[twtso? pj-oftol ‘Ot] ajssfol) s}+5 sfsols? a}fssowlao]s os ‘Felts -e}otsa a}ssal[sol} of s[tols a}soho}ol pss ot] 3H «olofts? -anlo ‘pufasfr}s, ‘aslo Stofol}of sfaof ‘[rafols, -tnlt sifels, -an}s Apa} a}afols, aslo -a}sslwl+ -altofols — -anfs ‘Sfso0ts fras}+o}of Joofoojwls? -}+ ‘ofeofo]! +} ‘FHo[ts! s[s}iots ‘ejca [s-ofwolalo; -a}sstw -sfofo; ‘efia ‘oje[s oltofo. ‘altafo? -Leer}+ cofto}s fs SP:3}ots ‘aslo; >Lo ‘ptrpwlals -a}ssfwls ‘pJoolafos ; ‘SJOJo [Oo ofrslasjo ‘Jatwlts Jo -ajcofs sta -asfsazf ‘Syso0ts. ‘ss j20fs, ‘olol+ >}oolts, sols >Jafof o[tfeptrpwla}e ofoftrto, ‘sft Ls-olsts >to}wofs Ja ‘O[to] aleoftrts. Jo e}oofs ‘ols ajofwlis] ohals ‘olisst+ ‘a}sstwls ‘oe ‘poltops Jo}ooftejo; [5 -ofs SPECIMEN 111 a}s>fool+ -pfsofs -Lso]s, -asJt ‘ole [5 -3}1s “Jro[ts ‘ojs0lt+ Fracls}ols ! 338. Qvodusqve tandem abutére, Catilina, patientia nostra? Qvanditi etiam furor iste tuus nods élidet? Qvem ad finem sésé effrénata jactabit audacia? Nilne té nocturnum praesidium palati,nil urbis vigiliae,nil timor populi, nil concursus bondrum omnium, nil hie muni- tissimus habendi senatiis locus, nil horum ora vultusque moéverunt ? Patére tua cdnsilia non sentis ? Constric- tam jam omnium hérum conscientia tenéri conjiratio- nem tuam non vidés? Qvid proxima, qvid superiore nocte égeris, ubi fueris, qvds convocaveris, qvid cénsili céperis—qvem nostrum ignorarearbitraris ? Otempora! O morés! Senatus haec intelligit ; cdnsul videt; hic tamen vivit. Vivit? Imm6 vérd in sendtum venit; fit piblici cdnsili particeps; notat et désignat oculis ad caedem tinum qvenqve nostrum. Nos autem, viri fortés, satisfacere réiptblicae vidémur, si istius furdrem ac téla vitémus. Ad mortem té Catilina diici jissi consulis jam pridem oportébat ; in té conferri pestem istam, qvam ti in nos omnés jamdit machinaris! 339. Ho wou GREEK VOWELS. d as in harma. je sno pate: dz 7 autos. de MA praxis. ec » khorai. [ i" egéneto. [+ n leips. (2 . basileus. Tt » aléthés. Tt He péi. [tt ae éuxandmen. { i hippos. Vowels began with the jz Orts. 341. ti CONSONANTS 113 = f+ asin krind. a0 Pwr | 2AOT08. = ae oikos, en © ee hofitos. — flee ©) ae déma. : Se = je , oide, ie Ap muia. = Nt SEES sikon. clear beginning (119 b): CONSONANTS. OS, oS vy) G)+, Ot Or, Or°, Or ite de Or q Aq | il a) How ue ul 7 RB asi _ bradus. diddmi, Ag6... | Aggelos, hippos. kakés, kheir. légo. mallon. 114 GREEK 5 asin nin. n == p fas « eae ies #3 ph = pp. ,, sophia. r = OF , éris. rh = os , rhefima. skotos. re { sbénnimi. t pee (Or a: tdte. tt (ss) = W ,, tatto. th eee OR Ae $5 tithémi. x = QS %,, ddéxa. Z =" POS; igs neneugon. 343. Voiced consonants were unvoiced after the as- pirates 4h, th, ph, the aspiration passing into them, as in khrénos ‘avs}5}s, phloisbos pto}stsp}s, thnéskd ‘OsT+sSaAyF. SYNTHESIS. 344. The Greek synthesis and organic basis was pro- bably similar to that of the languages of Southern Europe now. Voice stops were fully vocalized, voice- less unaspirated stops were followed by a voice-glide as in téthémi ‘o'Looterl. SPECIMEN. 345. ‘Ft OLt0, ‘Ft ‘oJIO[s oof}, ETH O[ts ‘O]z0e feregzes [of-s[as[t(s, Jo-w) ‘rjwlso] -e[3 ‘ale ‘oboj}rs plwolyze of5] ste ‘ajc ‘pos(s]s [s-oo[epol, -[+ Oo ‘Jos -[actisis {sl}roy+s, ‘Oleols ‘e[3 J2otes -ajoe []zoles [as‘yro[ts ‘aje pos} jrm[ts -[a ‘apie ‘aje co} mjwte p}ctesof, SPECIMEN 115 eTFL5 Ol ‘O}Ls W}ro}cs ‘OT Of -aclsops Jojow) apes OFT [OTOsO per [If+F ‘p}oeHs -‘Ojol ‘ae spop os! js spe} wt. 846. Mé dét’, 6 pantes theo}, médeis tatth’ imén epi- neuseien, alli. mélista mén kai todtois beltid tind notin kai phrénas entheféte, ei d’ ar’ ékhousin aniatés, totitous men autotis kath’ eautots exdleis kai prooleis en géi kai thalattéi poiésate, hémin dé tots loipots tén takhis- tén apallagén ton epértéménén phdbon déte kai sotérian asphaleé. be ¥ Ae oes: ¥ at * pe soe: a oi es + if oh ¢ Hy nt Rts Seip 7 Bue re , 4 x : » # weet Fe) 4 r i roti a a pt hood d ia el 4 2 ay pi th ; iy - Cia 4¢ Le 4 rd : ’ or ; eure tes) ? £7} + es F i A tisestak we | a mi ae Der fe rey "ERE EP 6 ERLE Ges aS SE “aa r Ha ‘ : af AD * ‘3,2 4 or y APPENDIX List of Symbols. (The references are to the paragraphs.) xi stop modifier 69 ec. | vowel in G. gade (second x] syllabic modifier 150. vowel) 50. - I voice, voice-glide 26. te gee gel Ir 31. 1» in better (second E voice-glide rounded. vowel) 5o. 1 vowel 5o. es eo BOe Be gh = EPP ou: KO. Vp 1h er 50. 1 » 50. Eo» 50 BP ein pub FO: IL >» In how 50. Weis: in be? t5 oe! ZEST Ro: FP AUN EGY 80; So. Doge SRR S RO: ‘ies... In fan'5O. fo eSB FR ee RO! ek. eh nna" 50. Deeinttk50. iy es 50: £0 S850} J , Inadl 5o. Peet ry 50; Toray 50. f ,, in Fr. peu 50. J 5, Innot 50. { 4» In men 50. I ” 5°. f 33 5°. FOL, #050. T >» in ar 50. T 9 52. t »” 5°: Be P52 IGOR T >» in man 50. 118 LIST OF % vowel 50. 1ete. diphthongicl ete. 153. x+ long 97. x* half-long (medium) 97. x+ short 97. xi open modifier 69a; spreader 41, 71 €. x= level force 102. ‘x clear vowel-glide 117 b, 121b. x' voice-glide 125. x! voice modifier 26. "x level tone 162. -x weak stress 107. ‘x rising tone 162. ‘x falling tone 162. rx rising voice-leap 162 ; high key 169. *x falling voice-leap 162 ; low key 169. x4 retracter, inner modifier x41 in-vowel modifier 37. x+ advancer, outer modifier | 36, 73: xt out-vowel modifier 37 ; interdental modifier 75. X+ raiser 36. x+ lowerer 36. x+x link 72, 76. SYMBOLS x* general modifier. x< increasing force 102. x> diminishing force 102. XA tenseness 70. xv looseness 70. x< in-breather 16, 89. x< click modifier go. x> out-breather 16, 89. vx compound rising tone 162. sx compound falling tone 162. “x weak stress 107. *x open stress 159. xy wide modifier 43, 66. xi narrow modifier 43, 66. xsnasalitymodifier 45,694. s1uvula stop consonant 78. 4 consonant 79. | ff in sing 79. £ 33 79: if Me =Ital. gn 79. 3 99 79: 3 9 =n 79- J ) 79: F 39 = 79- xs trill modifier 69 e. o breath 25; basis for group-modifiers100,112, 172, 183. Or 31. LIST OF X° aspirate 122, 129 2 vowel-aspirate 118. *x gradual vowel - glide 117a, 121a. x* breath-glide 125. xx glideless modifier 143. x3 breath-modifier 25, 48. ‘x strong stress 107. :x medium stress 107. ;x extra strong stress 107. ¢ consonant in G. ach 79. € . in G. sage 79. fe) " in G. ich 79. y) 7 in you 79. O 9 79> 0) Y =r 79. 9 93 79: 3 9 79: CG . in G.auch 79. & ” 79: a) 2 =wh 79 re 39 =Ww 79: q * = a: a 39 =F 79: © ” 79: @Q ; 79: e) Mt == E708 D . =d 79. D 9 =P 79: Lc) 39 =6 79: SYMBOLS 11g V consonant in thin 79. v F in then 79. > 3 =f 79: ? %9 Re eE g , 79. s 99 79: 0 ‘9 79: 09 7 = Ital. g/ 79. @ % 79: uy ‘is =¢ 79. 3 %9 79: 8 3 79. x( back modifier 71a. x\ front modifier 7 1b. X\ point - modifier 71. x) rounder (lip modifier) 39- x) non-syllabic modifier 114, 150. Xe inverter 71¢. xo protruder 40, 71c. X glottal stop 29. x§ blade-modifier 71d. x) inner rounder (lip-back modifier) 39. 0 whisper 27. » whisper modifier 27, 49, 65. \ throat-stop modifier 130. 47; OXFORD PRINTED AT THE CLARENDON. PRESS BY HORACE HART, M.A, PRINTER TO THE UNIVERSITY Books published during the twelve months ending June 1, 1913, LE a SE AI GS AO I II Ey or in the press, or in preparation, are underlined. 5A AREA MAR SE a EA BARRA bE TE TRE I PETS YS RS I A ROTTS TCR ESA He ENGLISH DICTIONARIES THE CONCISE OXFORD DICTIONARY. Adapted by H. W. and F. G. Fowrer (authors of The King's English) from the Oxford English Dictionary. 1056 pages. 3s. 6d. net; on Oxford India Paper, 6s. 6d. net. THE OXFORD SHAKESPEARE GLOSSARY. By C.-T. Onrons: Based on the Oxford English Dictionary. es. 6d. net.; on Oxford India Paper, 3s. 6d. net. CONCISE ETYMOLOGICAL DICTIONARY. By W.W. Sxeat. New Edition, arranged alphabetically (1910). New and Corrected Impression (I9Ir). 5s. net ; thin paper edition, 6s. net. STUDENT’S DICTIONARY OF ANGLO-SAXON. By H. Swerr. 8s. 6d. net. CONCISE DICTIONARY OF MIDDLE ENGLISH, from rrs0-1580. By A. L. MayHew and W. W. Sxeat. 7s. 6d. GRAMMAR, PHONETICS, ETC. ENGLISH COMPOSITION. By A. Cruse. as. 6d. ELEMENTARY LESSONS IN, ENGLISH GRAMMAR... By: H.C: WyLpD. 2s. NEW ENGLISH GRAMMAR. Based on the Recom- mendations of the Joint Committee of Grammatical Terminology. By E. A. SonNENSCHEIN. Jn the press. GRADUATED PASSAGES FOR REPRODUCTION. Selected and arranged by M. L. Banks. 2s. JHE KINGS “ENGLISH : Abridged Edition. “By Hi Ww; and F. G. Fowier. ts. 6d. Complete Edition, 5s. net. WRITING OF ENGLISH. By Puitie J. Harroe, with the assistance of Mrs. A. H. Lancpon. Second Edition. as. 6d. ESoenvew RITING, RHETORIC, AND PROSODY. ere VeERETONOMITH, 1.25, 6d. 10h, cy Min Parc, LENSING ae EXERCISES IN DICTATION AND COMPOSITION. Selected and arranged by N. Notman. as. ‘ PROGRESSIVE COURSE OF .PRECIS) WRITING, By F. E. RoBeson, 2s. 6d. Key, by J. A. MortLock, 5s, net (applica- tions must be made direct to Zhe Secretary, Clarendon Press, Oxford), i CLARENDON PRESS SCHOOL BOOKS HAKLUYT: Select Narratives. Containing the Voyages of Gilbert, Hawkins, Drake, Frobisher, Raleigh, and others. With portraits. By E. J. Payne. Firstand second series. Ed.2. 5s. each. Voyages of the Elizabethan Seamen (Hawkins, Frobisher, Drake, Gilbert, Amadas and Barlow, Cavendish). By E. hy Payne. With notes and maps by C.R. Beaziey. With illustrations. 4s. 6d. Separately, Voyages of Hawkins, Frobisher, and Drake, 2s. 6d.; Voyages of Gilbert and Drake, 2s. 6d. HOOKER: Ecclesiastical Polity, 1. By R. W. Cuurcu. 2s. MARLOWE: Edward II. By O. W. Tancocx. Third Edition. 2s. (boards) and gs. (cloth), MORE: Utopia. (Robynson’s Translation.) By J. CourTon COLLINS. 2s. NORTH: Translation of Plutarch’s Lives of Coriolanus, Caesar, Brutus, and Antonius. By R. H. Carr. gs.6d. Separately, Coriolanus, and Caesar, 1s. 6d. each. Caesar, with Shakespeare’s Julius Caesar (Oxford Plain Texts), 2s. SCENES FROM OLD PLAY-BOOKS. By P. Simpson: 3s. 6d. "SHAKESPEARE: Edited by G. 5. Gorpon. Midsummer Night’s Dream, As You Like It, and The Tempest In one vol. as. 6d. Richard II, Julius Caesar, and Macbeth. Inonevol. as. 6d. Hamlet, Coriolanus, and Twelfth Night. In one vol. as. 6d. Separately, Hamlet, Coriolanus, Midsummer Night’s Dream, As You Like It, The Tempest, 1s. net each. By W. G. CLark and W. Atpis WricHT. As You Like It, rs. 6d, Coriolanus, 2s. 6d. Hamlet, 2s. Henry 1V, Part J, 2s: Henry V,-9s. \ Henry VIII, 2s. “Jultus; Caesarean: King John, 1s.6d. King Lear,1s.6d. Macbeth, 1s.6d. Merchant of Venice, 1s, Midsummer Night’s Dream, 1s. 6d. Much Ado About Nothing, 1s. 6d. Richard II, 1s. 6d. Richard III, 2s. 6d. The Tempest, 1s. 6d. Twelfth Night, 1s. 6d. The Oxford Plain Text Shakespeare. As You Like It, Coriolanus, Hamlet, Henry IV (Part I), Henry IV (Part I1), Henry V, Henry VIII , Julius Caesar, King John, Macbeth, Merchant of Venice, Midsummer Night’s Dream, Much m, Much Ado About Nothing, Richard II, Richard III, The Tempest, Twelfth Night. 6d. net each. Julius Caesar, with North’s translation of Plutarch’s Life of Caesar (ed.Carr) ase) ae Richard II. By Henry Newsott and J. C. Smiru. Is, 6d. net. et AY My Pa 4 ENGLISH SHAKESPEARE (continued): Merchant of Venice. By H. M. PEercIvAL. 25. net. Shakespeare Glossary. See Di: SIDNEY: Apologie for Poetry. By J. C. Coins. as. 6d. *SPENSER: Faery Queene, LIIl. By G. W. Kircuin and A. L. MAYHEw. 2s. 6d, each. SEVENTEENTH CENTURY “BUNYAN: Pilgrim’s Progress, Grace Abounding, and Im- prisonment. By E. VenaAsies. With Portrait. Second Edition, revised by M. Peacock. 3s. 6d. Holy War and The Heavenly Footman. By M. Peacock. gs. 6d. CLARENDON: History of the Rebellion, Book VI. By T. ARNoLp. Second Edition. 5s. War Pictures. Edited and arranged by R. J. Mac- KENZIE. 28. 6d. net. “DRYDEN: Selections. Prose and verse. By G. E.Hapow. 2s. 6d. Dramatic Poesy. By T. ARnotp. Third Edition, revised by W.T. ARNoLD. 3s. 6d. Selections. Including Oliver Cromwell, Astraea Redux, Annus Mirabilis, Absalom and Achitophel, Religio Laici, and The Hind andthe Panther. By W.D.Curistiz. Fifth Edition revised by C.H. Firtu. 3s. 6d. Absalom and Achitophel. By W. D. Curistiz. Fifth Edition, revised by C. H. Firru. as. Select Essays. By W. P. Ker. Two vols. 8s. net. * See page 5. 5 CLARENDON PRESS SCHOOL BOOKS “MILTON: Areopagitica. By J. W. HALEs. 3s. Comus. By R. C. Browne, 6d. By O. ELTon, ts. Comus and Minor Poems (L’Allegro, Il Penseroso, Lycidas, Arcades, andSonnets), ByO.Etron and W.WorRRALL. 2s. English Poems. By R. C. Browne. New Edition, revised by H. BRaptEy. Two vols. 6s. 6d. Separately: Vol. I, 48¢*Volalls asi Vol. I: Early Poems (L’Allegro, Il Penseroso, Arcades, Comus, Lycidas, &c.) ; Sonnets ; Paradise Lost, I-VI. Vol. II: Paradise Lost, VII-XII; Paradise Regained; Samson Agonistes. Il Penseroso. By O. Exton. 4d. L’Allegro and Il Penseroso. By H. M. PErcivat. 1s. Lycidas. By R. C. Browne, 3d. By O. Exton, 6d. L’Allegro. By O. Etton. 4d. Paradise Lost. Book I, by H. C. Bercurne, ts. 6d. Book II, by E. K. Cuamsers, 1s. 6d. Together, as. 6d. Samson Agonistes. By J. Courton Co..ins. Is. 6d. Sonnets. By W. WorRALL. 6d. KIGHTEENTH CENTURY ADDISON: Selections from the Spectator. By T. ARNOLD. 4s. 6d. ADDISON and STEELE: The Coverley Papers from The Spectator. By O. M. Myers. as. *BURKE: Selections. By E. J. Payne. Second Edition. Vol. I: Thoughts on the Present Discontents ; The two Speeches on America. , With additions and corrections. 4s. 6d. Separately: Thoughts on the Present Discontents, American Taxation, Con- ciliation with the Colonies, 2s. 6d. each. Vol. II: Reflections on the French Revolution. 5s. Vol. III: Letters on the Regicide Peace. 55s. “BURNS: Selections. By J. Logie Ropertson. Second Edition. gs. 6d. Poems Published in 1786. By M.S. CLeGHoRN. 3s. 6d. *“COWPER: Selections. By H. T. Grirritu. Vol. I: Didactic Poems of 1782, with some Minor Pieces 1779-1783. 35. Vol. I1: The Task, with Tirocinium and some Minor Poems 1784- 1799. Third Edition. gs. * See page 5. 6 ENGLISH *COWPER (continued): Selected Letters. With introduction by E. V. Lucas, and notes by M. L. Mitrorp. as. 6d. net. *DEFOE: The Life and Adventures of Robinson Crusoe. PartI. By A.C. Lippert. Illustrated. as. 6d. “GOLDSMITH: Deserted Village. By G. G. WuisKarp. Oday dext only, ad. hn jprec ah Good-Natur’d Man. By G. G. Wuiskarp. 1s. 6d. net. With Sheridan’s vals (ed. BAtston), 3s. 6d. She Stoops to Conquer. By G. A. F. M. Cuatwin. 2s, The Traveller. By G. Birxspeck Hitt. 1s. With Deserted Village (ed. WuHiISKARD) and Byron’s Childe Harold. III (ed. TozER), 2s. 6d. Selections. By Austin Dogson. 3s. 6d. Vicar of Wakefield. By L. J. Stein. as. 6d. GRAY: Elegy, and Ode on Eton College. (Text only.) 2d. Selections. By Epmunp Gosse. With additional notes for schools by F. Watson. Is. 6d. “JOHNSON: Life of Milton. By C. H. Firru. 1s. 6d. (paper boards), 2s. (cloth). With Vanity of Human Wishes (paper boards), 1s.9d. Rasselas. By G. BirxKBeck HIL1. as, Rasselas, and Lives of Dryden and Pope. By A. Mixngs, Second Edition, revised. 4s. 6d. Separately, Lives, 2s. 6d. Vanity of Human Wishes. By E. J. Payne. 3rd Ed. ad. With Life of Milton (paper boards), 1s. 9d. LOCKE: Conduct of the Understanding. By T. Fow er. Fifth Edition. 2s. 6d. PARNELL: Hermit. ad. *POPE: Essay on Criticism. By J. SARGEAUNT. 2s. Rape of the Lock. By G. Hotpen. 2s. With Essay on Criticism (ed, SARGEAUNT), 3s. 6d, Essay on Man. By Marx Pattison. 6th Ed. 1s. 6d. Satires and Epistles. By Mark Pattison. 4th Ed. as. “SHERIDAN: The Rivals. By T. Batston. 2s. With Goldsmith’s Good-Natur’d Man (ed. WHISKARD), 38. 6d. STEELE: Selections from the Tatler, Spectator, and Guardian. By Austin Dosson. With portrait. 7s. 6d. SWIFT: Selections. By Sir Henry Craik. Two vols. qs. 6d. each. Separately, The Battle of the Books, 2s. “THOMSON: Seasons and Castle of Indolence. By J. Locir ROBERTSON. 48. 6d. Separately, Castle of Indolence, 1s, 6d. * See page 5, 7 CLARENDON PRESS SCHOOL BOOKS NINETEENTH CENTURY *ARNOLD: Selected Poems. By H. B. Georce and A. M. LEIGH. 2s, Merope, with the Electra of Sophocles translated by R. Wuitreraw. By J. Caurton Cotiins. 38. 6d. AUSTEN (Jane): Pride and Prejudice. By K. METCALFE. 2s. 6d. *BROWNING: Menand Women (1855). By G. E. Hapow. 3s. 6d. Strafford. By H. B. GEoRGE. 2s. *BYRON: Childe Harold. By H.F.Tozer. 3rd Ed. 3s. 6d. Separately: I-II, as.; III-IV, 2s.; III, 1s. 3d. With Goldsmith’s Traveller (ed. Hiv) and Deserted Village (ed. WuIsKaRD), 2s. 6d. *CAMPBELL: Gertrude of Wyoming. By H. M. Firz- Gipson. Second Edition. Is. «CARLYLE: Lectures on Heroes, Hero-Worship, and the Heroic in History. By P. C. Parr. as. 6d. Sartor Resartus. By P. C. Parr. 3s. 6d. *COLERIDGE: Ancient Mariner. By M.A. KEELING. Is: Poems of Nature and Romance (1794-1807). By M. A. Kretinc. 38. 6d. See also under Wordsworth. *DE QUINCEY: Spanish Military Nun, and Revolt of the Tartars. By V.H.Cotiins. With 2 maps. 2s. *DICKENS: Pickwick Papers. Abridged, with introduction, by Russett Scorr. With glossary by Puitip T. STEPHENSON, 2s. 6d. Tale of Two Cities. By Mrs. F.S. Boas. as. 6d. Tale of Two Cities. Abridged by Russe. Scott. 2s. 6d. Dickens Reader. By Mrs. J. C. Smiru. With preface by J. C. Smiru. 1s. 6d. *ELIOT (George): Millon the Floss. By R.O. Morris. 2s. Silas Marner. By H. Marwick. 2s. “IRVING (Washington): Sketch Book. By T. Barston. 2s. LCS Bh SRN TE PEE SN NG A CE UTE *KEATS: Hyperion, I. By W. T. ARNoLp. 4d. Odes. With illustrations. By A.C. Downer. 3s. 6d. net. Poems published in 1820. By M. Rospertson. 3s. 6d. *KINGLAKE: Eothen. Introduction by D. G. Hocartu. Notes for schools by V. H. Cottins. With Map. as. 6d. *KINGSLEY: Hereward the Wake. By A. D. Innes. 2s. 6d, Water Babies. Slightly abridged. With notes and illus- trations. By J. Horace-SmirH and M, L. Mirrorp. as. 6d. * See page 5. 8 Date Due = x My 8 LTT TE : CETTE ume » PP i Hs . 4s f re. 44a a0 ey vito os i Hill ill ii ll } ll & o rT) = Te = o P=} w s) = = a.