^ ^ ^ "V^.-' ^ \r 1^ > IMAGE EVALUATION TEST TARGET (MT-3) 1.0 I.I 2.2 IL25 III 1.4 ■ 20 m 11.6 Hiotographic Sciences Corporation ^ \ \ c\ 23 WIST MAIK TRIIT WHSTIR,N.Y. l4St9 (71«)t72-4S03 '^ CIHM/ICMH Microfiche Series. CIHIVi/ICIVIH Collection de microfiches. Canadian Instituta for Historical IMicroraproductions / Institut Canadian da microraproductions historiquas Tachnical and Bibliographic Notaa/Notaa tachniquaa at bibliograpliiquaa Tha tot Tha Instituta has attamptad to obtain tha baat original copy availabia for filming. Faaturas of this copy which may ba bibliographicaily uniqua, which may altar any of tha imagaa in tha raproduction, or which may significantly changa tha usual mathod of filming, ara chackad bolow. □ Colourad covars/ Couvartura da coulaur □ Covars damagad/ Couvartura andommagia □ Covars rastorad and/or laminatad/ Couvartura rastaurte at/ou palliculAa □ Covar titia miasing/ La titra da couvartura manqua I — I Colourad mapa/ D Cartaa giographiquaa m* coulaur Colourad ink (i.a. othar than blua or black)/ Encra da coulaur (i.a. autra qua blaua ou noira) r~~| Colourad plataa and/or illustrations/ D D D El Planchaa at/ou illustrations an coulaur Bound with othar matarial/ Rail* avac d'autraa documants Tight binding may cauaa shadows or distortion along intarior margin/ La re liura sarria paut cauaar da I'ombra ou da la diatoraion !• long da ia marga intiriaura Blank laavas addad during rastoration may appaar within tha taxt. Whanavar poaaibia, thaaa hava baan omittad from filming/ II sa paut qua cartainaa pagaa blanchaa ajoutiaa lors d'una raatauration apparaiaaant dana la taxta. mala, lorsqua cala Atait possibia. caa pagaa n'ont pas «t4 filmAaa. Additional commants:/ Commantairaa supplimantairaa: PAGINATION AS FOLLOWS: viil, [5]-148 p. L'institut a microfilmi la maillaur axamplaira qu'il lui a ttt possibia da sa procurar. Las d*tails da cat axamplaira qui sont paut-Atra uniquas du point da vua bibliographiqua. qui pauvant modifier una imaga raproduita. ou qui pauvant exiger une modification dana la mithoda normala de filmaga sont indiqu4s ci-dassous. |~~1 Colourad pagaa/ Pagaa da coulaur Pagaa damagad/ Pagaa andommagias Pagaa raatorad and/oi Pagaa raataurtes at/ou paliiculias Pagaa discoloured, stained or foxei Pages dteolorias. tachatiea ou piquAas [~~| Pagaa damagad/ r~l Pagaa raatorad and/or laminated/ f^^ Pagaa discoloured, stained or foxed/ Tha poa oft film Ori| bag tha aior othi first sion or 11 □ Pagaa detached/ Pagaa ditachAas r~T^howthrough/ \j^ Tranaparanca r~1 Quality of print variaa/ Qualiti inAgala de I'impression Ineludas supplamantary matarial/ Comprend du material supplAmantaira Tha ahal TIN whi Mai diff( anti bag righ raqi mat D Only adition available/ Sauia Mition disponibia r~7r Pagaa wholly or partially obscured by arrata LlJ slips, tissues, ate, hava baan refilmed to anaura the bast possibia image/ Lea pagaa totalement ou partiallement obacurcias par un fauillat d'errata, una polurdt, ate ont it* filmiea i nouveau da faqcn A obtanir la maiilaura imaga possibia. This itam is filmed at the reduction ratio chackad below/ Ca document est filmi au taux da reduction indiqu* ei-daaaoua. 10X 14X 18X 22X 28X 30X 12X 16X 2DX MX 28X 32X Th« copy filmed h«r« has b««n r«produc«d thanks to tha ganarosity of: Mttropolitan Toronto Library Littraturt tSapartiiMnt Tha imagas appaaring hara a^r* tha bast quality possibia considaring tha condition and iaglbility of tha original copy and in kaaping with tha filming contract spncif icatlo^s. Original copies in printad papar covars ara fllmad beginning with tha front covar and ending on tha last page with a printad or illustrated Imprea- sion, or thii back cover when eppropriata. All other original copiae are filmed beginning on the first pege with e printed or illustrated impree- slon, and ending on the iaat page with a printad or iliuatratad impression. The last recorded frame on eech microfiche shall contain tha symbol -^ (meaning "CON- TINUED"), or tha symbol Y 'meaning "END"), whichever appilaa. L'axemplaira fiimA fut reproduit grice A la gAnArositA da: Metropolitan Toronto Library Litaratura Dapartmant Les imagas suivantas ont AtA raproduites avac la plus grand soin, compta tanu da la condition at da ie nettet* do l'axemplaira film*, at 9n conformiti avac las conditions du contrat da fiimege. Les exemplalras originaux dont la couvartura •n papier est imprimAe sent fiimto en commenpant par ie premier plat at an tarminent soit par la darnlAre page qui comporta una amprainte d'imprassion ou d'illustration. soit par la second plat, salon la cea. Tous las autres axempleires originaux sent filmis mt commenpant par la pramlAre page qui comporta une amprainte d'impression ou d'illustration at an tarminent par la darnlAre page qui comporta une telle empreinte. Un dee symboles solvents apparaftra sur la darnlAre imege de cheque microfiche, selon Ie cas: la symbols ^-^ signifia "A SUIVRE", la symbola ▼ signifie "FIN". Meps, plates, charts, etc., mey be filmed at different reduction ratioa. Those too ierge to be entirely included in one exposure ara filmed beginning In the upper left hend corner, left to right and top to bottom, as many framea as required. The following diagrams illustrate the method: Les cartas, planches, tableaux, ate. pauvent Atra fllmAs A das taux da rAduction diff Arants. Lorsqua ie document est trop grand pour Atre reproduit en un seul clichA. il est filmA A partir da I'angia supArleur geuche, de gauche A droite. at de haut an bas, an prenant la nombre d'imagas nAcessaire. Las diagrammas suivants iiiuatrant la mAthoda. 1 2 3 32X 1 2 3 4 5 6 ■> ,-'N J TREVELYAN PRIZE ESSAY. / ANALYTIC ORTHOGRAPHY: AN INVESTIGATION OF THE SOUNDS OF THE VOICE, AND THIIR ALPHABETIC NOTATION; THE MECHANISM OE SPEECH, AKD ITS BKAHIMQ UPON ETYMOLOGY. m S. S. HALDEMAN, A.M., - , , PBOrE880R IN DILAWABB COLLSaB; ■mUB or TBI AHiucAif rRiUMOFBicAt Boonn; or the academy or natvbal sciences or pniLADELrau; or Tni abibicA!! okieiital soiiitt; or TBI mrEUAL economic sociitt orst. rETn$Bi;RdH; rELLoworTHS amebicax acadkmt or a:its axd sciencu; bonorabt mehbeb or THB bistobical socibtt or wiscoNsin; coRBE«ro»DEHT or THE natural history society or NVBEHDEBo; or the BOSTON SOCIETY Or NATURAL HISTORY; or THE NEW YORK HIBTOBICAL SOCIETY; Or THE HUTOBICAL 800IBTY or rENNSYLTANIA ; Or THE MABYLAND HI8T0BICAL SOCIETY; AND or TBB AMERICAN BTBNOLOOIOAL SOCIETY. PHILADELPHIA: J. B. LIPPINCOTT & CO. LONDON: TRUBNER & CO PARIS: BENJAMIN DUPRAT. BERLIN: FERD. DUMMLER. 1860. V !' » > ""»" ' »• ^.a I /. i, i. J* ^ Jl/^,S-M- Pi 'V i!' li^i i i - .. I '' vt ^ A mUsi Xi :fK:lPi^t>i PREFATORY. Altrouoii the Essay following owes its form to the prizes offered by Sir Walter C. Trcvelyan, A.M., tlie material has been accumulating during a number of years, in connection with Ethnology, Kpccch being an important characteristic of man. But in taking cognisance of speech as it occurs in nature, it is found very different from its representation in books, as we learn when German, French, and English are really compared. It was considered necessary to record such vocal phenomena as we had observed, and out of this a notation has arisen which those whose knowledge of languages is based on books may regard as too minute, whilst those who are familiar with languages we have heard but casually, will prntwbly discover that our chief error has been a want of nice discriminating powers. But whilst it is the duty of the explorer to record the minutest phases in a given language, the natives themselves will determine how far these distinctions are to be expanded or curtailed when represented in an alphalnst. Our "Latin Pronunciation" grrv, out of the question of alphabetic notation, and in that we determined that if the Roman Alphabet is used as u oasis, the letters must have their Latin power; and this is the view of others, as will appear in the course of our Essay. Unfortunately, there are many who admit the justice of this, as long as they suppose that Latin was pronounced as they pronounce it, who change their view when there are discrepancies, and are ready to pull down the structure n Latin etymology and prosody to their own barbarian level. « ' When the ancient testimony on the power of a letter is undisputed, an argument is used which satisftes most literary people, namely, that the modern corruptions may have existed in ancient times. But whilst we grant the justice of this supposition (§ 131,) we cannot allow ourselves to depart from the letter of the ancient grammarians to theorise on unwritten dialects for the purpose of vitiating the normal form. The Tuscan haaa for casa has nothing to do with the power of cay in the written dialect, and he who said bufalo for bubalcs, was using / where another used b ; he was not pronouncing 6 as f, but using an unwritten form, like a Roman saying piano for planus, or an Englishman pronouncing the same planus in the two modes piano and plain. In January and February, 1854, there were four Alphabetic Conferences held in London at the house of Chev. Bunsen, in which fifteen scholars and scientists participated, but they arrived at no common conclu- sion. There was no difference of opinion in regard to the power of nearly all of the consonant letters of the Roman alphabet ; yet we have not followed them in regard to k, v, w, y, z, in which they sacrifice and cor- rupt a fifth of the Latin alphabet. (V) vl PREFATORY. The new Lawa of the ifechnniHtn of Speech, and the PhyHioIogy and Physlojynomy of Words, m cxhil»ito«l here, wc l»elicvc to »mj tlie true haHin of etymolojry, and they will tw taken up and expanded here- after into an cdiieationa! uericii on tlie PiiiloHopliy of Etymology, Affixes of the English Language, Ac. At preHciit there in no lietter proof of the low condition of lingnifitic education amongst ns, than the use of Deacon (more correctly Diacon) Trench's books in our schools and colleges, where they hare been introduced l»y illiterate admirers. These books are equally popular in England, notwithstanding the more accurate views of scholars like Kills, Garnett,* Guest, Key, Latham, f Wedgwood, &c. Although we have furnished to the phonetic periodicals several articles on etymology as a matter of speech rather than of spelling, we do not belong to the "reforming" class, and we are not aware that we have hitherto expressed an opinion of phonotypic English. Intent on a literary, rather than a philologic view of the subject, Trench and his imitators have overlooked what we consider the strongest argument against it — though not a valid one. ' The authors animadverted upon have been those whose books were at hand, although many others contain similar views, and nothing farther has been intended than to represent each author in quotation, as he has chosen to represent himself. The supposed errors of those who have taken some pains to be accurate, have not l)eon alluded to. Thus a careful author might, from a knowledge of several European languages, limit the number of pos3il)le consonants to twenty or thirty, and conclude that a sard cannot follow a sonant consonant in the same syllabic, although contradicted (§ 682) by the Hungarian words for one tad /our. The circumstances under which the following pages were printed, have caused defects in the typography which arc quite independent of the notation. The several sets of type used did not combine well, — some were wanting which are sufficiently" accessible in large offices, and the compositors, unaccustomed to technical matter, and unacquainted with foreign alphabets, had much difficulty in understanding the manuscript. Me- chanical corrections were difficult to make, and often resulted in typographic inaccuracies, so that it was thought best to be satisfied with an approximate regularity, rather than risk the dropping out of an accen- tual, or the turning of a letter which had been inverted intentionally. * We regret that so great » philologist should have allowed his prtyudloes to make him unfair in his review of Webster and Riohardaon, in which he condemns the errors of the former, and palliates the greater errors of the latter. Webster Tas the first lexicographer of English who placed definition and etymology on a proper basis; Richardson was not competent, even to follow, in either. Webster (apart from his erroneous semitism) saw the value of oriental etymologies, and he thought Tooke an unsafe guide. Richardson despised oriental etymologies, and adopted Tooke's errors— 4rat seemingly not to his dis- oredit, for the Quarterly Reviewer, so free from "narrow-mindedness," and of such "vast erudition, maseuline energy of diotiou, and scathing sarcasm," says that Richardson's defects "are not so much chargeable on himself, as on the guide whoso diets ha implioiily follows." t We quote this, and the last sentence of 1 820, in vindication of Webster,— preferring fairness to sarcasm. t The first to call attention to the affinity between French o in oie (J 213,) and English w in now. CONTENTS. CHAPTER r. I 1, General statement of the subject. Necessity for universal standards of comparison. 6-«, Errors of (rram- marians. 10, The spelling of Shaicespeare relinquished. 36-8, Trench's argument 4?i. Ten Paradoxes. CHAPTER II. 1 42, The Roman alphabet the proper basis of notation. 47-64, Six Rules of notation. CHAPTER III. i 72, Kinds of alphabets. 72-5, Hieroglyphs. 78-82, Philosophic alphabet*. 87, Cherokee a conventional syl. labic alphabet ; probable theory of its formation. CHAPTER IV. 3 88, The Latin alphabet. 93, Vowels. 94, Diphthongs. 95, Nasal T0*els. 118, Double consonants. 109, Hebrew in Latin letters. CHAPTER V. 2 114, The Greek alphabet 122, Accent 126, The digamma. 129, Greek orthography of Latin names. CHAPTER VI. 1 133-47, The Anglish (Angloaaxon) alphabet CHAPTER VI L { 148, Organs of the Tolce. 152, Organs of speech. CHAPTER VIII. 2 154, The Elements. 156, Vowels pure, nasal, whispered, independent, and glottal. 157-8, Consonants. 163, Coalescents. 168, Dieresis. 172, Syneresis. 174-85. Consonant characteristics. 186-89, Nasality. 190, Possible extent of consonant variation. 199 note, Notation of Sir Wm. Jones. CHAPTER IX 2 202-18, Phases of words. 208, Metathesis. 205, Epenthesis. i'07, Cydesis. 208, Eduction. 214, Induction. CHAPTER X. 2 221, Elision. 223, fbur and Five cognate with the Latin forms. 227, Plummet and lead cognates. 228, Absorption. (Vii) VIM CONTENTS. rilAI'TKR XI. I 229. Miititlinn or aimthp«ii>. T.W, rrrcPf««li>n. 2:i2, RcrPMion. 23.1, TraniPe«iinn. 2.14-ff, (314.) AntlUxii. 2:i7, (:»I2.) MetttlliixiH. 2:»H-4:J. AdlnitioH (if the vowel*. 245. Mark* of matation. 'iW-Mi, Kxamploi of inturiniitation. 257-8, Vowola oftun unchanged. 259, Vowel rolatiou irreguUr. 264-5, t'ommutttion. 26G-72, roruiutation. ClIAPTERXII. { 273, TranBmutation. 279-86, Otosia. 287-9, AHHimilation. 290-.3, Disgimilatlon. 294-307, Qlottosii. 3(M, caije from cavia. 312-14, Motallaxig and Analluxiii of coniouanU. CHAPTER XIII. I 315-38, Etymologic bearings, with examples. CHAPTER XIV. I 339, (1S6,) Vowels. 350-1, Stopt vowels. 352-8, Quantity, or relative length. 359-68, Notation of quantity. 369, Scheme of the pure vowels. 370-440, The pnria vowels in detail. 444, Scheme of intermntation. 44.'i-8, Independent vowels. CHAPTER XV. { 449, Consonants. 451-63, Labials and Labio-dentals. 464-8, Lingui-dentals. 469-83, Dentals. 484-6, Indistinctness. 487-90, Arabic linguols. 491-4. Sanscrit cerebrals. 49.5-513, Sigmals. 514-24, Palatals. 525-4G, Gutturals. 547-51, Faucols 552, Laryngals. 555-6, 559, Spiritus lenis. 560-2, Hiatus. 664, Chinese abrupt consonants. 565-7, Aspiration. 668, Hamza. 570, Arabic and Hebrew ain. 671, Sanscrit vi$arga. 573-4, Laryngo-faucals. 577, Scheme of the consonants. CHAPTER XVI. Examples. | 590-609, English. 610-13, German. 614-22, French. 624-8, Cherokee. 629-32, Wyandot. 633, Nadaco,orAnadahhas. 634, Kansa. 635-8, Ohippeway. 641-2, Greek. 643, Italian. 644-8, Latin. 649, Orebo. 651-724, The numerals from one to ten, written from dictation in seventy-five languages and dialects, sixty of them being token from the months of natives. Tui following corrections, &o., may be made : | 52, 1. 5, transpose ah and sk. I 167, note c, read eonitrxteled. I 201, 1. 4, for cay read gay. Heading of i 312 read ({ 237). The arrangement in { 577 was published by us in the Linnaean Record of Pennsylvania College, for June, 1846. { 649, 1. 4, read probably. { 669', for t' read t'. 3 681*, omit I, leaving its mark stand. { 721, put Malay in parentheses. The heading (fi) of i 405 has been turned into (<)) in some copies. The Hottentot cluck on a t basis (i 447) is the only one we have heard in nature. Page 402, line 2, omit or. i 19, 1. 5, for e, t, read e, i. k 724', the third letter is *. \ 668. According to Smith, in Robinson's Palestine, vol. 3, p. 90-1, Boston, U. S., 1841, "llie Hamzeh is in no sense a breathing. . . . When it occurs in the middle, or at the end of a word, the voice must be entirely stopped before it can be pronounced." This valuable Appendix has been omitted from a later edition. \ 639, note. According to Dr. L. Loewe, (Diet of the Circassian Language,) "the pronunciation is so difflcult, that even the most distinguished linguists find it hard to imitate the sound of a syllable as uttered by the mouth of the Addee-ghey people." Klaproth says it is one of the most difficult in the world to pronounce, no alphabet being com- petent to represent it accurately; and that it has a daoking of the tongue, and several throat consonanta, which a European cannot reproduce. AN.\LYTIC ORTHOGRAPHY; a.. . { 26:i, 1. 2, read raisin*, jj 800, 1. 7, read CattrtSn. { 808, 1. 6 of note, read (p. 21) and plaoe tlie aooent of "region." after e. 8 088, the a in the table wanti the anplrale mark. Hl», 1. 5, first word, read %bn. i 01^19> 1-2, read dA,; line 12, for 'a read «,. J 084, the inlllal of the word for brow is the same as that for et/e. In the word for ihirt, for J read I. { 001*, for 7 read '. { 007', the outer branch of inverted b ■hould have been removed. | C78>, for r read r. { 080V>, invert the t. | GSils, tlie nasal mark belongs to the vowel, i 084*, third form, for c read /. { *18&i, for * read f . { 087*, the first vowel has an acute accent. J 088, read Coptic'**, and in the note •*Memphitio. { 088*, read ff, and in the note ff !■> our mr. { 080*, for I read |t, and Xt '■> tliB lo^t "ote* I ^^^> , insert initial S. g 702', vSliS. The mark of * belongs to v, and also in 704^ and 707*. | 704', for iiS read Sii. Place " over v. J 705*, plaoe a grave aooent over the first vowel. { 700*. the second letter is j. { 718*, after I insert s. S 714-10, the k is for o. { 050*, for p read n. The final aspirate in | 702^ is that described in { 508. It is not indicated in all oases, but is assigned to one of the forms ({ 720*) of Chinese. It is remarkable that the opposite Chinese phase (} 504) should have been indicated as occurring in the word for seven in two Freuob dialects ({ 000^70) taken independently. The PuBLisHERg of this Volume are in no manner responsible for the typographic execution of these pages, the sheets liaving been fur- nished to them from the forms as prepared for the "Transactions," of which they form a part. S. S. II. Chemistry. 2. -vw-Maa v«ta>t«V«l UtttrtWO* -x '^v VIII CONTENTS. CHAPTER XI. i 229, Mutation or onathosis. 231, rrocessinn. 232, Recession. 233, Transcession. 234-6, (314,) Anallaxis. ' 237, (312,) MeUllttxiH. 238-43, Amnities of the vowels. 24.i. Marks of motation. 246-,56, Examples of intermutation. 257-8, Vowels often unchanged. 259, Vowel relations irregular. 264-5, Commutation. 266-72, Permutation. \ Addee-ghey people." KTSprotn saynrs one or tne most dittcolt iil the world to pronounce, no alphabet bemg com- petent to represent it accurately; and that it has a clacking of the tongue, aod severe! throat consonants, which a Buropean cannot reproduce. ANALYTIC ORTHOGRAPHY; %n |nbes%tion of i\t Sranbs of tjc Mtt, anb IJtir ^Ipjabetit ftolation, BY S. S. HALDEMAN. (From the Traniaotions of the Amerioan PhiloBophionl Society, Vol. XI.) CHAPTER I. CuMPARATiTE Oraminar cannot acquire a scientific shape until it discardn the pedantic fetters of orthography, and writes all languages according to one system; fur things of a kind admit of a just comparison only when compared by the same standard. In this respect, philology is iu its infancy, and we place difficulties where none are to be found in nature. — Bapp, Grundrisz der Oramraatik des Indisch-europaiscben Sprachstammes, 1855, p. viii. § 1. The present tendency of science is to adopt standards of universal application, and it is usual for learned societies and associations, to have a permanent committee of research, consultation, and correspondence, with a view to bring about a uqiformity of weights, mea- sures, and coinage. 2. The advance of linguistic science demands a uniform nomenclature and notation for the phases of speech, so that the same syllable may be written in the same manner, wherever there is occasion to use it, just as every known plant and insect is recognised by a uniform Latin name among all who are familiar with botany and entomology. 3. Although the want of a uniform mode of representing languages is felt as an urgent necessity, they have not been provided with a letter for each sound; whilst chemistry, (which is not studied by one in a thousand,) has a perfect notation, an alphabet of dotted or marked letters, to represent some sixty elements; and, as it were, spell all their ascer- tained combinations. 4. Berzdku did not hose his symhdlii on his native Swedish, but upon Latin,* without even looking at the inconsistent and cumbersome notation which his predecessors of the last century had used, and which may be seen in their books, or in the £ncyclopa>dia Bri- tannica, as late as the year 1798. * " Beneliua has properly selected them from LaUn names, as being known to all civilized nations." — Turner' $ Chemiatry. 2. ^ •«i»> 6 ANALYTIC ORTUOGRAFUY. 5. Alphabets of hundreda of characters have been cut for Arabic, Sanscrit, and Greek; * the Greek vowel iota requires the fifteen types i, i, i, i, f, ?, t, ?, ?, ?, ?, ?, i', i, i, the general al- phabet proposed by S'ufiic', (Schunjitsch in German letters,) requires seventy-two vowel modifications; astronomers and mathematicians have a sufficient typography, and the com- plicated hotation of modern music can be set up in detached types. 6. The chemic alphabet came from the hand of a philosopher; English writing has been controlled by the literary and superficial, as distinguished from the scientific public; the alchemists rather than the chemists — astrologers rather than astronomers — linguists like Trench, rather than philologists like Bapp, who "settle" questions in spelling, pronunci- ation, and grammar, according to English analogies, without knowing what these analogies are. f 6 a. (hold Brown writes a ponderous " Grammar of English Grammars," after consult- ing about four hundred authorities, but instead of producing a cyclopaedia on the subject, the work is worthless for deciding questions which depend upon general principles. With him, (and probably nine-tenths of his four hundred grammarians,) aioe is a tripthong, be- ginning with a; and with Trench, (in lectures, and therefore clear of spelling,) *^anl and cniw?c< were originally different spellings of the same word," (as "gad'' and "jail," or "plough" and "plow" are at present,) but he does not tell us whether the " same word " that "ant" spelt, was emmet, or the reverse, "emmet" spelling aw^.J Q b. A college student asserts, in a published communication, that one of his professors *' " Where ligatures and abbreviations abound . . . 750 boxes are required for the different sorts of a fount of Greek ... It must, however, be observed, that almost 300 of these sorts are the same, and have no other dif- ference than that of being kerned on their hind side; for we remember to have seen Greek with capitals kerned on both sides." — Printers' Grammar, 1797, p. 242. f Thh Essay owes its form and matter to the following circumstances. In the year 1857, Sir Wr . C. Treveljan, A. M. (Oxford,) of Wallington, Newcastle-on-Tyne, offered two prizes for essays on a Reform in the Spelling of the English Language, to contain, among other features, an Analysis of the System of Articulate Sounds — an Ex- position of those occurring in English — and an Alphabetic Notation, in which "as few new types as possible should be admitted." The last requisition has, in a few cases, resulted in a double notation, one of which represents the author's preference in a new form of type, the other being a form in use, but not approved. The investigation was made from a natural history point of view, and the results are here presented. A Report is yet to be made to the American Association for the Advancement of Science on the Subject of an Alphabetic Notation for exotic Languages. Suggestions and criticisms are solicited towards this end, to be addressed to the author at Columbia, Pennsylvania. X Similarly Webster, the chief of English lexicographers — « nations differ in the orthography of some initial sounds. . . . Thus the Spanish has Uamar for the Latin clamo." This is a difference of "orthography" in the same sense that English "knee" differs from the Saxon "knee." People who hold such views must consider tear tear, sow tow, boppye in the same line, and Jonstonus (1657,) uses eels and eeles. Chaucer has egre eyer, mdUjre maugre, lest litt luste, Itwed lewde, knicn knene, hackenaie luxkeney. f "These are not his spellings; ho edited no play, and the Tempest was not even published in his life time. They are printer's spellings, probably more regular than his." MS. note of A. J. Ellis. /" ANALTTIC ORTHOGRAPHT. ^m. into thin Ayro ; ! And, like the baselesse fabricke of this vision, The Clowd-capt Towres, the gorgeous Paliaces, The solemne Temples, the great Globe it sclfe. Yea, all which it inherit, shall dissolue. And like this insabstantiall Pageant faded, Leaue not a racke behinde. — Tempett, Act 4, So. 1, 1G28. 11. " Here then are England and Wales, with their sixteen millions of people, with near* ly eight milHona unable to write their name, and not less than five millions unable to read their mother tongue."* In the United States, even in the states which supply the educa- tion at the expense of the treasury, the number of illiterate people is very large. The time for attending school is limited among the poor, and schools are rare where the popu- lation is sparse, so that minds of a high order may remain undeveloped. Energy indeed may overcome great difficulties, but this may form no part of a mind of high generalising and inventive powers. 12. The millions qf/reemen kept in mental and moral darkness, instead of loving an or- thography, know not what it is, whilst the great mass of readers despise it; — some think- ing it a trick of the schoolmasters to extend the period of tuition — whilst others regard it as a means of separating society into a lettered and an unlettered class.f 13. A child aged thirteen, who can read, has within a few days spelt as follows : — b-a-o-t boat, (not knowing the position of the "silent" letter,) 1-oo-k, 1-o-k, lock, (putting "double" before o is not suggestive of a different sound,) m-u-r-o-u-r mirror, {"you" and "eye" are equally unsuggestive of the first vowel of this word,) c h-i-r chair, (saying c-h-ai-r instead of c-h-a-ai-r.) 14. Among the moat mournful of theatrical scenes, such as are most likely to call up feel- ings akin to those of the poet who sung — Srdce moy szarcze ach hui deos sadnissa! Kard man hiort ag ouige diz sathinassus? are those in which an illiterate character slowly spells out a letter, commencing " D-ee-r C-u-r," and is greeted with a shout of laughter from people who would spell cur (which has a cay sound) with a consonant called see, and a vowel called you, and then pronounce this s-you-r as cur. * British Q. R. Nov. 1846, Art. VIII., p. 472, quoted in Ellis's P'ea, 2d ed. p. 66. f "It is better for criticism to be modest . . . till the pardoi tble variety of pronunciation, and the true spelling by the vulgar have satirized into reformation that pen-craft which keeps up the troubles of orthography for no other purpose, as one can divine, than to boast of a very qnesfionahk merit at a criterion of education." — Dr. Jumes Rush, Philosophy of the Human Voice, Philadelphia, 1833, p. 383. r 1. 3^ • ANALYTIC ORTHOGRAPHY. 15. Three milliona of people can support a literature in all its branches, from primers and almanacs to encyclopoodias and universal histories. This may be given in round num- bers, as the amount of population supporting Danish, Swedish, and modern Greek ; and about a million Albanians are divided upon three alphabets, the Italian, Greek, and a na- tive one of 52 characters, more dififerent from the Greek and Italian than these are differ- ent from each other. A journal is considered to be well supported when 2,000 copies can be disposed of, and in the Book Trade, works devoted to special branches of knowledge are often printed in editions of 250 copies, not as rarities for bibliomaniacs, but to supply the probable demand. 16. Wlien more rational modes of orthography arise, there will therefore be much danger, not from the dearth of books, but from the multiplicity of alphabets which will be pro- posed — and it is possible that there may be half a dozen in the British Islands, and twice as many on the Vesperian side of the Atlantic. 17. There is a politic reason for a reformed orthography. The age demands it, and the population is moving steadily towards it, unconvinced by platitudes on the Study of Words by those who have not exhibited that acquaintance with the science which the discussion of its principles demands. The reform should be undertaken with all the aids that science and scholarship can command. Let the fields of philology, physiology, epi- graphy, and living speech be explored, and let an alphabet be erected, so free from those national perversions which national vanity might wish to be legitimate, that no one will have the power to say — " They are only exhibiting the dress of their vernacular," — "This letter has a purely English power," — " l^hat is a French corruption." 18. Let i?ie alphabet be capable of enlargement, to render it adaptable to all languages, whether English, Italian, or Tahitian, and equally suitable for the dialect of the peasant and the tables of the comparative philologist; and let it not run counter to the great ety- mologic and metric principle which requires that all records, statements, and comparisons, shall be made in symbols, each of which shall represent the same phenomena. 19. The great succeaa of phonography shows that not a single concession which is false in principle, need be made to conciliate English sympathies, (§12,) or to preserve so-called English analogies; and it would be unkind and ungenerous to all nations having the allied pairs of vowels in they them, marine mariner, he his, were the attempt made to assign cha- racters to them as diverse as a, e, for the former, and e, i, for the latter. The unlettered five millions yeeZ theaffinity between the vowels of break and wreck, who would see no more fitness in the dissimilar forms a, e, than the chemist finds in the cumbersome notation of the alchemists. ANALTTIO ORTIIOGBAPHT. 11 20. It is admitted by Mr. Ellis, (Plea, 2d ed., 1848, p. 130,) that his English alphabet of 1848 would injure the visible etymological connection between Italian and Latin; "but toe should us much injure the visible etymological relation between Emjlish and latin by any other mode of spelling." But as Italian is nearer to Latin than Englieh is, a proper or- thography would show it. English has no right to seem to have a certain resemblance to Latin which it has not — to pretend, by spelling secure with an e-character, that it is nearer to the Latin securus than is the Italian sicuro. § 256 — 8. 21. The English consonant th, and the vowels in at, up, not being Latin sounds, should hot be represented by Latin letters, but by new or modified forms, so that the eye could detect strange or unlatin elements in an unlatin language, as readily as the eye detects Pc^lish by its crossed I, and distinguishes Portuguese from Spanish by a nasal sign, which also separates Polish from Bohemian, as it should separate French from Italian, to exhibit its afSnity with Portuguese. 22. German should not exhibit a seeming resemblance to English in th for t in tJteil, (a part,) a cognate of deal, because English th is not wanted in German ; nor should French have th (for Greek tbeta) in tliime, where the English are entitled to it. The Welsh, having the /sound, should not writejf for/of the twelfth century, and having English v (for which *w' was used in the twelfth century, and *u' in the thirteenth) they should not write it with the / character — although this is a trifling error compared to that of using the Latin V {toay, § 106) character for the English vee sound. In short, Welsh, German, Latin, English &c. writing should resemble when the words are alike — when different, it should dissemble. 23. ^loe can pronounce French aiid Polish, we can appreciate the relations between the following pairs, in which the Poles have sought to secure an identity in the vx)rd rather than in the sign : — b^asse, behas, (snipe ;) paragraphe, paragraf ; paraly tique, paralityh; page,^2'/ bagage, ia^az'; ^9xaao\, parasol; parapluie, 2)ara/)%', (umbrella;) Triest, Trst; German meister, Polish majster; English Mr., Bohemian mistr. 24. A phypiological basis has been advocated, and the alacrity with which the Standard Alphabet of Professor Lepsius, (London, 1855,) has been adopted by various missionary societies, seems to be an evidence in favour of such a basis. Unfortunately, the acknow- ledged merits of the learned author have caused this work to be adopted without due ex- amination. This "admirable treatise," (p. III. of the preliminary recommendations,) wherein the author "clearly explains the scientific principles," (V.) the result of his "close and profound attention," (VII.) and "Fleiss," (VIII.) or industry; "principles which Professor Lepsius has so ably sketched," (VII.) and which are to diminish " the 12 ANALTTIO ORTUOGRAPUr. difficulties encountered in the formation of a language previously unwritten," (VI.) — this treatise, as a System, is unphilosophic, inconsistent, vacillating, and superficial.* 25. Dr. Lepeiw concedes that an alphabetic system should admit of " reduction and enlarge- ment without alteration in its essential principles." Yet a uniform mode of enlargement is not proposed, and whilst 7 is allowed to represent an aspirate 2, n is not allowed to re- present an aspirate n, because I is "fricative" and n "explosive," by a false theory; nor is there a substitute suggested for the forbidden splritds aspSr mark. The diacritic marks used are not restricted to particular phases of speech ; but, on the contrary, one mark is assigned several heterogeneous values. 26. Profeaeor Lepdvs has not quoted Mr. Ellis, who is much his superior in this intricate subject, nor Dr. Latham, who would have informed him that a diphthong b not compoi^d of two vowels. Nor has he given the Latin alphabet a critical revision, if we may judge from his notion (p. 41) that the Latin diphthong (E is the German vowd o, and that (XELUm ends with German m, and that this Latin word is, in German letterp, kolum, rather than (in Polish notation) kojlu,, or (French) coylou". 27. English spelling has a redeeming feature to which the late period of its reform gives incalculable value. Its corrtiption ia ao great, that any consistent alphabet would have so many discrepancies from the present one, that the few concessions which the new could make, would be of very little aid to any one already able to read the corrupt one, whilst the drudgery of learning the irregularities of this, would be lessened but little by the form of a phonotypic one previously learned. Hence, as far as English is concerned, the new alphabet might be Greek, Russian, or phonography, because the labour of learning to read a consistent new alphabet is not great.f 28. The Cherokeea, who have a cumbersome and imperfect syllabary of 85 characters, which must be laboriously written in their printed forms, when advised to adopt the Roman al- phabet, express the?r distrust of ours, stating that the best argument in favor of their own is the fact, that when their children have learnt the characters, they are able to read. 29. English epellin^ is so irregular that any reformed orthography in Roman typography * See my Report on the Present State of our Knowledge of Lingnistio Ethnology, made to the American Asbo- ciation for the Advancement of Science, (Tenth Meeting,) August, 1866. f The use of a corrupt alphabet induces bad habits in a phonetic one like Greek. A girl of fourteen, who knew the sounds of German and French, learnt the Greek alphabet in one hour, about one-fourth of which was taken up with a work on inscriptions, to account for the writing forms; but when words were to be spelled oat, ^v was converted into English on; e?c (instead of having the initial vowel of «teA) became ice; and to words like xda/ioc with the genuine but short vowel of cSast, that of cost was assigned, (for even in the modem tongue, o and w have the same quality.) Similarly, several persons have been met with, who read the Spanbh article el like the first syllable of alley; because, Spanish e being English a, a4 must spell al. ANALYTIC ORTHOGRAPHY. 13 must present radical differences, because one mode of notation nuist replace nniiii/ modes. Hence if o is adopted with its correct power in host, the word lost must vary from its pre- sent form, and nothing in the new can recall old forms like lore, lose, and the seven or eight thousand words spelt with final e, which must disappear from the whole, perhaps to be transferred to other words which have been spelt with a different final character. Di- graphs being wrong in principle, they should not even be hinted at, as in using a character like GO to recall the old oo, which ought not to be recalled intentionally, and for ages to come. Compare door, adore, oar, four; rot, rote, root, groat, slough;* mote, moat; they, met, meet, meat, mete; great, grate; bate, bait; bite, bight; heel, heal, fealty. 30. " Writers on phonetics . . . adopt the present letters as far as they go, adding a few new ones to complete the list. They wish to retain the old letters, so that the pre- sent generation may be able to read the new way with little trouble. Grave as this con- sideration may look, it is but a slight one. A man can learn a phonetic alphabet which is altogether new to him, in a few hours; a labor insignificant in an alphabet intended to spread over the world. There is no advantage to the learner, in retaining a letter as to its shape, and changing its character. We may retain the letter e, but when we restrict if, to one of the many sounds it now stands for, we make a new letter of it. It occasioned me more trouble to remember that a particular sound belongs to the printing a, and another to the written a, than to attach those sounds to new characters, because in this latter case the other sounds of the letter a are not constantly occurring to my mind." Condensed from An Endeavor towards a Universal Alphcthet; hy A. D. Sproat, Chillicothe, Ohio, 1857. 31. English spelling can he reformed thoroughli/,where(ia, in Spanish, Italian, and German, the imperfections are fewer, and their removal less imperative. The Italian syllable qui corresponds with Latin QVi, but Spanish qui has u silent. Italian uses J nearly in its pro- per Latin sense, Spanish corrupts it to a guttural aspirate, and uses y instead of Latin J; Spanish ch is tsh, Italian ch is k, that is, h keeps the cay pure in Italian, and corrupts it in Spanish. It may be long before such discrepancies are removed. 32. The English word chew (tshoo, Walker) would be expressed by chu in Spanish, ciu in Italian, tschu in German, tchau in French, ?/y or mmy in Russian, czu in Polish, csu in Hun- garian, and wfi in Hebrew. The Greek and Latin alphabets are incapable of representing it — for in tshu, the sh should have their power in mishap, and s being already an aspirate, it cannot be treated like the lenis t, to form th. If the English word /avor were German, it would be spelt fewer; and if the Latin cor (heart) were English, it would be spelt care, as in fact it is. * As words, 'groat' and 'slougb' are uokoown to the writer, except the latter as a medical term. 3 14 ANALYTIC ORTHOORAPUY. 33. 1/ Eugliuh spelling had been reformed earlier, it would have been badly done, by per- sons ignorant of the bearings of the subject, and before a correct enumeration of the sounds had been made. Now physicMsts like Willis, Herschel and Faber, and philologists of the first class, contribute their stores, based upon a more refined analysis of the operations of speech. Formerly, had there been an educated class, (educated in linguistic science,) this class would have stood aloof until an alphabet as corrupt as the present one would have been fastened upon the language, making English the laughing stock of civilised and savage nations, indirectly checking its influence — cutting off the English people from the antece- dents of their language, whether Anglic, classic or Celtic — depriving them of the inci- dental etymologic knowledge which is suggested through the eye of a population where information is acquired by reading rather than by conversation — and surrounding tbem with a literary Chinese wall, not to exclude the barbarians, but to keep them within the circle of their abominations. 34. If Walker had used a phonetic alphabet instead of his figured notation, he would have done much towards a reform in spelling; but he would probably have allowed b-a-r to spell bare instead of bar; n-o-t to spell not rather than note — sanctioning corruptions which a better educated age might have a difficulty in removing. 35. Walker'B notation ia not chronologic, as in tar, which he marks with Oa instead of ai, or simply a, as the original power for which the character was made. A chronologic notation would run something like /«ir, ajl, whaj, faj,* aje, fa^re, (French e,) umbrella^, ma^ny, pla^it; mariiue, wiiH, wiaW, fiiV; Shang-hoB^ (-high,) Gai.}ic,CkiB^8ar; o{tm,o^,ho^or,'moji)e, wo^rk. If, with such a notation, the orthoepists had represented agiven sound with the letter having the lowest figure, the tendency would have been from corruption toward purity, and the figured pronunciation would have been a collateral aid to etymology, especially if characters which want the original power in English, had been started without the lower numbers, as in rhyjthm, {y^ being the Greek vowel, and yj the French «i,) rhy^ifne, my^rrh, y^^r. 36. Mr. Trench uses an argument which deserves attention.f He considers it an as- sumption of the spelling reformers " that all men pronounce all words alike, so that when- ever they come to spell a word, they will exactly agree as to what the outline of its sound is. Now we are sure men will not do this from the fact that, before there was any fixed * Mr. Ellis thinks that ami had the vowel of fat formerly; — that all what were not early sounds; that within three hundred years, made lade were mad lad, with the vowel lengthened; and that the historic order of the powers is — arm, fat, all, what, fare, ale. Mr. Ellis will present a history of English pronunciation for the last three centuries, in the third edition of his Plea, to be published in the United States. f EiiglUh, Past and Present, Lecture V., a production which, in sixty years, is likely to be regarded as a curi- osity, if we may be allowed to reason from the condition of chemical notation in 1798. ■^' ANALYTIC OKTIIOGRAI'UY. 15 and settled orthography [pronunciation?] in our language, when therefore every hody wa» more or less a phonographer, seeking to write down the word m it sounded to him, for" like the Hebrews, Hindoos, Greeks, Latins, Welsh, and Cherokees, " he had no other law to guide him, the variations of spelling were infinite. [*] Take for instance the word »xul- den; which does not seem to promise any great scoj»e for variety." 37. Certainly iwt, if we spell all the variations of nuhdnn (with silent h as in ««/> — f) to suit the Latin SUBiTANeus, or conform them to the French mwda'nt,-e. He proceeds to cite fourteen spellings, assumimj that they represented the modern word, and not the lost forms from which our nudden is derived. Double forms like soden and middain, perhaps of different age and locality, may (apart from the blunder of the double d) have bien as correct formerly as are now urban and urbane; human and humane; travail and travel; costume and custom; clarify, glory, glare, glair, and cJeiir; einmet and ant; decking and ticking; or brest of Wiclif, Chaucer, Shakespeare, and the lettered vulgar, bowide breast of those who know not the use of letters, according to Prisc''ian's definition. 38. Granting tlutt tJiese fourteen spellings stood for the same vocable, h iving the vowel up in the first, and of end in the second syllable, these sounds were unprovided with spe- cial characters, so that sud- might be spelt sod-, with o in xoorth, or mdd-, sudd-, some writing dd to shorten the vowel, as we spell mid sad, will wilful. Thus, sodain may have had the vowels of worth and said ; sodaine — Jolous in honor sodaine and quicke in quarrell. — Ai you like it, 1U23. the e of imagine(-ation;) sodan the vowel of many; aodayne that of says, {sayd;) sodein-e that of heifer; sodeyn that of they pure, or modified as in its derivative them, as silent b turns break into wreck. Other forms would have been justified by fri'end, jeopardy, dead, foetid, guess, panegyric, {Ellis, Plea 2d Ed., p. 155,) English being more irregular here than old English, with the difference, that the moderns corrupt a wider field with their irregularities. Abner Kneeland thus answers the foregoing objection.^: * As in the variations of the Latin word duo, which have been spelt as in two, twioo, Iwaia, twelve, duodeci- mo, (fodecahedroD, dual, deuoe, double, doubt, d (, 9<^y b V t> dh {wanting.) ^ P t t < cay f f fl th K ch W m ^ n 9 ng Here m is acknowledged as a nasal 5, n as a nasal d, and perhaps ti^ as a nasal gay. As- piration is indicated by a line, which on the left of the gay character, would have given it the sound heard in Belgian, and as this is wanting, the laws of permutation which would place it in a word, cannot bring it forward. Hence gafr (Latin caper a goat) be- comes dy afr (thy goat) instead oidy ghafr—i\iQ analogous form. ANALYTIC ORTUOGRAPHY. CHAPTER TIL ■ " • ALPHABETS, PICTORIAL, PHILOSOPHIC, AND CONVENTIONAL. > .... o'o.Ht quel'pcriture est iin ouvrn<;c encore bien itnpnrfait dcs hommcs, et quo la parule est une ctoUion de la na- ture. — Olivier, Des Sons de la Parole, Paris, 1844. § 72. It is agreed thai the diverse Latin, Greek and Hebrew alphabets have been derived from the Phenician, and that the earliest form of this was hieroglyphic, each letter bciii}^ the picture of an object whose name commenced with its power. The letter Qoo pictured the human head and neck, the neck being made as a vertical line below, until writing in tw^o directions threw it 1o the right (q, q,) or ( p ) left. R was a side face looking to the right, the tail representing the beard; but, as this was sometimes omitted, we find that r has two forms (R, P,) in different Greek inscriptions. 73. The earlier form (t ) of the Hebrew j gimel represented the head and neck of a camel (Hebrew gamal) looking towards the left, the direction of Hebrew, Etruscan, and some Greek writing; whilst the Greek ( F) gamma represented it looking towards the right. One of the forms ( < ) of this becjvme rounded into Latin Cay, acquiring a new power as readily as the word acquired initial cay in the Latin camelus, ah in the French cJiameau, and c?2/i in Arabic. Hence, 74. There are several objections to hieroglyphs. Every language would require a different set of symbols; the symbols for allied sounds would be dissimilar, and the power of the characters would vary with the name of the objects represented, until variations in the written forms would cause the originals to be forgotten, so that instead of more accurate pictures of an ox, a house, a camel or cynocephaltis, and a door, we should find the apparently conventional figures a,b,c,d. 75. Hieroglyphic or picture alphabets would be readily suggested at the invention of writing, and they are more easily learned and remembered than any other kind. On this account, a French hieroglyphic alphabet has been proposed — Lea Hieroglyphea Fran^ais, par C. Chesnier, Paris, 1843, in which a pointing finger (i/i-dex,) stands for the nasal vowel in, an aw-gel for an, the numeral 1 for un, a pink (oeillet) for short eu, a sword (epee) for the vowel of fate, a hatchet for short a, the head of an ass for a, a pipe for p, and a bomb for b, &c., with symbols for hi, pi, cr, &c., requiring fifty-five characters for the French language.* * It is applied to foreign languages in the most perverse manner, the aspirate of the Spanish word evangelicot being given as English gsh (in egg-shell,) and the nasal an is placed in tanto, and in the Greek amphi. In Italian, French na^al in is placed in denti, esempio, and nasal on in contare. In English, the same vowel is assigned f? Hi! }iii:.| ♦^•■1 m |2j| analytic ORTIIOGKAPIIY. 70. Fif/iircn r/ ffte oryans of speeth, either pictorial or mnemonic, must have attracted attention at an early period; and it is probable, that when the knowledge of the hiero- glyphic origin of the common alphabet was lost, the form of the letters was influenced by tlie position of the vocal organs, as in figuring the closed lips in B, and their circularity in 0. Pownal (Study of Antiquities) accounts for the vowel characters in this manner. I (in marine) would represent the linear aperture, the figure being turned to range with other letters. A (in Arm) would represent the mouth well opened. T might figure the tongue rising against the palate; the tongue forming an obstruction in the middle of the mouth; ^ a similar obstruction by the two lips, but with a vertical line to distinguish it from 0. The middle line of E (in vEin) was originally as long as the others, and might represent an opening of the mouth nearly as narrow as that of I. H was much like E, being a square with a horizontal medial line, and in some Greek inscriptions, the character H represents the consonant h, in others the vowel e. 77. Such a 6yetem is impracticable from the difficulty of figuring the position of the in- ner organs; and as the number of essentially distinct elements is not great, a pictorial re- presentation of them would be as little worthy of attention as a proposal to use the sign III instead of the numeral 3 in arithmetical processes, as being more suggestive of three. 78. An anonymous author issued a sheet from Lockport, New York, in 1853, proposing a set of characters to indicate the organs. Here b is h, its reversal a (with the apex of the semicircles angular) makes p, and a (with the curve angular) is/, the base represent- ing the lip and the top the teeth. This reversed, or facing to the right, is v; a character like m (with the left side rounded like the right) is m, and w when inverted, leaving Eng- lish wh, German w and Greek (p unrepresented. D i» taken as the base of the dental let- ters, the curve being the palate and the stem the tongue. Yet, whilst w is a nasal d (as 7n is a nasal h) the first and second lines of N are assumed to represent the nose, and the third line the tongue. 79. A philosophic alphabet would represent the same phase of Bjjeec*. in the same man- ner, arid A. D. Sproat has endeavoured to accomplish this, as in i i, i. d, ^ n, t th. Here the base line indicates vocality, the angular one aspiration, and the medial one na- sality ; but the n is discrepant, it represents a surd », it wants the base line to make it in- dicate the common sonant n. This system has a shorthand form. 80. Htman'e Phonography has a philosophic basis, as far as this is compatible with rapid to /or, of, none; men is men (mane,) have is ^v, and has nasal in, Imth is §s, the has French z; despised is des- pa-ist, with pure st; others in French orthography would be itzzoeurss, and Goldsmith and GBthe ought to have spelt their names Golshmeet and Got. In German, ench is made up of short French a, long au, and French ch or English sh; zu is made (in French spelling) the impossible tzou; German, English, and Greek initial h is silenced, and Greek 9 x are turned into k, t. ANALYTIC ORTHOGRAPHY. 25 writing, and it might perhaps be adapted to print.* But as it is an essential feature of shorthand, that every available sign shall be employed, that for English th would be as- signed to some other sound in a language without this lisp, which would destroy the uni- formity of notation between different languages. 81. Script and print are eesentiaVy different in this, that as facility in execution must be a primary object in writing, the most complicated character can be printed with the same ease as the simplest one. But, notwithstanding this feature, a uniform notation for writing and print is perhaps desirable. The two kinds of common print, roman and italic, are copied after manuscripts, and the forms of written and printed Greek do not differ. 82. The Cosmo8ph(»iography of Gouraudf is an attempt to construct a condensed writing character, which may be printed with separate types, specimens of which he gives. The author is said to have been a fluent lecturer in French, Spanish and English, but he has made no critical observations on pronunciation.^ 83. Of conventional alpliuhete, the Cherokee is a good example. Sequoyah the inventor had a book in the European characters, which, as he inferred or was informed, conveyed intelligence, but in a mode as obscure to him as the Egyptian hieroglyphs to father Kir- * Henry M. Parkburst (Ploughshare, Washington, June, 1853,) has proposed such a "Gosmophonetio Alpha- bet." His alphabet is inconsistent; because, for example, surd and sonant marks were deemed necessary for j), h, but not for w, m. f Practical Gosmophonography : a system of writing and printing all the different languages, with their exact pronunciation, by means of an original Universal Phonetic Alphabet, based upon Philological Principles, and re- presenting analogically all the Goniponent Elements of the Human Voice, as they occur in Different Tongues and Dialects : and applicable to daily use in all the branches of business and learning, illustrated by numerous plates, explanatory of the calligraphic, steno-phonographic, and typo-phonographic adaptations of the system ; with speci- mens of the Lord's Prayer in one hundred languages; to which is pretixed a General Introduction, elucidating the origin and progress of Language, Writing, Stenography, Phonography, etc, etc., etc., by Francis Fauvel Gouraud, D. E. S., of the Royal University of France, New York, 1850. \ In his opinion (p. 76,) there is an "absolute identity" between the English an, 1-en-t, f-on-t, s-un-k, and the French nasal vowels an, in, on, un, respectively. He assigns the C'eltic vowel mfat, to French, German, Ita- lian, &c., and he considers the English ou in fount to be the vowels in nor and put. He says of the French vowel in peu, vaeu, that it is " a sound which the English learners of that language generally think so difficult to pro- nounce, although they use it a hundred times a day." He assigns the French vowel in caur to English cur, and finds French u in rapturous. The numerous versions of the Lord's Prayer are given in their peculiar orthography, without pronunciation or translation, so that such series of Chinese or Cherokee characters must be useless to the great mass, even of philologists. No. 33 is a specimen of " Gothic" in Gothic characters, with some of the words improperly divided; No. 61 is "Mceso-Gotbio" in Koman letters, being the same thing. The latter is credited to Ulphilas, the former, in the Ulphilas character, to Stjernhjelm, who gives a plato of it in his version of 1671. The foreign alphabets are in bad, and often inaccurate lithography. Some of the versions commence with the prayer, as the Hebrew, Irish, Armorican, and Croatian ; others commence with the verse (Matthew vi. 9,) as Gaelic, Welsh, llussian, and Cherokee, so that comparisons may be thwarted at the oommencement. LiM^- ^v-v^ m* ti l fm rm 26 ANALYTIC ORTnOGRAPnY. : Ifil cher or the characters on a tea-chest to a London grocer. He used them in a syllabic sense, varying their forms, and adding others to complete the number eighty-five. Here K became tso, and J coo, which latter is not so bad as making it zh or dzh. The second and fifteenth word of the Lord's Prayer in Cherokee, is, in French orthography — ca-l'un- la-ii, (heaven,) but with German flat k and t, the last vowel as in English pit, accented, the o in art, and the second syllable exactly the French Vun (the one.) In Gouraud's Transcript, No. 30, this word stands first in the third line; and the third from the end of the first line. The characters are read towards the right. 84. Although the Cherokee alphabet is syllabic, beginning with a consonant, as lo, tlo, tso, a word may begin with a vowel, so that there are vowel characters, as d in arm, r in yeva, t in field, &c., and this being the case, it may seem singular that the inventor did not fall upon a strictly alphabetic notation, seeing that, when writing W la, d le {lay,) r li (lee,) M lu (too,) he might have used wd, ^d, or rD, for la; wr, &c., for lay; and wt, &c., for lee. 85. But there ia a great difficulty in getting an abstract idea of a consonant, as distinct from a syllable. The con-sonant 'P' is nothing when alone^ 'L' is something. But pa and la are alike in termination, with an initial difference. Their notation must be ana- logous, and if syllabic, it can be appreciated. But if the initial and final effect of la have each a character for the sounds which are so readily appreciated, pa must have the same a final, whilst it has nothing corresponding to I in the sense of an element which can be pronounced independently. The p' o£pa cannot be detached from a, it is a nullity with- out it, pa must therefore have a single character, and if pa, so also la. 86. The same course of reasoning perhaps, causes Dr. Lepsius to assign single characters to the Hottentot clacks, which are made with a consonant position followed by a vowel po- sition of the organs; — to term m, h,p, equally explosives; and in fact, m is whatever h is, with nasality added, differing as a nasal vowel differs from a pure one. If then, h is an explosive, so is m, and if m is not explosive (and it is not) neither are h, p. 87. Those who term P an explosive, take the Tsa-la-ki view, mistaking two phenomena for one. P may be compared to a gate in a water course: if quite open, the water flows like a vowel sound, if let down nearly close, the flow may resemble that of /or th; if closed entirely, or closed or opened suddenly, the gate acts like P on the current of the voice or breath, or like B, should the water continue to gurgle and dam up behind the obstruc- tion; or like M, should the stream flow over the gate, or find a side passage; and when the stream issues suddenly, in an "explosive" manner, it is the current, not the gate or obstruction, which is explosive. /■ ANALYTIC ORTIlOGRAPnr. CHAPTER IV. THE LATIN ALPHABET. At present, ancient Latiu usages are the only feasible basis for an alphabet that the learned in all nations can use ; the letters, as far as possible, having their ancient' Latin values. — FAlh, Universal Writing and Printing, Edinburg, 1856. J!ho life of all language is pronuntiation. — Rogtr WUliama, Key into the Languages of America, London, 1043. La prononoiation est la chose la plus importanto dans r6tudo d'une langue .... La prononciation est ii une languo ce que les couleurs sent aus figures d'un tableau. — Robello, Orammaire Italienne. it will be found upon critical and candid inquiry, that much, which at first sight strikes us as barbarous, is only ancient. — Pennington, An Essay on the Pronunciation of the Qreeli Language, London, 1844. § 88. Moat of the languages of Europe for which the Roman character is used, preserve the original power, except that the greater number of sounds in some of the modern lan- guages prevents each of the characters from being restricted to a single power. 89. The cJuiractera of the Latin Alphabet are the tioenty following: — A, b, C, d, e, p, G, H, I, L, M, N, o, p, Q, K, s, T, V, X; and of these, nhie had the same power as in English, namely: b, d, p, h, N, p, Q, t, x. 90. The names of the letters, according to the ancient grammarians (Schneider's Gram- matik, Berlin, 1819, p. 2,) are, for the vowels, their power, and for the consonants, the following syllables, given in English spelling, — bay, cay, day, aif, gay, hah, ail, aim, ain, pay, coo, air, ace, tay; to which Schneider adds leah, and I consonant, V consonant, these being called by Eichhoff (in English spelling) yee, and way. Sometimes Greek K was used in writing calendae; and t, z, appeared in unnaturalised Greek words, with their Greek power of French M, and English 2!. v., 2d ed. 1854. {114. Figure, Name, Power, As in Figure, nnme, power, as io Aa dlifa a arm art. ^ Nu w n noon. /ts(i (ir^ra b 6ay. , ' s$ Vi C8 axis. rrr rd/t/M g.ng ginng. ; Oo S /tixndv obey. jd diha d deW. II n vi JTl P pea. lit i ^^r;6J/ e epsom. f'i>m fxo [rh as in Welsh.] ^: Cr;r« zd visJom. 2' «T c atY/m «eek. . tt^ ^ra e there. > 7'r7 Ta'j «e, ro6«on. ' M fi /fj m moon. Qoi (L /tiya u own. 114a. Ov, 01), «, properly a diphthong like o-w in no-wonder, which should be preserved. At an early period it was pronounced both by Greeks and Romans, like French ou, La- tin U, the 00 in fool. 115. ' sriRiTUS ASPER (rough breathing,) English A, placed over the second character of diphthongs or digraphs, as 06 where, read ho-w or hoc. The (*) spiritua lenls (smooth breath- ing) indicates the absence of the rough breathing, as in the English owe. It is not indi- cated in inscriptions. 115a. As it is hardly possible to commence a word with a vowel, without allowing a little inaudible breath to pass before the vocal ligaments begin to vibrate, this, as Chavce suggests, may be the smooth breathing. 116. Aa zd haa (lis single cJiaracter z,* so its cognate si is sometimes written with a single letter, as in i^pov or darpov (a star.) In writing the Doric and Eolic dialects, ^ was re- placed by ah, as if the double sound varied from that in wi«(/om to thf.t in miscfeed. The character a is used, except as a final, to which c is restricted, as in ^/o<5oc, (wise.) * See Haldcman. Investigation of the power of the Greek Z by means of Phonotic laws. — Phonetic Jour- nal, Sept. 24, 1853. 32 ANALYTIC ORTHOGRAPnT. liii 117. The characters E, n, P, x, have not the same power in Greek and Latin, which causes great inconvenience, and tends to prohibit the use of the proper Greek characters, for manuscript forms, most of which arose in the 7-lOth centuries.* This difficulty should be removed by using e or 6, for which authority may be found in Greg. Placent. p. 106, plate; and in the Elementa Epigraphices Graecae of Franz, Berlin, 1840, p. 245 below. P should have the upper projection cut away, the angle rounded (P,) and the curve thick above, and tapering downwards. H might have the Coptic form (H) and (X) would be nearly the Coptic X. 118. r, Y, r, before y, x, {, jj, has the proper of ng in sing or n in ancle, angle, as in dfxukoc (curved,) Latin angulus (an angle.) Words like sing cannot be represented in Greek and Latin, because the ng sound is not made except in connection with a following guttural. In these pages T will be used for the nasal sound. 119. is written with ph in the Roman alphabet. It differs from F in not being made by the lower lip and the upper teeth, but by the contact of both lips, as in blowing. 120. V was originally a Greek letter with the power of ooze, and from this the later Y, Y (French w) seems to have been formed, either to indicate the pursing of the lips by the contraction of the base, or to show its relation to I. Y had not the pinched sound of French u in the Eolic dialect, nor as the second element of the labial diphthongs; hence av agrees with English ofw and German au, in brown, braun. 121. Diphthongs, Ai as in atsle; or like o-y in go-ye; «>( the same lengthened; a like e-y in get-ye; »7t the same lengthened; and in all cases, the first element has its proper power. ACCENT. 122. T7ie accent of Greek differs from that of Latin in falling upon the last syllable, as well as upon the second and third from the end. There are three varieties, the acute ( ' ) and grave ( \) used with long and short syllables, but the grave restricted to finals; and the circumflex, (" ~) which is a union of the others, used with long final or penultimate syllables. 123. Th^ acute accent indicates the chief stress, the grave a secondary one. A word bearing an acute accent on a final syllable, may have it changed to a grave in the middle of a sentence (as being weaker among other syllables,) although the acute would be pre- served at the close, as in the English sentence (writing detain in Greek characters) "/ *See Epitome Gbaecae Palaeografhiae et de beota Graeci sebmonis fronunoiationk dissebtatio ATOTOBE B. P. D. GBEOOBio PLACENTiNio, ROMAE, MD. 00. XXXV. This work IB abnndantly illustrated with figures. t lialdeman, Proo. Aroer. Acad., 1849, p. 171; Castanis, The Greek Exile, Philad., 1854, p. 246; E. A. So- phocles, Greek Alphabet, 1854, p. 113 — 14. ANALYTIC ORTHOGRAPHY. 33 will iwt dtrijv iny longer^ I will not divjv. So the second syllable of renewed is acute com- pared with the first, but '*f we say " buds are renewed 6very spring," it becomes grave in comparison with the acute accent of a'cry. 124. In strict accuracy, the acute accent seems to have been rather at the end of the vowel or syllable, the grave at the heginning, and the circumflex in the middle, corresponding respectively to the crescendo "•==:, the diminuendo r=-, and the atoell -=::::=— in music. The following are offered as English approximations: — sea-dog, seed-ing, strai-ning, cara- van, caravansery, careful, elecampane, undeviating, unconstitutionality, incontestibility. 125. As English Jias sounds unknown to the Greeks and Romans, it would be difficult to find a line of English which they could represent or read correctly if written in their alphabets. For example — " The proper study of mankind is man," .. .. t np..np aT.,8.. .. fji..uxatu3 ,. fi..v — DHI PR..PR ST..D.. .. M..NCiEND .. M..N — cannot be written, because the power of th in tJie, the vowels of study, the vowel and v in of, the vowel and z sound of is, have no proper characters, and the existing ones do not allow of the English latitude of power. Similarly, the line — Those things Aanging trithi/i — contains but four letters (o, h, y, and final n) which would be written and read by a Ro- man in this connection. In the following examples, the Greek, Latin and English elements are nearly identic. arm hold pure bind hero cone scheme town sweet useful wine fed dpfi 6X3 ntup ^axvd "put xwv axcp. rauv ann to6ff,.iiX uaiv ..ed ARM HOLD PJUR B^END HIRO CON SCIM TAVN SVIT JUSFUL V^N F..D Here, the Greek t, a, being properly vowels, mup and ohit admit of being read as dissyllables, so that they are not true representatives of pure, atoeet, nor would the Latin forms have been, before the modern separation of I, J, and V,U. THE DIGAHHA. 126. The inconvenience of one letter for the sounds of ooze and well, although not felt by some who have proposed English alphabets, was appreciated to some extent by the ancients. The sixth Hebrew letter vxm (in wound from toind) was represented in archaic Greek by the 'digamma' f (the original of the Roman F,) and it is possible that in some dialects this had the power of German W and Ellenic (Romaic) /?, the sonant of ^ § 1 19, that is, a con- sonant akin to English v, but made with the lips alone. 127. Wis the proper character for this aspirate b, it was made for it, and is still in ex- tensive use as its representative. " W is of German origin, and occurs first in the name of Witiges, anno 536, on coins." — Kraitsir's Glossology, p. 98. 1 i r' i 34 ANALYTIC ORTHOGRAPHY. m 128. The elements of woo are sequents in English and Latin, as in wool, wltvrnvs, but not in Greek, where they would be likely to be submitted to a naturalising process akin to that which produced the three forms— English tool/ (=Ang. Vulf,) German wolf, and Swedish ulf. This process would be used with caution in proper names, which some would naturalise and others present in their true pronunciation. Except in the case termination, Anxtoi; is a genuine transliteration of lvcivs; 'fuA^wc and Br^po^ are naturalised forms of TVLLivs and vervs, the former with French m, the latter with 6, an interchange (English b, w,) which is common in Sanscrit. But Greek /9, «, w, are no more identic than English 6, w, and German w, in the proper name Weltzhoover, which is pronounced in these three modes (and sometimes written and printed with i,) in Maryland and Pennsylvania. 129. We can now account /or the want of uniformity in the Greek orthography of Latin names, such as — ObaXi pto<: 0Xaoi)eoz N e podaz VALERIVS SKVKRVS FLAVIVS NERVA VARRO ObdppWV AVRELIVa AbpT^ltOZ OCTAVIVS Oxraobioz iVLivs lobkioQ WLTVRNVS ObooXro'jpvoz liaXiptoz I'eu^po^ Ss^Tjpoz ^Xabtoz 0Xdt9coz Nipfiaz Bap ^wv ARIST0BVLV8 AptaroitiXoz TIBVR Tt^upa LIVIVS AeSto^ VKRRBS Bipfttjt: 130. Appreciatiny the ina/:curacy of seeming to string four or five vowels in a line (V being oo,) the Romans sometimes used the digarama inverted (to keep it distinct from their F,) writing ocTAiiAE, ser^vs (the modern servus,) and the like, to be seen in inscriptions. Dialectically, this j may have had the power of German w (Spanish h between vowels,) as we find bervm for vervm. 131. TJierewas probably a Spanish dialect ofL&iin paralleled byan Ellenic dialect of Greek, an Arabic dialect of Hebrew, and a Sanscrit dialect of some unknown original. For, in some cases, a language pure in the morning, may have sloughed off a dialectic ulcer in the afternoon of the same day, and the organs which could open sufficiently for brig and kin in summer, might close to the aperture required for bridge and chin, when opposed to the blasts of winter. 132. Th^ greatest corruptions occur when the language instinct has become enervated. Then sixt is perverted to "sixth," although forbidden by a law of the language. Then some one may say "of like" for alike, as "almost" is said for amo«< (perhaps an old dative akin to the German am meisten,) and " out of doors" for om< adoors — mistaking for a plural sign the adverbial -s of fmcards, whence, sirux, twice, else, vix, bis, 8c<:, d^ (backwards,) the u of ah (back) becoming n or tp, as in Ellenic. Compare Xa~jpo<: and }.dQpoz (violent.) ANALYTIC OKTHOGRAPHY. 35 CHAPTER VI. THE ANGLISH ALPHABET. With all the prejudices of an antiquarian taste, and an eye long familiar with the form in which the words had been accustomed to be read, in what has been called the Anglo-Saxon character, and with the difficulty of recognising the same words when presented in a different dress, it required a strong reason to justify tlte rejection of the old letters. Nothing but a thorough conviction that the Roman character would be the moat legible, and would best show the identity of the present English with Anglo-Saxon, as well as the clear analoer existing in the words of all the other Qermanio languages, would have led to the adoption of this typo. Boaworth, Dictionary of the Anglo-Saxon language, Lo!idon, 1838, p. clxxi. § 133. Anglish ortJwgraphy is nearly like that of Latin and German. The characters diflfer somewhat from the Roman, which are frequently used instead. The letters are a, b, c, b, e, p, 5, b, 1, 1, m, n, 0, p, ji, p, c, J», », u, p, y. 134. A as in arm, art, and probably as in fall, what. Compare amklsniall; stal a stall; fram from; nat not. 135. Cay always pure, as in Latin. Compare corn corn; cirnel kirnel; cepan keep; brocen broken; ceac keg, cag; ece acJie; cennan ken; cynn (Irish cine) Jein; cyning (old German cuninc) king; citte kit; cealf calf; cinne, cinn (old German cinni and kinnc) cJiin; cild (old German cind, kind,) c7u7rf. 136. e, e, in they, met; a as in fat. Compare fatt/a<; %0Et tlmt; aepl apple; hoEbbe Juxve; band band; paSn wagon. Care should be taken never to use » for this letter, but (if the proper type is not at hand) to file off the right hand side of the Roman letter. This would form the basis of a good letter for the vowel in fat; whilst the use of the unaltered Roman letter would tend to corrupt Latin. 137. P, (u at a later period probably as in of, vine (its Welsh power.) Compare ofer, ouer, oi?er; efen even; lufe, luue, loue, love; hafe, haue, Jiave; Mer fever; MJive; fers a verse; ff as in 0^ its Welsh power. In Belgian, v often replaces English/, which is a Devonshire peculiarity. Compare Belgian, Anglish, English, Oerman. Toet fot foot fuss vloer flor floor flur vrij freo free frei geven gifan give geben. 138. Gay pure, as in get, give. In gear year; gearn yam; geolca ydk; geolo yellow; ciorl churl; cealf calf; the element after the initial is probably English y, which remained in the Eoglish yolk after the g was lost. Some regard ge as equivalent to English y, but as 'guard' is, (or was) provincially gyard, and *cow' iscyoio; Anglish geard (a yard, gard-en) was probably gyard, or in Latin letters — gjard. 36 ANALYTIC ORTHOGRAPHT. I !' I! 139. O perhaps as in ii^t. Compare the double Anglish forms mon and man, lond and • land, sond and sand. 140. S, r, doubtful; perhaps pure in some dialects, in others as in izeal, misery, a Somerset (zomerzet) and Devonshire form. 141. V as in thm. The Greek 6, (?, may be subtituted, or i. 142. D, %, as in then. When this type is wanting, a may be substituted. The sonant and surd ih were interchangeable to such an extent in the various dialects, that the let. ters of both fell into Snglish th, with which Anglish words are often written without taking the difference into account.* 143. p English w, and represented by both w and v, but as the letter is a manuscript and italic form of Latin Y, with the second line turned into the stemt and as it has no connection with Germanic W, v is its proper representative. The following may be compared, in which the Gothic initial probably agrees with the others. Latin, Gothic, Anglish, English, VENTUS vinds pind wind =vind VELLUS vulla pul wool sVUl VIDUA viduvo pidpa widow=vido VOLO viljan pyllan will =vil VERMIS vaurms porm worm ssVBTjm 144. Dr. Bostcorth virtually admits the necessity of measuring languages by the same alphabet — sounds by the same letters; but his use of W (where Diefenbach, Kaltschmidt, and others use V,) removes Anglish from Latin and gives it a forced and unreal resem- blance to German. On the other hand, some will have *cinn' read like chin,, to bring it down to the English level, by removing it from its cognates, the Belg. kin, Gothic kinnus, Greek r^vwc, &c. The Latin V is used in the next examples — Latin, Anglish, English. A VAD-ERE vad-an wad-e =:ved ;. , VOLV-ERE vealov-ian wallow =vblo VAST-ARE vest-an waste =ve8t. 145. Y, y, has its proper power of French u, German ii. The dot indicates nothing. It is not placed over the small *. 146. cs is preferred to x, and c? to qu. 147. In the change from Anglish to English, the derived language often retained old forms which were allowed to become corrupt in the original. The English toagon is older than the Anglish paen, (as if wine?) whence wain; and the modern rain is precisely the Anglish *ren,' a corruption of *regn. *The English use of th for two sounds recalls the Greek double letters, which had different powers in different dialects; ^ being k< or x<; ^, «■(, /3(,^(; and ^, zd, td. Without a similar reason., the Oreeks would hardly have used such an UDphilosophic mode of writing. tSoe Einman. Thesauri, Inscriptiones; Coloniae Brandcnb. 1671, p. 414, and many old books. ANALYTIC ORTIIOGBAPHY. 37 CHAPTER VII. OKGANS OP THE VOICE. Ce qui doit encore r^suUer de ces considerations c'ost I'admirntion qu'inspire ce ro^oaniHmn morveilleiix du plus parfait de tous les instrumcns, I'organe de la vois. Alil sans doute, il a pour nuteur lo plus ptirlait do tous los artiHtes. — Abbi Sicard. § 148. The larynx is the organ of voice. It is composed of five yielding cartilages united by ligaments, and various muscles, forming a mass at the head of the trachea or windpipe, of which it is a continuation. Although large enough externally to render the front of the neck more or less prominent, the larynx is reduced within to a narrow opening, ex- tending front and back, named the glottal fissure (rfi.na glottldis.) 149. Each aide of the glottal fissure has an elastic band with the inner edge (next the fissure) free, and the outer edge, as well as the ends attached to the cartilaginous frame- work. These bands are the vocal ligaments; they have no independent power of vibra- tion, but are as passive as the reed of a clarinet, until acted upon by a current of air. Their tension and length vary in speech and song, but they are never quite relaxed. 150. Wlmi the larynx is in repose, as in ordinary breathing, the glottal fissure is widest at its posterior end. In this condition there is no vibration, even with increase of breath; to cause vibration, and consequently voice, the glottis must be narrowed to a uniform slit, (Willis.) The singing voice is due to a greater approximation of the vocal ligaments than is required in speech. (Faber, inventor of the speaking and singing machine, in a verbal communication.) In falsetto singing, the extreme edges alone vibrate. (Johann Miiller.) 151. The parallelism of the vocal cords is the effect of volition, end is chiefly due to the action of two triangular cartilages (the arytenoid,) the anterior angles of which approach each other, and the cords with them. As every sonant element of speech requires the parallelism of the vocal cords, and every surd avoids it, there is a continual quiver of closing and opening, which can be viewed in the throat of some birds; and as eight sylla- bles (like pity, Fopocatapetl,) can be pronounced in a second, there are sixteen motions in this short space of time, not like the unappreciated trills of the tongue, but controlled and individualised by the speaker. This is about double the rapidity of the motion of the eyelids. ORGANS OF SPEECH. 152. The mouth and nose act on the voice or breath proceeding from the glottic, by means of the lips, teeth, tongue, palate (roof of the mouth,) and its continuation, the suft 6 f 38 ANALYTIC ORTUOGKAPHY. palate, or palatal veil, which bears the uvula, and acts as a valve to close and open the nasal passage posteriorly. 153. The pharynx is the cavity of the throat behind the uvula. It extends up to the posterior nasal passages, and is concerned in modifying the vowels. CHAPTER VIII. THE ELEMENTS. At the proaent day, in pliysius nnd uliumiHtry, we have no longer theor'iRts in the sense of the scliuols of the Inst cen- tury Siiuli men are indeed still to be found, but only in those departments of science which have not yet acquired a trulv sciontilio foundation; and in which, partly for convenience, partly from a deficiency of logic, such speculations uro tolerated. — Liehig, Principles of Agricultural Cheniistry, 1855. § 154. This cliapfcr beiny introductory to the succeeding one on the Phases of Words, its subject is not treated fully, but will be resumed farther on. In the mean time, the words "diphthong" and "coalescent" will be used, although the English syllables oy, /, ou, as a vowel followed by a consonant, have no more right to a special name than the syllable odd. Capital letters will be used with their Latin (or Greek) power — others as in English, unless there is a statement to the contrary. 155. The old division of the elements into two classes (vowels and consonants) is philoso- phic and proper. Those systems are unphilosophic which make three classes for vowels, sonant consonants, and surd consonants; or which separate a class or order of sibilants; or include ?, w, in an order of liquids. 156. Vowels (^vocalis vocal, sonorotts,) are made of the uninterrupted voice, the distinctions between them being due to slight modifications, chiefly of the cavity of the mouth and pharynx. Vowels ave pure (or normal;) nasal, as some of the French, Portuguese, and Polish vowels are; whispered, of which some of the aboriginal American languages afforded examples; independent (of expiration, inspiration, or voice,) being a vowel effect succeed- ing a clack; and glottal, in which the vowel is accompanied by a scraping effect along the rather close glottis. Its type is the Hebrew and Arabic ain. 157. Consonants are 'he result of interrupting the vocalised or unvocalised breath. Their quality depends upon the point where the interruption is made, and upon the nature and extent of the interruption. They are classified according to the points of contact where they are modified or interrupted. 158. The consonants of uieb, whip, and the vowels in ore, ooze, belong to the labial con- tact; those of five to the labio-dental; thin then to the lingui-dental; debt, le&n, to the ANALYTIC ORTHOGRAPHY. 39 dental or basi-dental; «eize to what may be called the oigmal contact (from the Greek let- ter, and from aij-fio^ a hissing,) for « has more affinity with t than with «/«, which, with zh, belong to the palatal contact. The guttural contact is formed by the back part of the tongue and palate, as in young, c&g. The vowels in pique, vein, ave guttural vowels. The glottal contact seems to be formed at the glottis, as in /toe. There are several glottal con- sonants in Hebrew and Arabic. The epigloida is passive, without muscles, and it is not an organ of speech, as some have asserted. 159. Tli^ fundamental elemetita are the (Latin) vowels U, A, I, and the consonants (mutes) P, T, Cay, corresponding to the lips, palate and throat, or to the outer, middle and inner parts of the mouth. When the contacts are half open, a series of intermediate consonant sounds result, which may be called liquids. These three kinds are related as represented in the diagram, the affinities running vertically, and the analogies horizontally, but as P, T, are equally close, and as A is much more open — more of a vowel than U — the affinity between A and L or R is much less than between U and V, still greater is the distance from A to T, compared with U to P. Vowels U A I Liquids V L J ., ,. Mutes 1 f Cay 160. The primary vowels, in natural order axe OU A EI, or IE A OU, and in forming them mechanically, if a tube of a certain length produces U, it must be shortened for 0, and so on to I, which requires to be shortened the most. A is the type of the vowels — the natural vowel — and the most agreeable of the whole. Closing the or- gans from A towards the throat, E and 1 will be formed : if towards the lips, and U. 161. Two complementary vouoels are wanted to occupy the spaces on each side of A, which are greater than those between OU, and between EI. These are awe, between A and 0, (formed on Faber's speaking machine by touching the and A keys simultane- ously,) and Mm on the throat side, between A and E, from the latter of which it is more commonly derived. Some, on the faith of mechanical experiments, locate tirn between O and U, thus making it a labial — a view which would vitiate philological deductions. Mr. Ellis would prefer at between A and £. 162. The secondary vowels are modifications of the primary and complementary ones, formed by a different aperture, and commonly, but not necessarily short. They occur long, whilst the primaries may become short and abrupt, or staccatoed. Any vowel is here considered secondary whose place is between those already named, as bit, bet, bat, bot, but, full. If naught and not differed only in length, the two would constitute but 40 ANALYTIC ORTIIOGRAPIIT. one vowel, and it ia worthy of notice, that whilst the secondary not has a closer aperture tlmn its primary n«?/ght, the secondary them is more ojmi than its primary they. But this seeming law would disappear with a change in our conventional nomenclature, if, for example, we were to consider foot the primary and fool the secondary. The following is a comparison of lip and throat vowels of about the same degree upon each side of the scale : odd add *^ , . owe there o-bey i them fool his foot he COALESCENTS. 163. TJie labial vowel ooze readily becomes the consonant way, and between them there is a shade of sound allied to both, but a variety of the latter, and a consonant, because it has the power of forming a single syllable with a vowel, which two vowels cannot do. Hence to connect I A U into a monosyllable, the extremes must be consonanted, making JAV (yow,) and the result is similar if the order is changed, as in AJV, JVA, &c. Conceiving the coalescents to be vowels, the ancient grammarians adopted the word dipJi- tJiong to account for two vowels forming one syllable. The labial coalescent is represented by u, w, in English, as in writing — out, house, mouse, (German aus, haus, maus.) 164. J/te ffuttural vowel pique may become the guttural liquid yea, as in mim'on, and between the two lies the guttural coalescent in aisle, ei/e, boy. The consonant relation of the coalescents is shown in the combinations how weW, my years, in which it is difficult to tell where the coalescent ends. A comparison of the former (or how-ell) with ha-well and the latter (or my-ears) with ma-years, will show their affinity. 165. A coalescfut between vowels is apt to form a fulcrum by becoming a more complete consonant. Compare (emp)7oyer with lawyer. Hence the Romans, who wrote AE be- fore a consonant in gr.plification, p. 41,) for the "precious advantage" of representing certuiu Arabic plurals hy re- versing the characters, as in dair a house, diar houses. This would bo paralleled in mu^,) Neapolitan chiano, with a cay, not transmuted from the p oi piano, but educed from the J. 216. Alliteration is a variety of induction in which an element suggests its repetition, as in peRdlx, Fr. perdrix, Eng. partridge; Latin amiTa, Fr. ) is a vowel change from a more open to a closer position of the organs, towards the lips or throat. The term is adopted from Crosby's Greek Grammar. 232. Revemon (marked < ) is the reverse of precession, and is much less common. It is the change from one vowel to another on the same side of the vowel scale, as from Latin UhsUs (a bear) to Spanish OsO; Latin dIgitUs, Spanish dEdO; Latin mIrabilis, French raErveille, English mArvel; Latin lIngva, Spanish lEngua, French lA"gue. 233. Transcession (marked >^) is the interchange of lip and throat vowels across the vowel scale, as between U and I in food, feed; 0, E, in English anow, German scJinee (=fNE;) Latin hOnus (good) bEne (well.) It may be combined with precession (x> ) as in passing from O to I, (a rare phase as in roll, reel; dole, deal; German ofir, English car;) and from E to U; or with recession (x < ) as in passing from I to 0, and from U to E, these three phases being extremely rare. 234. Anallaxis is the change from one clement to ttoo others, one of which stands on each side of it. As E stands between A and I (§ 238) it may happen that in the attempt to produce it, the organs may fall successively into the positions on each side of it, pro- ducing A-I, or (in case the I is coalesced) JE, as in the German mehr echnee (more snow) which becomes mai scJmai in low Suabian.* The following are examples from ancient and modern geographical names, assuming that the derived forms have been diphthongal at some period — ebEi.linu'" Baillo, bEthsan Bai'son, mentEsa Bentaez. 235. Upon the lahial &ide, becomes A-U or AV, as in sonus sound; old Suab. lob, German laub; Ger. korn, melone, Austrian Teaum, melaun;* French bo°t6, English hounty; * WocTier, Allgemeinc Phonologie, Stuttgart, 1841, p. 244-5. ANALYTIC ORTHOGRAPnY. 6f English bow nnd bote, and in the Irish dialect of English, where bold, hold, cold, &c., bc< come howld, &c., inOuenced hy I. 236. Reversed a N<(/Zaxt« appears in the SwcdiHh JAg compared with Latin Ego (I,) and in the modern lAlea, from the ancient YAjJEA. The following are Rliaetian examplco — tErra ilAra earth, yErmis vIArm worm, vEsi'A vIAspra tca»p. " Some words that might be supposed to be under Wa [English w, (t in far,'] ore to bo found under 0, as the syllable wa is often pronounced like o, and o like wn." {Daraija, Otshipwe Dictionary.) The latter [o-^^wn, § 2i5,) is an example of reversed anallaxis, the former (mw-ko) of metallaxis. 237. Metdllaxia is the replacement of two elements by one that is intermediate, being the reverse of anallaxis. It occurs in passing from AI to E and from AU to 0, os in Latin bal-ZEna, Italian balEna (a whale,) Latin cA VsA, (a cause,) Italian cOsa, French chOse ; Latin cA' Vdex and cOdex (a stem.) 238. The following tables of the aflinities of the primary vowels may be used in study- ing intcrmutation. In the second one the complementary vowels arc placed; in the third, the close of the organs to French u is indicated, and the probable manner in which the letter Y was suggested from its relations to the vowels V {po) and I. • krm Owe v'Ein , , . , pVll machine A A awe urn E V I E V I 239. Intermutatton being mostly in the closing direction, when U and I are reached, the recession continuing, U may become the labial, and I the guttural coalcscent. But let the vowel of the German Icuh (coo, a cow) be closed to English w, and the result (cw in qu-een) is hardly pronounceable until a voicel is interposed, when the English form cow appears. 240. If Ihe closed upon sufficiently to form the guttural coalescent, this mugt be aided in a similar manner by a vowel, for coalescents appear in no other manner in English. Hence the French cri, thus treated becomes cry, (that is, in Latin letters CRiE,) hy pre- ceaswn and epenthesis, not by anallaxis. * CastelH, Worterbuch der Mundart in Oesterreich unter der Enns. Wien, 1847, p. 13. M ANALYTIC OKTIlOORAPUr. l»li 24 1 . T/ir i-iHilcMieiil in the prlncipnl clement of a diphthong. In EUenic (Modern Greek) at; hiiH been closed to afp and a'fi, conHequently it has no coalesccnt, and consequently it in not a diphthong. 242. 'f/iere i'm a Hunt fo inter mutation, bo that it is hardly possible to find an example of a departure from A to O and U, and a return through I and £ to A, and a circuit in the opposite direction would be still more difficult. 243. Ah a V ran return to 0, and A J to E by metallaxis; and as the former can be- come U and the latter I by the loss of A; the triplets 0, U, AV, and E, I, AJ, furnish two sets of elements which circulate among themselves, apart from the more open vowels. They may be tabulated thus: — B AV U 1 AJ These relations, and those of Y and German 6 are shown in the next diagram. A 6 E AV U Y I AJ 244. Anallaxis ia older than metallaxis, and vowels precede diphthongs, so that when both occur in cognate words, those with a vowel may be considered the older, although immediately derived from diphthongs. Thus, although the Spanish col and French chou (cabbage) are derived from the Latin cavils (a stalk, cabbage,) and Greek xauXbc (a stem,) the original vowel was A, as in the Sanscrit rala5 (a stem) the initial of which is less old than the cay of the other forms. 245. MARKS OP MUTATION. •M- indicates an interchange, as 0-h-U, P-i-i-B. •{^ or H- is placed between a derivation and its primary, the crossed end indicating the root, or earlier form, f indicates a primary, a genuine form, or a true voot. 4. indicates a false original, as in J^hine, -^alione, where shine is not the true original whence «7M»ie is derived; one or both having come from an earlier form. The Greek \xld!iu) (to make a noise) is not the true original of clangy clank, because the gutturals of these are older than the palatal ^. The following are examples of precession. 246. 1. Sanscrit dvA; 2. Danish tO, Irish and Persian do, old English two; 3. Eng- lish two (too;) 4 old Nordisb tvau; 2' Belgian twEe; 3' German zwie-, Lettish diwi; 4' German zwei. ANALYTIC ORTUOGRAPHY. 50 4 3 2 1 2' a' 4' DVA to twoc • two zwio tvau zwci 247. 1. Sanscrit dAnta (a tooi a;) i!. Angl'ih tOth; 3. Greek orJoy; Gothic tUnOus, Eng. tooth; — 2'. Latin dEns; 3'. Turkish oTr (de*)wh;) Eng. tine, in Latin letters tjrs. 4 3 2 I 2' S 4' dAnta tU0 Dif, — tJFjU 248. ^ we pronounce ou of the Greek form like ou in roimd, (he word, as far as this part is concerned, will occupy the fourth place of the labial side, j,nu be a newer word than tooih, which is newer than toth, although the use of o in spelling tooth, might cause one ignorant of the sound, to suppose the Anglish and English forms to be of equal age. 249. Precession is commonly confined to one side of the vowel scale, as in most of the following examples. A -K -»-^ U -K AV. Latin frater (a brother) Gothic br06or; German brlJder; Welsh brawd. Latin sanus (sane =sen/) Belg. zOnd; Angl., Dan. sUnd; Eng. sound, with d educed fromn. Latin pal us; Isl. >:;<., Ang. pul a pool. Lat., Sp., Ital., corO'na, Belg. krOon; Rhaetian, crUnna; Eng. crou^n. 250. A -K -}- U -»~ I Here U, instead A becoming AV, crosses to I. Latin fA'gus; Angl. bOc; Ger. bUche; Eng. beech. The Rhaetian fau is from FAgUs by elision. Latin illO'c, illUc, illlc (thither.) 251. A -J- E -}- I 4^ ^. Latin alA^cSr; Fr. lE'ger; Sp. llgero; Eng. light (active.) Ang. nAther; Old Eng. nEther; Eng. neither; and (vulgarly, as if) nigh-ther. Isl. bad! (both,) old high German bethe; old Fris. bide; German be 'k' 252. A regular transition has occurred in English from A thorough h ol, and the se- condary vowel of it. This is shown by the fact, that the character *A, a, a,' used through- out the world with its proper power in arm, far, has in English acquired the power and name of the European ' E,' this in its turn has been confounded with the European ' 1/ which, by a similar perversion, has become the partial representative of an epenthetic A. 60 ANALYTIC OBTUOGRAPUY. 253. The following are examples of Latin words passing through French to English: — pax paix peace race mils raisi" raisin aqvila aigle eagle 1" ration-Is raiso" • reason tractare traiter treat domlnare dominer domineer eatio saiso" season factu" Mt feat macer maigre meagre clarus clair, ^cl6r clear acer aigre eagre balatu" O.Ger. bleat bleat. Old Ger. slafan, Goth, slepan, Eng. sleep, =sSLIp. Ger. bart, Ang. herd, Eng. beard, =BiiiD. Latin gravis, Ehaetian grev, Eng. grave, grieve. 0. French spare, Ang. spere, 0. Fris. spiri, Eng. spear. Ger. bahre, Fr. biere, Eng. bier, =bir. Latin clavis, Fr. cle/ =CLE, Persian kelid. Hung, kults, Eng. key, =ci. Sp. vinagre, Fr. vinaigre, Eng. vi- negar. Latin strata via, Old Eng. street, =stret, Eng. street, =strit. 254. The apolofjists of English spelling will observe, that these English words with I, derived from an original A through an ai or e spelling, follow neither, but represent the derived I sound in the six modes ai, ea, ee, e-e, ie, ey: — raisin* alone taking the form of plait = PLiT. This literary irregularity does not appear in Latin, where precession is equally present, as in jacio I throw; ejecto and ejicio I cast out: — capio I take; accepto I accept; accipio I receive, whence Iceep, =cip. « 255. The name of the English people, language and country, affords a good example of this change. The couatry was anglia, the adjective and personal noun of which was an- GLictJS, whence the Anglosaxon language will be called Anglish. The A of this became E in met in the Germanic dialects and old English, and the vowel of it in proper /nglish, Ital. Inglese, &c. And as ihglish is almost as old as Nglish, we find these words spelt with I in some of the earliest records of the language. Thus Craik (Sketches of Litera- ture, 1844 ; 1, 208) quotes the date 1113 for " Ingland is thjne and myne." ' Yet to this day,f this venerable /nglish language is ignored out of deference to J^nglish, (from which many of its forms are not derived,) and to the dialects of Scotland, Ireland, Yorkshire and Holland. 256. In passing from Latin to Italian and Spanish, E is usually retained, although it may become I, as in — * Walker's pronunciation — but now pronounced in the Irish mode. The etymologic spelling (so important with litterateurs,) being rait-, both in raisin and reason, the Irish mode was ns proper for the latter as the former— for English speech and writing do not follow the same laws. t February 5th, 1858. m ANALYTIC ORTHOGRAPUY. ei allevare Spanish alllviar to alleviate creatura <( criatura creature deus « dios deity ecclesia it Iglesia, Fr. 6gllse cJiurch ajqvalis (I Igual, Old Fr. Igale eqmd respondere Ital. rrspondere to respond securus It sicuro secure. 257. A vowel may be preserved / or ages unchanged. The following are examples of vowel identity between Latin and English. obedio obey regno I reign vena vein redimo redeem precor I pray velo I veil situs seat niarinus marine verbena vervain croc-io croak arma arms puppis poop. Here the etymologic E is represented by ey, ei, ay, ai; etymologic I by ee, ea, i-e, (§ 2u4,) and etymologic 0, U, by oa, co. Thus, an orthography which represents different forms as similar, must represent identic forms as different, and must still be considered etymologic. 258. The following words exhibit an identity of vowels between old Frisian and English. fri free hi he swet sweat hir ' Itere mi me wepn iceapon iven even thi tJiee hwer where del dale breker breaJcer tema tame hel hale stil steel niar near spiri sp&ir tron throne saterdi Saturday Here a genuine I is represented by e, ee, e-e, ea. Here me is torn from its affinities Latin Mlhi, Italian MI, German Mir, to associate it with Anglish me, or perhaps French me, which is neither ME nor Ml. 259. The vowel relations of allied languages are often irregular, as in the following Flemish and English examples, which have the same vowel (o in floor, door,) in the Flemish, but dififerent ones in English. « voor fore sermoon sermon voor for doof deaf loos loose soon sun oor ear droom dream boom boom zoon son rood red stroo straw. 260. A'J and A'V have arisen in the English hide (a skin) and German haut, from the old high Ger. HUT, which took the German form at one step, whilst the English form 9 62 ANALYTIC ORTHOGRAPHY. i ' m hiwi to pas8 through the Anglish ht/d, hid. Hide is newer than haut, but not derived from it, as represented in dictionaries; nor is bound derived from bind. 2G1. A'V becomes awe in EngUsh,'hy metallaxis (§237) varied by recession from the O point. C/iato has therefore not arisen from cheio, but from a form like the German Iciiuen. The Saxon (Lower Saxon) kluven precedes the Anglish clavian (clawian) and this the English to claw. 262. A V cannot occur be/ore labials in English, as it can in German. Hence, old high German bom (tree, pole,) became baum in German by anallaxis, and booni, beam, in English. German forms like the following are unknown in English, nor are they the antecedents of the English equivalents, although often quoted as such. haufe Jieap saum seam laub (leaf) haupt head laufen leap saufen sup auf up raum room. 263. A'V cannot occur be/ore 6c (brother-in-law) and Latin socer. Having the original element ah, the Germans preserved it in their form schwager, and the Hungarians (s as eh) in sogor. The Latin took a by induction in both cases, because as an initial, a occurs about twenty times as often as 7i. The onj of the Latin soger is probably older than the palatal of the oriental form, which may have been shwacura originally. 283. Ch^-+F. WJien the old English ch {x) began to fall into disuse, its sound was either dropped, as in \}\ough, ihvo\xgh, plow, not, or confounded with /, as in tough, cough, rough, enough. So p^iy, which by permutation gave Gall and Colic to English, gave Fel (gall) to Latin. Contrariwise, the English craft, soft, after, are the Belgian kracLt, zacht, achter. 66 ANALYTIC ORTnOORAPHT. In 284. Th-M-Ch, Ph, S. Gr. opvil, gen. opvteo^ (a bird,) Doric dpvi^, gen. opt^iXoi:, — 0M(o and SMtij, to bruise, — Doric al'dua for aBdva, Minerva, — -loc for Stb^ a god, Eng. Theodore Ru8s. Fedor. D-m-G, B. Doric J« for fu, the earth; but ov6(foz (darkness) for )-vofoc is by assimilation. Aeolic Ihltplv for Jdiflv a dolphin, — adMIia).ov for adNJukov a sandal, — Ital. coDardo (a coward,) Sp. coBardo, partly influenced by o. 285. G-M-B, — C-M-P. r^X"'^ Attic li)ji'/u)v pennyroyal, — //tiavoc and A'uavoc a bean. 286. T-M-P, C. Aeolic afldutov for a Tddcov a race course, — Latin VeTulus (old,) Italian veCchio. Although T is more easily formed than Cay, if the number of the latter greatly predominates over the former, the rare occurrence of Cay derived from T may be the result. In a paragraph of Hauaian containing 160 consonants, 28 per cent, were cay, whilst a Latin paragraph furnished about 9 per cent. The former example contained no T, so that any word coming in with this sound would be likely to fall into cay by induction. ASSIMILATION. 287. Assimilation is the cJiange of a consonant to adapt it to another with which it is brought in contact. The n of in becomes m before jp, h, m, by assimilation, as in im-plore, im-bue, ira-mense, but remains unaltered before /, v, w, as in in-fect, in-vert, in-wall. a. Latin n always became ng before gay, cay, ch, q, as in in'^certus, Tn'genuus, an'chlses, in'qviro, (§ 101) these words being cited for it by the ancients. 288. Latin had a peculiarity still preserved in Italian, of doubling a consonant as tt in attendo, and nn in annuncio. One of these consonants is in most cases absorbed in Eng- lish, as in attend, announce, in writing which, the second character is a mark of shortness for the preceding vowel. There is but one /in a flfi n i ty, Fr. a flf i n i 1 6, Sp. a f i n d a d, but the Spanish alone shows its etymologic relation to the Latin affinitas (gen. affinitat-is) and Italian a f f i n i t a, because there is no dissimulation about it, no misrepresentation, it pretends to nothing but what it is entitled to, and claims no addition but that of voca- lity for the t. 289. If 'accept were a Latin word, it would be written axept; but its prefix ad, (which became ac before cay in ac'^cepto,) became s before an a sound, as in as-s6ciare (to associ- ate,) so that assept would have been the Latin form of the English word, and in fact, the true English form, because ad- stands in inscriptions unassimilated, as in adcenscs, ad- FECTUS, and as the assimilation was a departure from the true form which could not be transplanted into English, the attempt should not have been made. dissimilation. 290. Dissimilation ia the reverse of assimilation. It prevents unusual combinations, and is due to induction. MF are incompatible sequents in Italian and Spanish, where they ANALYTIC OKTHOGRAPHY. «7 break the law of assimilation and transmute (§ 273) m to n, turning nympha, symphonia, into ninfa, ain/onia. 291. In Italkm (as in Latin) mm are compatibles, as m commissione, commismrio; whilst in SpanisI), one vi is dropped from comision, comimrio, as in the English eqtnvalents com- miaslon, commissary. When one m is not absorbed in Spanish, the n is unassimilated, as in conmoci-on cotimiseracion, conmemorar. Dissimilation occurred in Latin, for although w/ occurs in the original o( circumflex, we find an- for am- (ambi) in anfractus (a turn;) and the inscriptive forms circvnflexvs, circvnvenio, circvndata. 292. The Oreeks spontaneousli/ rejected two aspirates in certain cases; hence in 0pi^ (hair) became T in the genitive case Tfuxo; in consequence of the presence of X- So Ifix^o (I run) is dpizoj in the future tense; and Tpiifu} (I nurse) is Bpifot. The -ish in the words Engl-ish, Span-ish," seems proper in Belg-ish, with gay; but if corrupt dzh is used, this Belgish will give way to Belgian or Belgic; whilst Bussish is rejected for Russian. 293. The English ordinal suffix -th in four-th, nin-th, is -d in thir-d, and -t in fif-t, six-t, in the speech of those in whom the language instinct has not been effaced. In old Eng- lish we find first, second, third, fourth, fift, sixt, seventh, eight, ninthe, tenth, — eight being due to the aspirate once present in this word, which with its loss, could take th in eighth. GLOTTOSIS. 294. Olottbsis* is an organic change to facilitate ease in utterance, and it depends greatly upon the number, order, and frequency of occurrence, of the consonants concerned in it; practice making that easy in one language, which is difiicult for those who speak another. 295. Aa the base of the tongiie has less room and is less flexible than the end, it is more difficult to adapt it to the production of its peculiar consonants, so that children replace them with dentals and palatals, saying do for go, and til for kill. 296. The cavity of the mouth being set for the following vowel whilst the consonant is about to be formed (§ 203,) the closer aperture required by the vowels of key, get, gay, af- ford so little room for the action required to produce their consonants, that there is a ten- dency to use the outer portion of the tongue, which is thinner and more flexible, and has more room in the outer mouth. This action, which is often united with cyclesis (§ 207,) converts gutturals to dentals and palatals, particularly before I and E. In some cases, where orthography is not properly understood, this has perverted characters made for gut- tural sounds, to enervated powers (usually called soft,) in various modern languages. * Glottosis, as a word, is formed from yXciTTa, the tongue, by analogy with certain names of diseases, (amau- rosis, pyrosis, phlcgusis,) — this being frequently as great a defect in speech as stuttering, which is classed with diseases. As the word language is applied to speech in general, because the tongue (lingv^) is its chief implement, so glottosis is proposed for organic transmutation between all the contacts. 68 ANALYTIC ORTHOGRAPHY. I 297. Ckmxpare GreeA; Greer^e; ore ttxch; haxh bargre; Latin legibilis, Fr. lisible, Eng. le- (/ible. The Englisli ttih is commonly replaced by te in German and «/* in French, as in Lat. Camera, Eng. c/tamber, Ger. dimmer, Fr. c7tambre.* 298. This change ia widely spread, for although the speech of different countries may vary greatly, its expression is due to the same organs. Volney remarked it as a dialectic peculiarity of Arabic ; and Morrison informs those who wish to use his Chinese Diction- ary, that words like (c7t in chip,) chang vary to tsang; and that h in the Peking dialect, " before e and i is pronounced as ch and ts; thus Icing is turned into ching, and keang be- comes ts'eang." Morrison does not state whether k becomes tsJi before i, and te before c, with any degree of uniformity, as in Russian, where, in certain inflexions, k becomes ts before i, and tsh before e.f 299. L -M- B. These two consonants are made so near the same point that they are rea- dily transmutable, and to such an extent in Hauaian, that they are used indiflferently. R is wanting in some languages, and L in others. L-M-R. Sp. milagro miracle " papel paper " peligro peril sabel sabre esclavo Port, escravo eneldo " endro (dill.) a It S^H•T, D. Ger. hass Juite " aus out " weiss white Dan.aadike Ger. essig, (vinegar) Ger. hat, Dan. har, has Gr. ^68ov Lat. rosa rose. 300. Interchange of th, sh, zh, r, 1, n, d, t, s, between ancient and modern geographical names. ALAMATHA Elamora PONTES Ponches Fr. ARAvsio Orange CHARADRUS Calandro CALiFFAE Carif(6 BERGUSIA Balaguer LACARiA Lancona ORONTES Eluend METELis Missil PALURA Balasor. 301. In consequence of the projecting jaws and teeth (prognathism, g pronounced,) of the * Mr. Ellis writes several notes, the purport of which is, that " tsh descends from k vi^ kj historically, and dzh from g Ti& gij, as also tsh dzh descend from tj, dj, as in nature, verdure. ... I think wo can as well helieve fg to have become tsh in Sanscrit as in Italian. . . . Wallis (1653) analyses sh, zh, tsh, dzh, into s-j, z-J, t-j, d-J, and Smith (1568) shows that the former are nearly related to the latter in sound. . . . Salesbury (1547) gives si as the near- est Welsh for sh, resembling it, says he, as copper does gold." f Grimm's Geschichto der Deutsohen Sprache, §382. r ANALYTIC ORTIIOGRAPIIT. 69 African race, it is not easy to place the tongue in thy proper position for making ih, even when English is their vernacular, so that it is often replaced with / as in muth, noffuiiif, &c. This renders th doubtful as an African element. Shakespeare's 'Moor' being a negro, his name, to have a rational form, must bo Otello, as the Italians make it. 302. There are four or five times m many Italian words in pia-, lia-, chia- (cA as k,) as in plo-, fla-, cla-, showing a preference for the former. This partiality caused the elision of I and the insertion by induction of I or J, rather than the transmutation of /. This from Latin produced the Italian forms — FLAHMA fiamma fiame pluma piuma phime CLARUS chiaro clear planus piano plain PLANTA pianta plant pluvia pioggia rain. 803. In the last example the corrupt g (in gem) is made from English y in pluvja, the V being lost, and the second *i' inserted to aid in spelling the corrupt g. The loss of V and the change of I to J (as in passing from fil-i-al to fil-ial) is the only diflerenco between the ancient geographical name salvia, and the Italian form Sagliu=sx-LJA. This irre- gular Italian orthography disguises the close relation between the ancient and modern geographic names — PAL-A-Ni-A Ba-la-gna OL-U-US 6-glic PAL-Li-A Pa-glia AL-Bi-Ni-A al-be-gna Hos-Ti-Li-A 6s-ti-glia SE-Ni-A Se-gna TER-BU-Ni-o X Tre-bi-gna co-LO-Ni-A Cologne, Fr. His-PA-Ni-A Es-pagne " BRi-TAN-Ni-A Bretagne " 304. That elision of L and epenthesis of I or J are concerned in flamma, fiamma, is proved by the Spanish forms, where both L and J (written H) are lieard, as in llama (flame)=UAMA, or in the English collier for coaler. Italian, piano pieno chiave piantaggine 305. By taking Portuguese into account, we find a newer form in which PL-, &c., are lost, and the J converted into French ch (Eng. sh, or dialectically into isfi,) by glottosis — 10 Latin, PLANUS plain PLENUS full CLAvus key {-FLANTAGINIS plantain Spanish. llano lleno (& cheno) Have llanten. 70 ANALYTIC OHTIIOGBAniY. Lnti'tif ItnlttiH, Spanithf Portmjuenc, CLAMAKK fo cry cliiumurc llamar chamar I'l.UMitu'" lead piombo plomo chumbo I'LORAKE to lament llorar chorat" PLAGA a hlmo piaga Uaga cbaga PLUVIA niin piovdre Hover chover. 300. A union of three vowels, as ate, or da, is contrary to tbe genius of English and its antecedents, and when, by the elision of a consonant, three vowels are thus brought to- gether, and the intermediate one is I or E, it first becomes J, and then perhaps a palatal, as English or French j. It is not, as we are commonly taught, the B of the Latin rablds that becomes zh in the French ratje, and dzh in the English rage, but the I. TU'ts is cotir finned by the Rhaetian form rahjia, in which i indicates corrupt dzh. The supposable intermediate steps between Latin and French (the first and fourth column) aro given 1: 3re in Latin Letters, but ahbreviare is not a classic word. ABREJAR DaUJE RAJE SAJB CAJE SEJE RUJE GUJE If the elided B of rabies had been D, rar/e &c. would have been examples of partial metallaxis (§ 312, 313,) the D tending to draw the J into the palatal contact. 307. As sa.Aa made French sauge (the plant sage) with a sonant *g' due to the sonant Iv of the original; and se..ia made secJie with surd *ch' due to surdj) of the original, we may account for sonant zh in fusion, and the surd eh in mission. 308. Although mission, nation, with sh, are derived from the French miss-i-o°, na-ti-o", (nasio") with 8; and fumon, with zh, from fusio" with z, there is no transmutation of *, /, z, to the English palatals, the French consonants being lost, whilst their influence remained. 309. Those go upon a false assumption who think they are justified in using c as an al- phabetic character for sh from the analogy of ocean. It is the e which is the real sh here; and the t in twtion has as little to do with the same sound, as the p of s^ia in seche, or in the Old French pipion, which, as an English word, is pronounced j^i^eon,"" as the Italian storion-e is pronounced sturgeon in English. * See Paradoxes 1 and 6, § 41 a. , ,. abbreviare ABRE..IAR DILUVIUM" DILU..IU RABIES RA..IES SALVIA SA.. ..lA CAVEA CA..EA SEPIA SE..IA RUBEUS RU..EUS Sp. gubia GU..IA a..br6ge'' > auridge s deluge deluge rage > rage > sauge X sage cage < cage seche cuttle-fish rouge ruddy gouge > gouge. ANALYTIC ORTIKXlRArnr. 71 310. The won! *o(canlr' (with «) is ohhr than *ocenn' (in two HyllftblcH.) nnd is not de- rived from it; and when both arc pronounced with nh, this sound Ls represented hy *«' in 'ocean' and by *ce' in 'oceanic,' where *e' does double duty as a consoiuvnt and a vow(d. The word ia more correct when pronounced o-«e-an-ic; so is pro-nun-si-a-tion, because making ah out of «/, elides the voted power of *t" and reduces the \vord one Hyll.iblo. 311. If, hy the converHoii of i into English y or zh, o-be-di-ent becomes o-bo-dyent (the writer's mode of speaking,) or o-be-dzhent, no speaker of real English can preserve htth dzh and i; yet Walker has coined a jargon with such forms as o be-je-ent, and cris-tshe- an-e-te. Similarly, if 'omniscient' has an «, it has four syllables, if «/j, it has but three. Compare the dissyllables Russia, Asia, conscience, and the trissyllables militia, malicious. METALLAXIS (§ 273) OF CONSONANTS. 312. Sh hehuj made posterior to the s position, and anterior to that of imj, it may happen, that in the attempt to pronounce the combinations N-ch (ff-;^,) ' * 1'.,' ■ r* • 72 ANALYTIC ORTIKWRAI'IIY. I CHAPTER xiir. ETYMOLOGIC BEARINGS. Wo iniifit not pormit oiimolvcH to lio giiidod Rolely )>y the ovo norby thogrnmnriarian either; butmuit, on tho contrary, coiidult tlio eiir. — Honnycantle, CliiHHicnl Muhcuid, No. -3, p. 32. § 315. Mr. Ellis has cahnlated (Plea, 2d ed., § 30,) that not more than ono person in 1000 Clin be benefited by an etymologic orthograpliy, and it has been asserted that all the countries of which English is the language, do not furnish five hundred etymologists. There are, in fact, more good mathematicians and good chemists than good etymologists, and whilst few chemists would be at a loss to give the rationale of their processes, the authors (Sullivan, Graham, Lynd,) of popular school etymologies, cannot explain their own examples, nor distinguish between mutation, elision, and insertion. 310. The chemist worlcH primarily with things, and secondarily, with symbols; the scho- lar does the reverse, studying symbols rather than living speech, as a deaf mute would be compelled to do. Hence Scheie de Vere'" calls the French word for water " eau (o) " a triphthong; he says most English radical words have been reduced to monosyllables "at least in pronunciation;" and that " the changes of sounds and their growth go on conti- nually, and thus the fipclliny of a language gives us the only true account of \ta first form and stihsequent historic changes. This is the principal and all-powerful argument against phonography." A perverse inference from a correct premise. " For nearly fourteen cen- turies of our Christian era but few persons in France and Germany could write, and how was it possible to judge of words and their etymology without seeing them?" Dr. Latham says — " To those writers who, denying the affinity between the Irish and Welsh, can iden- tify the Erse with the Hebrew, I apply the term nyctalopia — the power of seeing best in the dark." Yet an Irish laborer who had acquired Welsh in Wales, when asked some questions about his own language, stated of his own accord that Welsh was " a good deal like it." And yet how different: but his language instinct had not been extirpated, and he could grasp the relations as readily as an American savage can disentangle an etymo- logy in his vernacular. 317. The Dictiotiary of Derivations ; or, an Introduction to Etymology, hy Robert Sul- livan, LL. D., T. C. D., meets with the approbation of " the distinguished Philologist and Anglo-Saxon scholar," Dr. Bosworth, and causes the Dublin University Magazine to "con- fess we have been startled at the extent of the ignorance of many previous writers on the subject." Dr. Sullivan, with many others, gives divinity (an older word) as from divingf * Outlinoa of Comparative Philology. New York, 1863. See also § 6 a. I ANALYTIC ORTIIOOHAPnY. 78 nnd ho represents h and i- as becoming ", cross, across, loss, long (cf. Gei'. lang,) was, oftea, orthography, coffin, order, God (cf, j^ot,) John, wander (cf. wonder,) hog (cf. hug, big.)f Compare the quantity of or ore hog hawk swan swoon alter older on own short hurt fort horse luiarse cross crease crusty long lung morn mourn burn. * Ellis, Essentials of Phonetics, London, 1848, § 9. •j- " There are great varieties of opinion and practice respecting the vowel in the words cited, both in England and Ame ica, There may be a real difference between awed and long odd, the latter may be closer. . . . Some of the differences you name arose from Mr. Pitman (speaking by dictionary) preferring a close sound and a stopt vowel in cross, loss, gone, often, oflSce, where a long or medial vowel is often or generally beard in London. In long ?re never lengthen o. The word god has the vowel unhistoriealiy lengthened by many," but not opened into gawd, "iicfoie r there is a dispute as to whether a long or short vowel should be placed. Isaac Pitman, who cannot irill i\n r, prefers the ancient short vowel, which to my mind can only be properly used before trilled >•.... I c«.nii,)t help thinking that in your cxpcriiuents on the length ot vowels, you < ,:i. by the process of measuring the f' 10, have been led to take the consonants into account." — A. .1. Elli*, MS. ANALYTIC OKTIIOGRAPIIY. 81 NOTATION OF QUANTITY. 359. The Romania considered the vowels as naturally short. They are naturally long, the consonants being naturally short. Long vowels were the first discriminated and supplied with characters, and in alphabets which do not discriminate between the two, it is safe to infer that the character was made for the long sound.* Theoretically, therefore, there should be no necessity to mark the long vowels or the short consonants. 3G0. Tfie marls of qutiniity should be placed above or after the characters, the former being preferable. In the latter case the mark of accent should surmount that of quantity. The number of diacritics would not disfigure the page, provided each were significant. It is only when they are meaningless that marks offend the eye, as in placing five dots over rijiditi, and yet these dots would not ofiend in a line of staccatoed music. Bcihtlingk has many Jakutish words in a modified Russian orthography, as kypyojax (a deserter,) where * p' is r, and 'x' y^. Sometimes these dots are surmounted by marks of length. Castren has Samojedic spellings like kiijii (birch) iijii (foot;) and there is a lake Abijijis in the State of Maine, and Ujiji in Africa. Compare Fiji, o/^oc (whey,) and Turkish qyjyq (oblique,) a form which shows that strangeness of appearance is as much due to new combinations of familiar letters, as to new characters. 361. If the longs and sJiorts were marked {'") the medials might be left unmarked, in- cluding such about which the writer hesitates — or, these might be marked with a superior dot (a') immediately after the letter. In Hebrew, three degrees of quantity are recog- nised, long, short, and very short; and in Sanscrit a figure 3 is used to denote a very long vowel. Let us use figures to denote length in approximate or nominal eighths of a second, as in f a'^n a^t i'^c, f a^n, a^r m (including the quantity of r,) C! is a full second, or a beat of the metronome at GO. 362. In the following CfieroJcee road c as ^ flat (§ l&l,) e strictly as in thei/, weight (avoiding ebb,) a in art; x as in it; o strictly as a true short in note, obey; and V as English w. Then we have — ce'^hi'^' (cehi') far, ce'hi,^' (ce^hi') very far, na'cvo'^' (nacvo) tiear, na^cvo' (nacvo) very near. 303. The CheroTcee word for wind (used figuratively for smoke) has the three vowels of foot, war, ebb, (u, n, e,) that oiwar being the open vowel o£awe, with a medial quantity, the word is u^nn^li, and it occurs disguised in rhe following word, where medial vowels * " In most languages the short vowels are not so accurately differenced as the long ones; this is the reason why tbo former were not indicated at all in the most ancient languages." — Lepnus, Alphabet, p. 51. 82 ANALYTIC ORTHOGRAPHY. are unmarked, u in vj), pure and nasal, i in feet; the acute accentual marks a short ac- cented votoel, and the grave would be used for a long one. tIcv«''tB"nnln'testi' — used by an old chief at a council, and incorrectly rendered by the interpreter — " the wind blowing from my direction will indicate where I am" — because the ordinary word for fimoJce was replaced by that for wind. On the prairies a column of smoke is a prominent object which may be seen at a great distance. The speaker wished to convey the idea that^— "the distant smoke ascending from my fire will inform you where I am," or, "the smoke at a distance will rise in the air from the place where I am," ti, at a distance; cva^ connects the subject with the speaker, the next t is probably a fulcrum to prevent the concurrence of the two vowels : testi', sJiall be blowing. 364. Quantity can be indicated in two other modes, and although the appearance of a printed page (whether of speech or music,) is secondary to its accuracy in depicting defi- nite phenomena, these modes will oflfend the eye less than the normal Latin mode. There are three variations in the width of type, named extended, medium and condensed, and these would answer extremely well for the three lengths of vowels, except that i, i, are not dis- tinct."' The following are examples : — Extended, A.EIOXJY Medium, A E I U Y Condensed, A E I U Y 4C5. In Italic typography, the termination of a, e, i, u, might be cut oflf at its lowest point, and be supplied with a separate type like that used to add a little flourish to finals in script printing. This addition could be broader or narrower according to the length of the vowel.f 366. Quantity is influenced ^ j consonants. Sonants, which have length themselves, may accompany long vowels, and surds may accompany short ones. In the following pairs, the second is longer than the first; and in German, zeichen tolcen, is shorter than zeig-en, to in-dic-ate. * "As short vowels and consonants are generally more frequent, it is practically most convenient to mark length only. . . . The condensed, medium, and broad-faced type would be very troublesome to distinguish accurately by the eye. I do not think you would approve of it if you had twenty pages of such type (especially in Bmall fonts) to read." ElUx, MS. f An economic provisional typography could be made by using italics (or small Roman letters) and spaces, but excluding capitals. Let the first and second line of «, n, r, a, d,Pf b, g, q, y, h, k, be formed of separate types, some of them meaning nothing except in combination; let a few new marks be made (like ^to form i^for t, to avoid the dot,) and let the required letters be built up from these, as in music printing. Dr. Bapp (Grundrisz, Vol. II., p. 8, &c.,) has formed in this manner a character for ni/out of ij (inverted italic /) the two members being not quite in contact. a e 1 o n y . a e i o u y. a e i a y. ANALYTIC OUTIIOGKAIMIY. 88 fierce fears leaf leave strife mtrive height hide late laid leak league bat bad jolut joined rope robe feet feed hart hard lout loud, 367. Consonants have a recognised qtutntity in Dacota, where s and sh occur short and long. •* When marked thus (a') the sound is prolonged." (Riggs' Dictionary, Washington, 1852.) Thus s'a (sh^a) is red, and s"a (sh"a) to shout. Dr. Lepsius has improperly transferred the mark of shortness to *«' to represent Engliijh «/», and to 'z' for zh. a. The n is long in Italian sedindo pufito (point,) but not in Spanish, which has it in 'Cervantes' =6erEaht£S, where it bears the accent. 368. 27t€ length of continuous consonants may vary with the sonant or surd phase of the succeeding one, as the short secondary vowels are seldom lengthened in English. The following are examples of n, 1, ng, r, m, s, thus lengthened. 8in"ce sin's han'k hong'd biuret blurr'd dam'pt dam'nd pinch impinge pence pens else ells dossed doz'd dint dinned wilt willed start starred etcirt edg'ed. 369. SCHEME OF the vowels. .A. a ^00. \t urn. {a74. * I1H02. i SuabUn.JSOl. t) odd. §405. JL odd. J378. Oit«i.Hn. Fr. §412. O §431. £ Suab. §3a0a. Oi»«. §416 g thfre. §388. Ofwy. O§430. I f* O Itat. §118. U S436. YHW. U?§437. a>? §421. q Swed. u. §440. U foot- §422 g Ru«B. §393 U full y Fr. tt. §436 marint. §399. §430, UWelsbu. ebb. §384. e Oud;rlt'hl. eight- i»^1- a §392. e ? §391a. t "> Longley, U, u; Antrim, o; and II. M. Parkhurst, u (Ploughshare, Boston, U. S., 1853.) It is doubtful whether the modification for open and close, should be made in the upper hook or lower dot; but the former is preferred, because it leaves the character more distinct from e. A, A, {&, a,) in add. 378. With very little affinity to A, this sound usurps its character in some alphabets. It is more nearly allied to ebb, but not enough to have a letter on the same basis, like that of Lepsius. Rapp writes it a; Comstock, A, a, Hart a, Masquerier a, a; and Pitman, Graham, Parkhurst, Kneeland, and Longley A, a. 379. TJie people of BatJi, England, are said to pronounce the name of the town long (= bi0J) and it is strictly long and short in Welsh, as in h^x a hook; hx'-/ little. It seems to be lengthened in the following words, but as the author speaks this dialect,§ the observation must be accepted with caution. « * Universal writing and printing with ordinary letters, Edinb., 1856. -j* Ethnography and Philology of the U. S. Exploring Expedition, 1846. . A valuable work philologically, but not phonetically. He does not think it necessary to indicate French u; ho uses A for the power in mart, mat, (but, probably, did he hear the latter;) E (or /ate, met; I for machine, pin — p. xii. 1846. J The Rev. J. G. Woods (Sketches of Animal Life, 2d Series, London 1855, p. 247,) mentions " The singular mode of pronouncing the word which is used by those who have resided there. Instead of enunciating the word Bath in a clear and open manner, it appears to be correct to elongate it into an effeminate drawl, thus — B-a-a-a-ath, pronouncing the towoI like a in cat." " The sound is common over Wilts and Somerset, and it may extend to Gloucester and South Wales. . . . The long sound is the name of the first letter of the alphabet in Iri^h English. . . . Our ladies often say graas, caalf, haalf, paas, aask. Many orthoepists (Worcester and Bell,) have recog- nised an intermediate vowel." — Ellis, MS. note. § Heard in Philadelphia, and used by Walker, who puts bis a* of/at, in grass, grasp, branch, grant, pass, fast, the proper sound being probably French &, as in pass, &c. . ^ 12 O i\ IMAGE EVALUATION TEST TARGET (MTS) /. 1.0 ^^ tsi u 184 I.I S? Ufi |2.0 L2g|||U 1 1.6 Fhotographic ^Sciences CorparatJon 23 WBT MAIN STRIIT WIBSTIR,N.Y. MSM (716)t72-4S03 \ i^' ■17 <^ 4S> 86 ANALTTIO ORTHOORAPHT. pan pfinic dam band banish dram fan fancy lamb man tan bad can, n. can, verb. glad bran ran bag Ann an, Anna cag Sam sample drag madder, a((/ madder ma'am mammon baa badger gas, gaz gash, as lass lash bread bred dead Dedham bed sped. ham ram lamp pad lad tag, beg wag, keg dragon 380. It occurs in provincial Oerman, as in bi'ric, (with the vowels of b&rrier) for berg (bero) a hill. A native of Gerstungen (= G^rstiiran) in Saxe Weimer, pronounced the first syllable of this name with s. in arroio. Compare thatch, deck; catch, -| ketch; have, •\ hev; scalp, -j scelp; German and English felt fat; krebs crc.6;' {eat /oat, adj.; Gr. rpixto (I run,) track, f 381. It has a long and open German provincial (Suabian) form, being used for long open a (e) as in bjL'r (bicr) for bar, (a bear.) This bears the same relation to add that French S in merae bears to e in memory. 382. This vowel is nasalised and short in the French Jin (end) =fi,; pain (bread) = pA,. But some consider this a nasal of ebb,* either because such a sound is used, (the Polish e,?) or because the French (being without the pure add) refer their nasal m to the nearest pure sound known to them. 383. The cJiaracter i.is a good one, and may be written with Greek a, into which '&/ degenerates in writing. The Anglish se is accessible for the open sound, whilst a small *&' would admit of being trimmed into several distinct shapes for varieties of sound. E, e, in ebb. 384. Most toriters pervert *e' to the use of this sound, an error which arose from regard- ing the vowels of thei/ them as variations in quantity. If the Roman alphabet is to be ad- hered to (( )» the half of ' e ' might be used for it, but a Romanised form of Greek e (like that of Mr. Pitman) is much to be preferred, — and it is shown as. a Greek form in Franz, p. 245, line 10 from below. 385. The secondary vowels ii, ebb, were not allowed to Latin, (§ 93) because there is no evidence that they were Latin sounds; and although ebb occurs in Spanish f (as in el e in use, to be assigned to French e. If the Roman alphabet, is adhered to, the type can be made by cutting away the right half of 'e;' and e can be made in the same manner, retaining the circumflex, — or excluaing it, and mutilating the type less than for e, giving it the appearance of Anglish 6. But (S ) a character formed from (">) Greek om^ga, is preferred for the e sound, and accentualised letters are not to be used to indicate quality. a. We quote here doubtfully, a Suabian open vowel perhaps between there and Mp, and heard in reiten, seide, weiss, fenster, stelle, and in regen (to move,) whilst regen (rain) has e. e, in vein, eight. 391. The English ay in pay, paid, day, weigh, ale, rage, is short in weight, hate, acre, A'mos, A'bram, ape, plague, spade. German weh (wo,)*reh (roe,) je, planet, meer, mehr (more, but mahr tidings has ^,) edel, ehre, jedoch. The Italian "e chiuso" has this quality, as in male, ottobre (with "o chiuso,") but it is nearly always short. Most authors assign this sound to French €, called '6 fermc,' but Dr. Latham assigns this 6 a closer aperture, for he says — <' This is a sound allied to, but different from, the a in fate, and the ee in feet. It is intermediate to the two." a. Dankovszky says the Hungarian "6 est medius sonus inter e et i," but his 'e' is uncertain. Olivier (Les Sons de la Parole, 1844,) makes e identic with I in the position of the mouth. 8 in -ment, -erne. 392. Tliere is an obscure vowel in English, having more aperture than that of ill, and less than that of ail. It is used to separate consonants by such an amount of vocality as may be secured without setting the organs for a particular vowel. It is most readily determined between surds, and it is often confounded and perhaps interchanged with the vowel of up. It occurs in the natural pronunciation of the last syllable of worded, blended, splendid, sordid, livid, ballad, salad, surfeit, buffet, opposes, doses, roses, losses, misses, poorer, horror. Christian, onion, and the suffixes -ment, -ant, -ance, -ent, -ence. 392a. Perhaps this vowel should be indicated by the least mark for the phase of the least distinctness (§ 484,) — a dot beneath the letter of some recognised vowel of about the same aperture. It is so evanescent, that it is often replaced by a consonant vocality without attracting attention, as in saying hors'z, horsz, horszs, or (using a faint smooth r,) hors'z. 3926. Bapp uses 'a' for this sound, and for the closer form allied to «rn, placing it in must, honey, a, an, master, fever. H. M. Parkhurst uses a tailed 'e' in present, convenient, universe, order, and in the suffixes -er, -ent, -ency, -ment; and the vowel of up, in up, money, impression, occur, SQme. Longley uses e in earth, verb, first, person, deserve, sir. # ANALTTIC ORTHOGRAPHT. 89 skirt, thirty, verge, — using the vowel of ebb in very, discovery, another, interest, and that of up in worse; so that he can hardly have the Irish dialect in view. Graham proposes a peculiar *e' for her, bird. 392c. With Rapp, we assign this vowel to German, as in welches, verlieren, verlassen (or even frlasn.) The vowel of up is not admissible in normal German, although it is common enough in dialects, and associated with short o, as in kopf, toll. In our examples, the theoretical vowel is that of eljb. 9, the Russian Li. (g, in Turkish.) 393. This vowel atrikea the ear like the pinched German o, ii, to which series it may belong; but the lips are not pursed, the effect being due to the enlarged cavity of the mouth. The quality is perhaps nearest to the vowel of if, but the jaws are more separated, and the lips are retracted as for I. It is long and short, and is said to be the sound represented in Polish by y. Eichhoff (1836) uses *y' for it; Castren (1854) the same, with an angular circumflex when long; and Ellis uses a small capital t. We propose a character formed from inverted fi, which is sufficiently distinct, whilst it bears some resemblance to the Russian and Polish forms — and 'y' must be restricted to its historic value. 393a. Gastrin mentions this as a Samojedic vowel, and he says that in making it, the end of the tongue is prest against the base of the lower teeth. § 344. He states that in several dialects, 'i,' in certain conditions, has something of this sound. 394. This Slavonic vowel occurs in Jakutish (Bohtlingk,) and is probably the key to an Altai-Tatar infusion, as it is said by Redhouse and Bohtlingk to occur in Turkish. But S^uiiic', who quotes Turkish very freely, in illustration of the elements, does not admit it. We have not been able to compare the two, having heard them with an interval of six years. They are closely allied, and our impression is that the Russian phase is based on ooze, and the Turkish on is. I, in pit. 395. The English vowel of it, pit, pin, &c., frequently formed out of a shortened I, and as 'e' is one of its equivalents, it often takes the secondary power, as in believe, regret, descend, which cannot differ from dispose;"' and we find in old English — biginiian, 1250; began, bithoute, 1280, without the unenglish gh; and Chaucer uses dispise, discent. 396. It is the German vowel of kinn (chin,) hitzig, brilig, will, bild ; and the initial of the Belgian diphthong ieuw (and perhaps, in some cases, the Welsh uw.) It is adopted for the English u in tube, (tiwb) in Comstock's alphabet — a diphthong known to the writer. * See the Phonotypio Journal, 1846, for this vowel in select, secure, reriew, degree, defect, desire, disease, denote, prepare, December, and many more. li 90 ANALTHO OBTHOORAPHT. 397. The form is accessible in a mutilated (U, u, t,) and it will best suit the languages in which I, J, are used correctly, among which it is hoped English will be one. Thus the series J I L ^"ill exhibit normal I turning to J in the closing, and L in the opening direction; and there are good reasons why they should resemble. Their affinity causes an interchange in Chria-tian (crist''jjan, -t'^jan, or -tfan,) with J, and Chris-ti-anity with L. Compare o-li-o, o-lio (oho, oljo,) fil-i-al, fil-ial, foliate, folio; il-i-ad, il-iad, va-ri-ous, va-rious, cordial-ity, idiot, previous, devious. 398. Tliia vowel is commonly confounded with I, but it has a more open jaw aperture, whilst each may be lengthened or shortened. When made long, it suggests long g, but they differ. The following notations have been proposed for e in they, s in Hiem, e theref i in he, and i in his. Rapp, 1836, Lepsius, 1855, Max Miiller, 1855, Ellis, 1856, Pitman, 1856, Pitman, Jan., 1852, Graham, Adair, Kneeland, 1824, Hart, 1851, Comstock, 1846, Masquerier, ]847, Antrim, 1843, Haldeman, 1846. Reynolds, 1846, (§545) _ Hale, 1846; Matushik, 1837, « Poklukar, S^unic', &c., 1, 1, i, in field. 399. Tlie universal I, is long in Italian id (Lat. ego, 1,) and short in f^lIcItarS, with true e. In English it is long in machine, marine, fiend, fee, tea, bee, grieve, eel. It is * Perverting I to eye. In citing the powers of English 'i,' that of marine is omitted, and not because it is i>e, for e, and not o — e is cited for the power in eve. The sixteen tone marks of "Comstook's Perfect Alphabet" . ..." not only represent accent, but inflection and intonation or melody." But as these differ as much as itress and pitch in music, they cannot bo represented by the same mark in a rational system. The inflection of unac- cented syllables is not marked, hence (p. 27) although "refined^." and "region^." close sentences, the final syl- lable of the latter is represented as unaffected, and the first syllable as falling, because this mark means both fall- ing inflection and accent. This notation has been used by its author since 1841. e 8 6 i I 1 i a i i t' . 1 i i i '.# ' e a i • 1 • 8 ee • I • 8 e e i it it ('( « (t u « « a tt i « e e & i <( E j« I a E • 1 a) a y e 8 1 (( « u (( e • 1 K « u ANALTTIO ORTHOGRAPHT. 91 short in Squal, educe, deceit, heat, beet, reef, grief, teeth. Qerman examplea are vieh, ivleder (against,) wider (again,) vile viel (how much,) vielleicht (perhaps.) It is medial in knie (knee.) French examplea are surprise, vive, ile, style, 11, \If, physique, Imiter, liquide, visite, politique, which must not be pronounced like the English physic, &o., with the vowel of pit. The following are perhaps medial, — prodige, cidre, ligue, vite, empire. ^ , (a, ft,) in aisle, Cairo. (§ 372, 4, 5.) 400. Proceeding in the labial direction from A, the first element is French a in ame, patte. The former is commonly received as the vowel of arm, the latter of pat. Du ponceau* in 1817 made the distinction. He says that French a occurs in the English diphthongs i and ou, and that the sound is between ah and awe, being aJi pronounced as full and broadly as possible without falling into aioe. The initial of English i (or e in height,) differs in being pronounced up and at; whilst the orthography *ou' was partly intended to repre- sent the French vowel of could, and partly the Saxon (Plattdeutsch) diphthong, which we have heard, and consider to have the initial of odd. 401. Ellis uses a with a horizontal medial line for it, and Comstock^, and a lowercase form (a) with the base open, and the left branch turned outwards. Pantol^n admits this sound when short, as in a, la, pas, ma, e", il a, and in both syllables of voila, avoir ; but he places the true A in a°, t'l a, car, toi, voix, naif, matelot. This is probably the proper vowel for grass, grant, pass, alas, (Fr. h61as.) See § 379, note. When accessible, we prefer Mr. Pitman's reversed a for French a. ' ^ H* (») in awe. 402. This sound lies between A and 0, and is common in several German dialects, and in Bengalee, where na'^e is nine. The Germans represent it very commonly by a, adopt- ing the Swedish mode, where however the sound seems to be a kind of o. Franz, Epi- graph. Gr. p. 246, line 1, has a Greek character very like 0. 403. This awe is not to be determined by its length, but by its quality. It is long in raw, flaw, law, caw, all, pall, call, thawed, laud, hawk; — medial in loss, cross, tossed, frost, long, song, strong, or, for, Lrd, order, border, war, warrior, corn, adorn, born, warn, horn, morn, storm, form, warm, normal, cork, wan, swan, dawn, fond, bond, pond, exhaust, false, often, soften, gorge, George; — and short in squash, wash, (cf. rush, push,) author, (cf. dath, pith,) watch, water, slaughter, quart, quarter, wart, short, mortar, horse, (cf. curse,) remorse, former, often, north, moth, fault, falter, paltry. 404. For the vowel pair in awe, odd, Ellis uses W),t t>;t Comstock, Pitman and Graham o, o; Bishop Wilkins Greek a; Hale u in a single character; Hart andKneeland with a horizontal medial line; Parkhurst &,t o; Lepsius o, p; Masquerier o, using one sign for * Am. Phil. Trans., 1818, Vol. I, p. 268. f With the appendage on the right M ANALTTIC ORTHOORAPBT. both, (like Wilkins, Hale, and Hart,) a sign made of b, — and d, p, q, would afford allied ones. For common typography, we propose n, 5, with ei (closed) as the writing form of n or its varieties. t> in odd, 405. 17iis d^ers/rom tlie preceding in being formed with less aperture. It is sJiort in not, nod, hod, what, squatter (cf. the open water,) morrow, borrow, sorrow, horror, choice, ponder, throng, prong; medial in on, yon, John, God, rod, gone, aught, thought, bought, caught, naught, fought, sauce, loiter, boy, and perhaps long in coy, oil. Some of these medials may belong to awe, and some of those to this head. 406. TJie accuracy of these examples is not expected to be admitted in detail, because practice between the two vowels is not uniform; yet it is probable that no one puts the vowel oi potter or the quantity of fall, in water, which is neither wawter nor wotter. In the following table, the medial examples have been chosen without regard to the vowel they contain. ^ gaud God nod gnaw'r nor Nor'ich Ki awe or orange rawed rod Rodney fawned fond astonish awed aught odd * thawed thought Thoth laws loss lozenge 407. In the next table, No. 1 is the long, 2 the short, and 3 the medial quantity of atoc; 4 is the medial and 5 the short quantity of odd. 1. 2. 3. 4. 5. pawned author po'nd roM ponder waw water wa^r Go^'d body squaw squash swa'n yawn want wa^n go'ne haw horse ho'rn Jo'hn horror thou'ght squat honest 408. Indications of quantity cannot be dispensed with here, ft, (or whatever character is used) might stand for the vowel of odd, and have a widened form, or a superior dot (') after the letter, for its medials; whilst n (or its representative) might be considered medial, and have a long mark for atoe, since the medials of the close vowel, and the longs of the open one are the rarest. 409. It is a difficult problem to supply avx and odd with suitable characters. They have no more right to be formed on an <0' than on an 'A' basis, and the available forms of *0' should not be drawn upon too largely for English, being required for French o, the two Italian kinds, and perhaps others among described or undetected phases, a. fl, n, or n, n, would form a good pair, and they recall A, 0, but n is perhaps too much like n for blurred print. The preceding, with ft, are not sufficiently alike, because the medial ANALYTIC ORTHOGRAPHY. 93 quantities may be oonfused by the same person in the same word. Other pairs are furnished by the rejected forms of the Phonotypic Journal, as a, n, or n. 410. The chief difficulty is in finding approximate forms which can be readily made with the pen. One of the preceding forms might answer for the open sound of atoe, and 9 (which approaches Gothic 0,) for the close one. The * A' part of the latter could be so much reduced as to make the character approach q, with the mark in contact. A pair like a with the upper or o part large for odd, and A the lower or a part large for atoe, would solve the problem in print, but they would be likely to take an e form in writing. A writing character formed of ei united, would answer for the atoe, and the script a recommended for aisle (the middle of the t portion broken towards the left,) for the closer sound; or, the closed et character (Hale's aw long and short,) might have the close power, and have the t part descending in a short tail, for the open sound, or the e portion with the break thrown to the left. 0, Italian "o aperto." 411. To an unfamiliar ear this vowel is referred at one time to and at another to av)e, and if an Italian speaks English with it, the word hold seems to be bald^ and bald seems hold. It is long in 'poco,' little; pdrto, port; sposo, husband; and short in trdppo, too much; ndttS, night; cdsa, tblng. Mr. Ellis's key words are roco (hoarse,) and rdcco (crozier,) and he refers to this sound, Swedish a° and Danish aa; and with doubt the French vowel of hotte, homme, with which we do tVil agree. Mr. Ellis's character is a good one, a Q form with the tail on the left — which might end unth a dot when the vowel is short. Dr. Comstock uses with a minute vertical tail below, for the short vowel of Fr. bonne (good,) Ital. dotto (learned;) and he places the vowel of own in the French trone (throne,) and Italian dolce (sweet,) the latter being "o chiueo" of the Italian gram- marians. An Italian grammarian compares the "o aperto" to the French o in hotte — " r aperto detto da Franc^si aigu o href, ha il suono dell '0 aperto toscano, come hotte (o-t.)" 0, French o. 412. This eoimd aeema to the writer to be more open than owey and closer than o aperto, and his impression is that the long and short sound have the same quality.* Gouraud * The Author's French pronunciation was acquired from heterogeneons sources, chiefly English and German, and although he las occasionally revised it in casual intercourse with Frenchmen, early hahits are continually ofossing later opinions. His practice is to pronounoe 6 as oim, bfinne as English hone shortened, and m6n with the same nasalised. His ideas of Spanish pronunciation were derived from a South American, whilst his English is partly provincial. On the other hand, his ear is good enough to enable him to tune a piano, (exoept in the low bass notes,) and to distinguuh across a room whether a speaker of German uses the (German to or English v, firovided the voice is fiuniliar. 13 94 ANALYTIC ORTHOaRAPHT. m cites three kinds of French o, referring that of poste, note, code, to o in not; sort, iilort, corde, to Eng. nor; and cote, fnute, beau, to Eng. note. 413. JMniar nibnit^ two, the o of Eng. opera in opfera, homme, loge, remords, offense, comme, notre; and o ototjer, in auteur, oter, impot, zero, faute, rose. Value admits two, as in Eng. no, nor; and Pantolton two, the first (without English equivalent,) in bonne, homme, trop, au, porter, octobre; the second (in Eng. old,) as in tr6ne, eau, beau, matelot. 414. Picot admits two, the first "close, that of o in trop, nearly that of o in nor;^ and " open, that of 6 in tot, nearly that of o in over." Chesnier admits two, as in homme, au- tcl ; and Olivier two, as in mobile, cor; and in beau, dos. In this treatise o (formed from Q) will be used provisionally for the short sound, and for the long one. 416. The New England or Yankee o in whole, coat, is a short sound with a wider aper- ture of jaw than owe, but not (perhaps) of lip. It has been casually heard, but not studied, and we refer it to the French o in bonne. Mr. Graham uses o for it. 0, English, in hone boat. 416. This well Jcnoicn sound is long in moan, loan, owe, go, low, foe, coal, cone, bore, roar, bowl, soul ; and short in over, obey, open, opinion, onyx, onerous, oak, ochre, rogue, oats, opium; and medial in going, showy. It does not occur in Italian. 417. O is long in the German ton, dom, hof, hoch, lob, tod, trog, mohn, lohn, moor, mond; medial in oder, also, vor, von, wo, ob, oheim; and sJiort in wohin, hofnung, ost, ofen, ober, koch, loch, zo-o-log. Hale, Ellis, Hart, Masquerier, &c., use 0; Pitman and Graham & with the tail on the right; Longley and Farkhurst a closed o; and Comstock a, to. Kneeland uses 6 as in know, holy, and o for its short quantity in home, wholly — having probably the New England vowel in view. 0, Italian " o chitiso." 418. For this sound we will use ot provisionally — but preferring the closed form of Mr. Pitman. It occurs in conca (ca»/ca) a shell; onda, wave; botte, a cask, (but botte a blow is open ;) no-io-so vexatious. It is long in aneora (&/ca>ra) yet, and short in ancora (a^c/, and German grons, iwth, oben, but these latter are English, as in gross, note, over. He has probably o chlwio in view. Ca8tr6n (p. 7, § 11,) mentions an open Ostjak n which approaches o, as in ud or od, the hand. We have heard such a sound in the Troquoi word for ten — U^JE'Ll"; and it may occur in the Irish (of Munster) mtJ hu mi/ eye; ciiidj five.* U. in pool; V in puU, (uj, m.u.u.) ,.„. 422. These two vovoeU are distinct in quality, and have the same variations in quantity. They are to each other as awe is to odd, and they require distinct characters. These, in the ordinary alphabet, may be u, u, with marks of quantity. 423. In passing through the series A, 0, U, it will be found that U in pool is labial in its character, and that this labiality is preserved in shortening foo'l to /oo''lish, whilst full, fullish have very little aid from the lips. We may represent /oo?/oo/tVi (often a medial,) by f ul, fulif. 424. If we compare fool with a word like fuel, rule, (avoiding the Belgian diphthong tew,) we detect in it (fyoo'l, rule,) a closer sound, which, when long, is confused with U, as in fool, rule, meaning by the latter neither ryule nor riwl, but rool, with a narrow aperture. This closer u is often preceded by y and r^as in due (=dju",) dew, stew, riiin, rude, where it is rather medial than long. 425. The Latin u is long in woo, two, too, tour, poor, do, who, move, prove, groove, lose, soothe, boom, tomb, moon; and perhaps brew, crew, threw, true, if these are not the closer U lengthened. U is medial in boot, shoot, root, troop, (all of which Walker marks long, like move,) goose, loose, moos, droop, stoop, hoof, proof, tooth. U is short in good, wood, hook, which is not who with k added, as Walker would have it. 426. U is short in foot, full, pull, could, (and if the same aperture is preserved, these do not lengthen into pool, coo'd.) In the following, i/ precedes the short vowel, — acute, dis- pute, refute, refutation. U is medial in rude, truth, fruit, brute, and long in fume (fju'm,) amuse, refuse, bruise. 427. r/te uoioeZo/yboZ occurs ZcMigr in the Italian piu (pju;) Saturno, Mercuric; tu, ihou; in the German pfuhl, uhr, fuhr, buch, and medial in urtheil, nur. That of foot occurs * We have heard an Irish vowel in loch lake, (sometimes 1b;b,) which seemed to lie between up and ope, but the o without labiality. We merely call attention to it here, and to Tschudi's work — Die kechua-Sprache, (Vienna, 185.S,) which contains details of pronunciation, but which we have not now within reach. 90 ANALTTIO OKTHOORAPHT. Jf': I I ;'i t ' short in Italian punto, point; and in German nusz, nutz, muBter, stumm, stunde. The French ou (in pool) is long in foule, and short in courrier. 428. Ibr the wnod pair o/ pool, pttll, Lepsius, Max Miiller, Ellis, Rapp, EichhofT, Bopp, Hale, Hart, &c., use u; Comstock, a character based on n, and u, (perverting u to tip;) Pitman (formerly) and Graham ui, u ; and Bishop Wilkins «. 420. There are two objections to m, m, — it ignores ' u' as made for a full open sound (note, § 369,) and it obscures writing and italics as in mmn (moon) for mun. This use of ui in Russian, for English sh, is inconvenient, as in writing " luumha," a pine cone. a 430. 7%ere ia a middle series of vowels between those of the throat and the lip oide of the scale, and akin to both. Between the o otobey (as being closer than owe, and e of ebb,) we place the close French eu in eux, lieu. It is marked 6 in § 369. H 431. The open sound o/ the preceding is heard in the French oeQ, beQrre, netif; of which some consider de, me, k the short quantity. Both this and the preceding are made with the jaw cavity large, and the lips pursed. Pantol6on writes but one French eu, making no difference for quality or quantity, in which he is not alone. 432. German has an allied or identic sound, long in schworen, schon, konig; and short in mochte, worter, loachen. To the writer, there seems but one German 6, that of French net^, with a tendency to the e side of the scale, a. But Lepsius refers konig to the closer of the two French sounds, and the l^ord Go-the to a position between this and the more open soifnd of French beurre. If this sound exists, there will be three allied characters wanting, H for beurre, (being an open character for an open aperture;) il for Gothe; and a for konig (a close character for a close aperture,) the letter to be unmutilated (h) when the varieties are not discriminated. 433. The first or moat open of these could be written on a u basis, with the break of the left side towards the left. Dr. Lepsius writes the vowel of konig (?,) with the mark of length above, when long; that of Gothe (o, and of beurre (2) with a line of length when long, and if this sound were to occur nasal and accented. Its letter, the doctor's notation, would be 5. 434. Rappi* writes the closer sound ^ as in peiir, leQr, seul, oeil; and the open one o, as in je, de, se, le; and 9 (of English hut, the article a, &c.,) in the French final of noble, &c., when pronounced in poetry. Thus for French 'redoutable' he writes rodutdbh. * PhysioWgie der Spraohe, Vol. III., 1840, p. 108. ANALTTIO ORTnOOEAFUY. 07 Y, y, y, (r, u,) Or.; Dan., Swed., Y. 435. If there ia any dijffbrence between French u and German ii, it is that the latter has a tendency towards I. It ia long in the French buse, vuc, mur (ripe,) and short in niur (a wall,) vu, une, fut. It is long in German iibcl, giite, natiirlich; and short in gliick, kuche, kiissen. We are unable to give an opinion whether the Danish and Swedish y, and Belgian u are exactly identic with the French sound. 486. TJte hiatoric character is Y, often used in Greek typography. Max Miiller uses u; Lepsius the same, with the dots below; and Comstock Y, U 437. Dr. Eapp usee this character (4, 114,) for a vowel between 6 and u, occurring in the German of Elsess (Alsace,) and unknown to us. i U,? 438. This Utter ia uaed by Castren (§ 11,) for a "close u" in Samojedic dialects. • ■ U. 439. WeUh u (y,) long and short, a distinct vowel according to Ellis, and made " with the tongue between the teeth." q- 440. The Swedish u is pinched, and is between 9 and it. Gastrin mentions it as an Ostjac sound. In the ordinary alphabet, q is at hand for it, as in Islandic oi^'a, God. a. Mr. Pitman has a reversed u, a tailed w, and several other forms of these letters, which could be distributed as required, among the vowels of §§ 437-40. See §§ 409-22. 441. The following table (§444) is compiled from Rapp (2, 119, 140, 160, 152, 171, 180;— 3, 161, 223, 265, 308, 312;— 4, 7, 111, 115, 118, 119, 127, 130, 134, 144,) and is in his notation, t^ circumflex indicating length and not quality; e being the vowel of ihey, and a of there. His key word for the fiflh column is * broad,' which does not suit English, the vowel being awe and not o; but as it suits other vowels, it is not altered. 442. Thia table ahowa the absurdity of what is falsely called etymologic orthography, and the impossibility of giving the history of a word in any single spelling. It shows that a phonetic representation of the various phases constitutes the etymology and distinguishes the newer from the older forms, and that in using the present alphabet, LIF, and JIR, are the only proper representatives of leaf and year; and farther, it shows that the vowel of vein has no more right to an a-character than o or i have, for if the original A became E in Gothic, it equally became atoe and o in other dialects. 443. Mchhoff'a table of mutation (Parallele des Langues, p. 91,) shows a similar result, the short Sanscrit A being represented by A, e, i, o, u, in Greek, Latin, Gothic, German, Lithuanian, Russian, and C^eltic. 9$ ANALYTIO ORTHOGRAFHT. idii i.rr Oriffinal, Gothic, ..Islandic, Anglisb, Friesian, Old Saxon,.. New Saxon, Old Suabian, English, ..Danish, ..Swedish, Belgian, Low Saxon, Upper Saxon, .. Old Upper Ger., Old Lower Ger., German, Alsace, Suabian, id. dialect, W. Frankish, .... £. Frankish, Bavarian, Swiss, 444. house, t2 « « « ou n uii oii au,9U (t au 9U au « SCHEME OF I>-TEBMUTATION, good, leaf, 6 u tt it u tt It no n 6 u tt tt uo A n a ua n tt 9U U9 tt a tt PM ea d tt 6 ou i 6 tt tt ou 6 au eao au tt A a tt a d,au year, d i & tt d d A a t d d tt o a A a tt tt d ft d A O 9U d tt broad, a tt ei o a & a i ei 6 i tt tt tt ft ei A e ai ft 01 thief, i ai a A a 09 d,ai m tt eo ia iu i ie I it j& I (( tt ie A e t ia tt A t tt %9 wide. A t tt tt tt it tt it tt ei I tt ai 91 m A t • It ai A • 91 tt : di . ° (( (.< «,9l INPEP8NDENT VOWELS. ^.'- ' 445. In using ih^ Uowpijpe to direct the flame of a lamp upon a small object, as in testing minerals, or in goldsmiths' work, a continuous blast is kept up by filling the cheeks with air, without interrupting the natural breathing through the nostrils; that is, the air may pass into the nostrils, and out of the lips, simultaneously. To effect this, the base of the tongue must close the back of the mouth in the ng position. 446. With the back of the mouVi closed in this manner, or by a deeper closure, the air within the mouth is entirely cut off from that in the lungs; yet it may be compressed and forced out from behind a p, t, cay, position, or dilated by a sucking action behind a AKALTTIO ORTHOGRAPmr. 00 '■,i'A ^54{ d, t, it, c,c?, position, and caused to produce a sound by the opening of the consonant contact, not :vith voice or breath, but by a resonance which some may consider an 'independent' aspiration. 447. 1/ an inverted aspirate sign is prefixed for inspiration or suction, p,e will indicate a syllable drawn inwards. Let 4 indicate independence from the lungs, of the rowel effect or resonance, before the character of which it is placed, when p |.,9 will indicate the sound made faintly by smokers when separating the lips under suction; — t|.,a^, one of the Hot- tentot clacks, the inverted accentual indicating force; — t|.,l^ (or with Av,) a sound made to start horses;— ^,hr a nasal trilled or vibrant inspiration, or snore; — p|.a (the air expelled,) a sound described to us, probably Dacota, for in Riggs' Dictionary, p (also t, k, c'=teA,) with a dot below " has a click sound," whence the word for elm is probably pj-ae (or p.j.e, if the effect is deemed aspirate.) 448. In the Nadaoo (an English name, An-a-dah-has of Schoolcraft,) a Texan language, we have heard such a sound following t, with an effect as loud as spitting, and somewhat resembling it, as in cabat|.o^ (thread,) where the resonance is modified by an o cavity; — n8'st.{.a^ (paper;) — tj-a^a^u^h (tooth,) with final h, it may be considered a dissyllable; — ha'vtj.0; (wind;) — q j.aas (thigh,) a monosyllable, the vowel of medial length. There is ah English click sometimei: heard, indicative of impatience. It is a rapid repetition of t{.,a.* CHAPTER XV. I in the i is, his, • air sed da THE CONSONANTS. No condition is more necessary for the success of a projected system of orthography than that it should be as mueh as possible a necessary deduction from fixed principles, and as little as possible a matter of arbitrary invention. . . . Now, the arbitrary elements of a reformed orthography should be as few as possible; since, as long ds they are arbitrarv, they Xrill vary with the peculiar views of the innovator — and as one innovator will rarely give up his own details for tnose of another, there is no means of injuring uniformity except by laying down preliminary common principles, and admitting Kkna common principle of reasoning upon them.— iVq/^ Latham, Feb. 1849. 449. The nature of the consonanta having been described in Chapter 8, it remains to give them in detail ; and in adopting the Roman alphabet we may associate each sound wit^ the character made for it, or indicate certain known sounds in the same manner that one without a letter would be indicated analogically. Premising that pA, thj cannot be used for simple sounds, because they must have their power in vphoid and pothook, we may in- ■i' Dh is a Bonnd peoaliar to the Oalla luigaage — and extremely difficult to be acquired, the d being followed by a JMfft of hiatos, or guttural approaohiug to the Arabic ain. — Ch. T. Beke, Bsq. Proceed. Pbilol. Soo. 1846, vol. 2, p. 89. I 1 I I 100 analtho obthoorapht. dicate an aspirate of o by '0, and of cay bj Greek x* causing a discrepancy which the use of 'c (with the aspirate mark above) would obviate. 450. Sjogren uses an h formed by continuing the termination down and towards the left, nearly in the shape of o, and this o is added to aspirate any lenis phase. Thus, using the Russian alphabet, F is gay, and the o mark added makes it a sonant aspirate ; — added to n it forms ^ and to the stem of T, ih, but the last is not correct, because t and ih {9) belong to di£ferent contacts. The lower projection of K similarly curved gives %• ^^i" mark forms part of the character, so that there is no economy of types, as there would be in using the Greek asper mark. a. Thia and the allied marks, when convenient to the printer, or when types are specially made, should be placed over the letter. LABIAL COKSONANTS. §451. m V P P {^) b 'b (w ff) m V V. v -v. 10 11 p, surd. B, sonant. M u Of these, p, b, m, have their English power; 'p is preferred to Greek p (§119) except in script; and its sonant form 'b to its proper letter (W, § 127) in the Roman alphabet, or to the Romanic S with which (or with a b with the stem broken towards the left) it may be written. This c is to have the centre open, as distinguished from true /3, which might be used in the modern language instead of (m. Bohtlingk assigns both f and ^ to Ossetian, Grusinian, and Armenian. 'B occurs in Ellenic G (sometimes u,) in Spanish h between vowels, and in German (W,) but some Germans use English v for it; German v and waste (=vest') to destroy ; Sanscrit Vrd to discern ; Latin Video to perceive; English tJoU and e-md-ent, where wit, -vid- are false spellings, wit having the right sound and the wrong letter, whilst the variation of sound in -vid- has not been accompanied by a change in spelling, according to Dr. Latham's sixth rule — '' That changes of speech be followed by corresponding changes of spelling." 456. Latin V has a surd aspirate in English wJi, which is always followed by V way, as in when =v vcn, which is not "ven, as some suppose, nor is it Jiwen, as hden is not then. A character commencing with (') would be suitable for print; and for script, a v with a break towards the left, in the descending stem. Unfortunately, this sound is departing. We heard wig for whig, the first time in July, 1848, and not unfrequently since. When this confusion is established between when wen; where were; which witch; wet whet; whey way; wheel weal; the language will have ceased to be a refined one.* The sound probably belongs to Welsh, provincial Danish, and ancient Greek. 457. "V occurs in several Vesperian languages, and the whistle which Duponceau attri- butes to the Ifnape (Delaware) language, is this tound, as in "vte (heart, nde, my heart,) "vtehim (strawberries,f) with flat^. In the Wyandot (vD'ndbt,) salad* vu (it burrows,) it occurs before a whispered vowel. Compare Penobscot nc^cVde^'s (six;''vtauac (ear;) vtauaGofl (ears.) 458. V^ a nasal English to, occurs in the Penobscot word for seven, — to'mba.'V.B^s. It is No. 1 of the Scheme, § 193. The labial coalescent (§ 451, No. 11,) is nasal in Wyandot) asm ne>£ta>-' Hie pine au'fr6,h,a,v, all unnter tfii > iGVaro^t. is green. LABIODENTALS. F, f; E, E, (v,) English v. 459. Saunda formed by {he contact of the lower lip and upper teeth, of which F is the * "Not neoesaarily. .... In the aontb of England so few people say wheUf whig, that this is the harsh and unrefined, the provincial pronunciation The sound wh is a dialectic pronunciation of khm in Welsh ; and, indeed, it would appear that wh in English came from khw through kw." — EUi» MS. note. f A heart-shaped fruit, but in Wyandot they are called itais, from their bright appearanco among the foliage. li !:!! ii 102 ANALTTIO OSTHOORAFHT. best known. The v of English, French, Spanish, &c., not being a Sanscrit, Greek, Latiq, or normal German sound, it was not supplied with a character in the Latin alphabet. Being a cognate of F, we assign E to it, of which the written form is v with a break to- wards the right, in the middle of the descending stem. The form may be seen at Rome on the tomb of Caius Foblicus Bibulus, in the abbreviation T E*. 460. The Utter y (and 'fr, found in some printing offices,) is not recommended, because it is scarcely distinct enough, and it does not differ sufficiently from Latin V, whilst our pair associates well with p, b, &c. (§ 70,) and if English v has an affinity with Latin V way, as in yaleo, valid, toell; yulgus, nbXxot:, vulgar, ^Ik, it has even more with h and f, (§267) as in prolate, proq^, prove.; s-criie, s-cra/>e, grave, gra/t, graphic; ro6, bereave, bere/t. 461. Should labio-dentdl p, b, m, occur, they can be formed out of these characters with the aid of the marks in §193. Most authors of ethnic or new alphabets use v, and many use w with their English power, the earlier ones having done so thoughtlessly, and the later ones to preserve uniformity — although uniformity from a false basis is not desirable. Mr. Ellis's recommendation of 'w' with its Grerman power, and va for English w, are the least objectionable — but he uses v with its English power. 4G2. There is no certainty in the accounts we have of English v and German to occurring in exotic languages, for when either is mentioned, we have no proof that the observer knew the difference. For example, although the modern Greeks asserted in the most un- qualified manner the identity of their f with English v,^hey were in error, and it has been but a few years since this question was settled. In a similar manner, the Spanish gram- marians are still mystified about their b and v. 463. The sonant labial trill is used in Germany to stop horses, and we have known a child who emphasised the word push by trilling the p when desirous of being pushed to the table after having climbed into his chair. LINGUI-DENTALS. 1, 7, {0, &,) in thin. Q, in is common in English, meaning probably the smooth r in far.) But the -ly in friendly is the li- of live (transposed in ill,) it is the -ley of medley, and if this word is pronounced with the final vowel suppressed, no ear can distinguish the then final I from that of meddle (smedl,) or the I in bulb from that of the transposed bubl, the difference between medley and medl'y being in the diaeresised vowel, (§ 169.) And the question may be asked — If the four English sonants ^medl' do not spell meddle, what do they spell? 483. ^ surd afflate (§ 195, 469',) we have heard in Cherokee (§ 624'°,) and a forcible sonant form (§ 469*°,) in Albanian, as in the word hun (noae), of which it may be a metathesis. INDISTINCTNESS. 484. A dot below a letter should not be used for any important phase of speech, for as the least mark, it should indicate the slightest sound, whether vowel or consonant. The Abb^ Proyart, in his History of Loango, 1776, says of the language — " There are many words which begin with m, n, as in mFouka, nGoio, but these letters are pronounced so slightly, that they who are strangers to the language would pronounce after them Fouka, Goio." " Some Dakotas, in some instances, introduce a slight b sound before m, and also a (2 sound before n." (These are examples of eduction.) " The letter n is hardly heard, and often not at all in the pronunciation of manji, [Fr. j,] in all the words that begin with it." — Baraga, Otchipwe Dictionary, p. 216. 485. We have heard this n in Wyandot, (= vD'ndb't,) where the speaker denied its existence, and would not have written it, had the language been a written (we. It occurs in ndd'cc (ndb^c, four,) and in the name of the town sca-nd^hte'ti' (beyond the pines,) Skenectady in New York — spelt echenectady, the sch being due to the Dutch. The h is the ordinary one, and a slight aspirate closes the word. The accent and the last three vowels are traditionally correct, to remain so until some phonetician fancies that the third syllable should have the vowel of /u<, as malady is supposed to have the vowel of the first syllable repeated in the second. 486. A slight n (not ng) occurs before gay in the Wyandot— unolxat* ihe'f da'nJ,o,'J,ii>'. ' nuta * he-eaie ^ tke-bear. nJ^o/jio' bear; (in Cherokee, jane/.) Here medial quantity is marked with (*). The r is smooth, and >> (§ 568) is the Arabic hamza. ARABIC UNGUALS. 487. Of the Arabic Unguals Lepsius says — " In their formation, the breadth of the tongue ANALTTIO ORTHOORAPHT. either touches or approaches the whole anterior space of the hard palate as far as the teeth, its tip being turned below." We have heard and pronounced these sounds casually, but not with the tip of the tongue turned down. Ellis (Essentials, p. 54,) says — " The tip of the tongue being brought against tJie back of the upper gums tightly, forms t, and loosely, forms d."* Here we think that the only difference between the t and d is the sonancy of the latter. 488. The Arabia letters of this phase are the following, to which we add our marks for lenis and aspirate, sonant and surd. Paulmier's is Algerian, and Volney's characters are cut with peculiar hooks, on the basis here indicated. Smith & Robinson, t s d z Lepsius, i 8 4 z Ellis, T t D J Max Miiller, T Z ? ? Paulmier, t' B' d' 2' Volney, t ~ d 8 8 Richardson, t 9 .1 Z n z S^unic', 't 'i 'd z 489. All the Arabia forma (as J, , t,) have in common a (,) vertical curve on the right, which we propose as being suggestive, and as more appropriate than the dot. Lepsius (Alphabet, p. 46,) adds a (theoretic?) n to the series, and we are inclined to place the Polish barred I here. (§ 478.) This would give the series — t, Sj d, 3, n, 1,. 490. The Ihlieh s', (and z^ although described as a mouill^ a, (z,) is perhaps near the * s,ad.' Vater (Gramm. Poln. 1807,) describes the Polish sound as between (German) aaj and each; and Bishop Pigneaux uses x for a sound between a and ah.-f We have heard such a one in the Waco {z= Veco) of Texas which we will mark provisionally with o, (or if sonant — d) as in iscveto (five,) a word derived from that for hand, as in Lenape and Hebrew. We attribute the same sound to the Chinese of Canton (cvoAof,) where the word for ten is oep'. * " The Boand differs very slightly if at all in the two pronunciations. The tongae is certainly not contracted and hard, when the tip is brought forward, but wide and soft. . . . The Polish Hs to lingual t, aa lia to t." — EUis MS. note. f " Ita littera x etsi sola indicat unam consonantem oujus sonus medium tenet inter litteras « et ch Qallorum et xa, xd, etc. proferunter modo dulciori quam apud Gallos et etiam modo molliori quam «c apud Italoi." — ^Dict. Anamitico-Latinum. Serampore, 1838. ANALTTIO ORTHOGRAPHY. 107 SANSCRIT CEREBRALS. 491. These are thus described by Wilkins (Gramm. 1808,) — "This series of consonants is produced by turning and applying the tip of the tongue far back against the palate; producing a hollow sound, as if proceeding from the head." Lepsius and Ellis add the common Sanscrit r of other authors. Wilkins says that in Bengal the d is "pronounced like a very obtuse r." See §199. We will assume that the Sanscrit r is a common trilled r, and that the Bengali sound is a trilled cerebral (not d but) r, and to be so written. Eichhoff (p. 80) excludes the I as fictitious; Wilkins makes it the Welsh IL Eichhoff uses his dotted l for a Birman sound, which others consider Polish I. 492. Lepsius, t 4 n S z r / t 4 Ellis, to dc nc fc JC re Ic ten dcH Miiller, t d n sh r I th dh Bopp, t 4 n •« r t. • Eichhoff, T p N s R L TH DH 493. Most of the Sanscrit forms have a horizontal curve below (J) by which we propose to denote them, placing it below or after the base letter, as in — t„ d„ n^ 8^ t„h d^h 494. Another mode is to use the small italic capitals, r.AJv,/^ &c., — and Bengali s, which would represent the point of contact as removed towards the throat. §471. In Ellis's notation, 'o' is a diacrit, rizon. 498. Lepsivs ry'ecta (he Emianic Cay on account of his third rule, which virtually rejects pronounced and etymologic Latin, and tends to render the barbarisms in it permanent. Yet, if he rejects Cay on account of its many powers, it had at least its correct power in several important living languages, whilst z has its correct power in no modern language, and its perversions are quite numerous, a. Its powers are, 1. Ancient Greek, as English zd; 2. Italian dz (and ta;) 3. German is; 4. English in azure; 6. as « in Hungarian and Danish; 6. French; 7. Spanish; 8. Middle high German; 9. Scotch, as in Dalzel or Dalyel, where it is derived from 6 through the Anglish s. 499. Bopp uaea s (§484,) for French, Polish, and English z, for which 's' and our surd mark might be used, but the space above may be wanted for marks of quantity. Most authors use z for it. 500. The Qreek and Latin R was trilled, as described by the ancients, and this accords with European practice. The letter 'r' therefore means this sound — however convenient the addition of a sign of trill (") might be found. Rule 5, § 63. We have heard trilled r in Albanian, Armenian (in part,) Arabic, Chaldee, Ellenic, Ulyrian, Wallachian, Hunga- rian, Russian, Catalonian, Turkish (in part,) Islandic, Hindustanee, Bengalee, Tamil, and other languages, in the pronunciation of natives. 501. The trilled r is assigned to English as an initial, although many people with an English vernacular cannot pronounce it. Dr. James Rush would have the trill reduced in English to a single tap of the tongue against the palate. This we indicate by tp with a dot above. 501a. The Spaniah (South American) r in pcrro (dog) as distinguished from the common trilled r of pero (but,) seems to be untrilled, and to have the tongue pressed flatly, somewhat as in English z, and doubled, as in more-reat. It may have arisen from an attempt to yotacise r. We mark it x (or if trilled, r,) with a line below^ in case it is distinct from the next. § 502. ANALYTIC ORTHOGRAPHY. 109 602. Armenian and Turkish have a smooth (i. e., an untrillcd) tactual r, much like the Spanish rr, if not^the same, and with that, requiring farther investigation and comparison. Our impression is, that this oriental r may belong to the series of the Arabic Unguals, in which case its letter would be r„ as in Turkish (with Latin letters) jlr.mi (twenty-one,) whilst Turkish t^rmeo (to give) has the ordinary or trilled r. 503. English smooth r, in curry, acre (a-cr,) begr, grey, curt, is formed by much less contact than the European and Asiatic r requires. It is the true liquid of the s contact, and allied to the vowel (e) in tip, a, character to be formed provisionally from italic x. Ellis writes it i. 504. The Sanscrit vofwel r, long and short, — written by Lepsius with /• and a circle below, and 'r by Ellis, should probably be figured on this basis. 505. A more open, smooth r, is found in cur, fur, far, more, which may bo marked in Ellis's mode, with an r having the stem continued down to the length of '1;' — or with (r) Anglish 8. We use the latter in our examples. 506. Mr, Ellis regards *fur' as /and this open r, without a vowel between, and Kneeland had a character for nr. We regard /ur as having the open vowel v (with which the consonant is allied,) short, the quantity being confined to the consonant (/ur=fu''r',) and the tongue moving from the vowel to the consonant position. 507. The same open consonant occurs in arm, worm, turn, ore; and although, for a particular purpose we have cited arm as long (§93,) it contains a short vowel (aVm) and long or medial consonant. 508. Jf voe write 'rn for urn, and fr, or fR, for /ur, we certainly cannot represent far, four, in the same manner. Moreover, we may dissyllabise pr-ay on a trilled or a close r, and monosyllabise it p'ray with the most open. 509. At one time the discussion of the English letters led to a curious result. When the difference between the open r of tarry (from tar) and the close one of the verb tarry, was ascertained, an identity of vowel and of consonant was represented, — a greater error than to spell more and moor, fairy and ferry alike, or prea-d for preat. 510. The Welsh surd aspirate rh ("r) may be the smooth element. We do not remember its character upon this point. The French -tre, -pre, is trilled, and perhaps rather whispered than aspirate. 511. The Polish rz, Bohemian f, is a trilled (and as we believe) aspirate r (sonant and surd) made simultaneously with zh (j) or sh (r.) See Ellis, Essentials, p. 50. One hypo- thesis has been given in § 200, another presents itself in the probability that it has arisen from an attempt to yotacise r, yotacism being common in the Slavonic languages. § 519. 15 , no ANALYTIC ORTIIOGRAPUr. 612. Lepaim rvprenmla Polish rz by r', Ellis, by ,e 3 ,) belongs, owing to the influence of the surd p. If there is no aspiration, its mark must be supprest. 513. Tliere is no guttural r, all the foregoing being made strictly in the anterior part of the mouth. But in dialectic German and French trilled r is replaced (by otosis) with a vibrant guttural, which is as far from r as German ch is from «. PALATALS. f , in potion. y, in nat^ur. tf, in etch. J, in brazter. 3 , in soldter. dj, in edge. 614. Every coimderation, philosophic and practic, requires that English ah (r) and French j (j) should have distinct characters, and that these sounds should riot be considered as having an aspirate or other aflinity with «, z. § 68. Our characters are as distinct as ' b, d,' and they have not been chosen that they may recall Latin S, J. Moreover, were it necessary to use a pointed ' s ' either for r or &, we would prefer it for the latter, as less likely to outrage affinities. The character * f was proposed by Volney in 1818, 'j' by Ellis in 1866, and both were used by us in ISdO.*" 615. The foiloxoing are some 0/ the forma which have been proposed for r, j. Bopp « I < Lepsius a w z S'ufiic' 8' z' Riggs * 8 z' Max MuUer a z Rapp ah fh Eichho£f 1836 9 J Hale 1846 f • 3 Comstock 1846 c J Ellis r J Longley « 3 Parkhurst « (( Pitman 1844 f J « 1856 • J Graham « « Matushik t t Masquerier k J Pickering sh zh 616. Among ihe toorst of theae and other forms, are those which were intended to recall the erroneous English notation, or to convey the impression, that f has some aspirate relation to s, § 68. Still worse is the desecration of Latin Cay. 617. The Sanscrit 5|t, according to Wilkins, "is produced by applying the tip of the tongue to the fore part of the palate, and passing the voice as in pronouncing our «." * Prooeedings of the American Fhilosophioal Sooiety, Vol. 4, p. 268. t An S ftcing the left, and a Z facing the right. ANALYTIC OUTnOGRAl'lir. Ill Eichhoff, who took bin pronunciation from the Tpouth of Ramtnohun Roy, makes it French ch, Eng. sh, and Max Miiller docs the same. Dopp makes it diiTercnt, he marks it >, the English fih sound being assigned to the fourth cerebral of § 401. I^psius (Alphal)ct p. 42,) and Ellis, thinks it x- % the description of Wilkins, it may be a sound between s and f, or a flat s, or one formed a little posterior to the ordinary point of contact. Wo have been accustomed to uso f for it, but as this is unsatisfactory for a doubtful sound. EichhofF's character 9 may bo used, x n^ust have occurred in the antecedents of Sanscrit, although it seems foreign to the genius of Sanscrit itself. 618. The liquidH of tfie palatul contact are a kind of J (ym) made at the palatal point, and as Eng. to, v, and r, z, are permutable, so 3 fulls into j, and its surd anpirate into f. 519. Hence the word soldier (sssoldjr, or sold^or,) is apt to fall into soldj-r; and nature (snet-"} Xf net"} } t, or net} x,) into netfi;, or netfai;. 520. Jotacism (i/otaoum) is the forming of J or some other allied sound HunnUnneouxly with a continuous consonant made with the outer part of the tongue, as «, r, /. Let the back part of the tongue be held in contact near the J [yea) point, and the apex upon /, in forming li in million, when the e£fect will be mil}} on, as the French fille is fil} , or fil/} , or (eliding /,) fi"} . 521. f, J, must be yotaciaed with the true J (yea,) because an attempt to do so on the } basis, would produce a sound between mute and liquid, like the fusion of English r, z. 522. There is probably no true yotacism (§ 519) after labials (p, b, m,) and gutturals (cay, gay,) or abrupts, for how are pj, gj, (or p} , g} ,) tj, t} , to be sounded except in suc- cession? The simultaneous effect, like that cited in million (properly miMjun,) is there- fore cut off in the Russian pjet"} (five, whilst in s>}Em (seven) it is apparent, with a surd liquid preceding the sonant which meets the vowel. 523. T?ie palatal liquid seems to be present in the French gn, Spaniuh ii, as in cigogne (ssioon} ,) nino (=nin} o.) 524. CastrevbS notation of real or supposed yotacism requires numerous types, us be passes a curved line (") through the stem of the affected elements, as 1, r, n, (the right side,) t, d, 8, z, c, (ts.) A small palatal or guttural 'J' (as the case may require,) would be more economical, and might be understood to be surd after surds. Mr. Ellis uses j, a character made by removing the dots of j. ' c, m car. * G, in g^et. " /■ (gJ sing. 525. GUTTURALS. * c, bucA. ^ q, \ch. * x, Swiss vibrant. * G, betro.9en. ^ £, (y ) konir/e * y, Ellenic id. 10 J, you. " J, hue. " J., nasal, §647. 112 ANALYTIC ORTHOGRAPHY. Hi 526. We adopt c, o, instead of k, g, proposing that a small (lower case) letter be made for G on the model of c. Mr. A. D. Sproat says — " The forms of the Roman and Italian letters (g excepted) are beautiful."* In fact, g is an ugly perversion in which the intended affinity between c, g, is destroyed. The dot of g is that of c, the circle is its body, and its tail is the distinguishing carvilium or mark of sonancy. The French have a lower case form modelled on (q;,) the written form, which associates it with its congener, normal J j. g itself should be curtailed (cr) so as not to project below the line. 527. We adopt Gay Gay as cognates in power and form, in the chief languages written in the Roman, as distinguished from the Greek alphabet. K is a foreign letter in Italian, Spanish, and French, where q is acknowledged — itself preferable to h, but q is required in its oriental sense. 528. Tlie use of h would tend to force it upon Latin, and although this has been done by Rask and Rapp, it is a dangerous course — but a course which shows the necessity of giving Cay its proper power in all cases. 529. It is true that *C' is an S in Greek, but deceptively, the Greek and Russian C being a form of i", S, whilst Cay is a form of F which in some cases had a semicircular form in Greek. See Franz, p. 25. Similarly, x in beattx is a form, not of Latin x, but of s, as French, Spanish, and English y is a form, not of Greek and Danish y, but of Latin ij, as is shown in the older typography of Latin, where they are often printed from a single type. 530. Gay cannot be igmtred, (§43-5,) because it will be always present in etymologic Latin, in Anglisb, Welsh, and Irish, Spanish, French, Italian, and old English. Probably every school, and the great majority of reading families, will have an etymologic dictionary, and scholars acquainted with the Latin alphabet, may be inclined to represent the pair, cay, gay, with the proper letters in exotic languages, where the natives cannot be prejudiced. 531. If hah is used, its stem should be shortened, as in Kneeland's character, and as one of its inscriptive forms is ;, this might even be adopted, such a double character being less objectionable than an entire k. The Punic cay is C with a vertical line through it, as in 0. The Albanian k is a semicircle {^). 532. The greatest concession that could be allowed to hah, would be a character made of h with the vertical line removed, leaving c with a break towards the left, which would be useful in distinguishing script c from e. But even this would be dangerous, because the * An Endeavour towards a Universal Alphabet, p. 10. It appears from a notice in the Am. Jour. Sci., 1840, Vol. XXXIX. p. 197, that this author addressed a letter to Prof. Silliman on the subject of bis alphabet, dated Feb. 22, 1834, (§ 80, 79.) It is noticed in the Phon. Jour., Feb. 20, 1855. ANALTTIO ORTHOORAPHT. 113 rejected c would then be at the mercy of every one who might want a new character; so that whilst c, e would be too much alike with c as cat/, the case would be different with a perverted power. Moreover, u and n are more alike than c and c, v and r, and various examples we have taken from native sources, are worthless on this account. 533. The QotJtic hemi-greek aljpJiabet has 7e, with which the Germans barbarised their alphabet, especially in the use of the hibrid ck, but ck and ch are concessions to the true cay, and the use of k has not caused kk and kh to replace ck and cJi. An Englishman will spell 'sceptic' rather than 'skeptic;' an Italian prefers 'chi,' and the Spaniard 'qui' to ki, whilst any one desirous of uniformity, who acknowledges 'ca' to be correct, will not object to ce, ci, if he is provided with the means of spelling tre, tri. 534. The use of kah for cay is to be deprecated in a highly latinised language like English. It is equivalent to granting that when words change, the spelling should not change (§455,) but that a new character must be placed in the unchanged words; letting t in the French 'nation' have the power of a, using a for z, as in rose, and going to Greek for a new r with which to spell words like 'narif.' This mode is always wrong — that which does not interfere with forms which retain their historic value, is believed to be always right, no matter how long the time during which it has been neglected or broken.*" 535. In old high'Qerman and middle high German, Cay and Kah were both used, and cay quite extensively, so that if the Germans were to re-adopt it, it would be a restoration rather than a novelty. 536. Etymologic relations. (§135.) c^sar, Ohg. caesar, keisar, cheisar. caseus, Ohg. kas, case, chase, Ang. cese, Eng. cheese, croc-io, to croak, Ang. circ, Eng. kirk, church. * Mr. Ellis puts a note here to the effect that English k, y, z, will prevail. " As I deny the effect of k for c in altering the relations — merely altering them to the eye, not the ear — the argument does not touch me. To mark the connection between English and Latin by the eye onli/, I consider false." This remark is just, and we admit that like letters should represent similar sounds. "If we know c^k," [and we know and have it as well for English as for Latin] "this is enough, wo may then change the Latin; writing (in palaeotypo) kaizar? kaisar, keesar, caeiar; kaaseus, caseui; krookioo, crocio; . . . Ang. keeze, cese, Eng. tshiiz, cheese, &c., where the real comparison is between the phonetic words, and the original spelling (and meaning) is merely added as a means of identification. We must thus alter Sanscrit, Greek, Hebrew, — why not Latin too? I doubt whether we shall ever get people to agree on a pronunciation of Latin, even by introducing such an alphabet as yours. Let us introduce the best we cai) get people to accept, even though we pay the price of letting Latin be like the ust, a language to be transliterated." To this we reply, that in transliterating Sanscrit, we do not falsify a single Sanscrit letter, whilst in thus meddling with Latin, the falsifications cannot even have the collateral merits of uniformity and stability — even if we do not take truth to the original into account. No one can yet predict the degree of perfection which people will or may be prepared to accept, but the fact is constantly before us, that the nation which has advanced furthest in civilization, has adopted a metric system in no way connected with the systems already in use, 8y!w, :i| ill Ang. cing, Irish, ceann, Welsh, cun, £ng. king, cancer, Ohg. cancur, Eng. canker. CARCER, Ger. kerker, Ohg. carcare, karkari, (prison.) cithara, Ohg. citara, Eng. guitar. GENU, Ohg. cneo, chniu, kniu, Ang. cneov, Eng. knee, coqvus, Ohg. koch, coch, choc; Ang. coc, Eng. cook,=cuc, Ic u Ic, kuk. 537. We ratlier prefer G. for the nasal of sing, because it tells what the phase is, and we are averse to associating the sound with an n-character, which would be paralleled by representing d with a g character. It is the English and German ng in sing,* finger,= Eng. iiG.GT, Ger. fiG.r, (§20-22.) It is common in Greek and Latin, rare in Italian, and unknown to Russian and normal French; but we have heard it in the Frovensal dialect at Marseilles, in Savoy, and in the Bearnais of Pau. It occurs in Spanish, Catalonian, Armenian, and in the Tonga group. 538. Pitman, Ellis, &c., use \s, g, and Comstock, i)> whicli have the advantage of being like 'j,' the representative of the allied liquid. Others use the same small letter with the end turned towards the right, which is less convenient in print. We recommend Mr. Pitman! 8 form, because it may be introduced into Latin (like J for I,) which we dare not do with G,. Bohtlingk and Sjogren use H. (a nasalised Russian N,) which is wrong in theory. In the less modern alphabets, ng is used. Eichhoff uses n; Marsden, ng circum- fiexed; Lepsius, n with a dot above, and Max Miiller, a capital N. We use r temporarily, because it is accessible in Greek typography. 'c 'q *-^, «u;-rf. §525. ^ ® G ^"2 '■ y, sonant. 539. There are three surd, and two sonant Germanic aspirates; the first (q) in ich is the smoothest and most anterior (§471,) forced forward by the closure required for close vowels; — the second (c) in ach, buch, the Greek Xt and according to some (but doubtfully) the Spanish j (jota;) — the third, the rough Swiss vibrant aspirate, as in ich=ldx — a sound we have heard in Lenape. We have also heard the Swiss sound untrilled, as in ;t;antdn, a canton. In the Swiss dialect, it does not vary before i, e, and from the slight vowel interposed in the word ich, the position of I seems too narrow for it, although they say giidt for gut good, gediant, &c., where id accounts for the German and English ie spelling. 540. We cannot determine the relation of this x to the oriental aspirate of Q, (or surd of ghain,) having heard them at distant periods; but they are probably distinct. 541. We ttse X provisionally (§386) for any Germanic or other allied ch sound which * We have knowo a distinguished scholar to contend that 'sing' ought to be pronounced sing-g, because it has a final g in the spelling. This shows that the advantages ot° fdnbttpi (compare oAnepi,) are not confined to the unlettered. See §27, note. ANALYTIC ORTH06RAFHT. 115 we by -a has not been particularly described or discriminated, although the proper power of x ^^ that in the German buck. Pantol^n, who speaks Ellenic, ascribes to Greek x ^th the sounds of buck and ich. The smooth c is heard (before e, i,) in the Spanish general, registro, (=ceneral, recistro.) 542. 'G is recognised in some dialects of German. We regard it as the sonant of 'c. By G* (f) we indicate the EUenic (not Hellenic,) or modern Greek soft vibrant r- None of these is the harsh oriental gh as we have heard it in Arabic and Armenian. This belongs to the deeper contact of Q. But most authorities consider the Germanic, Ellenic, and Oriental " gh " identic. Lepsius uses r for (g ) the incorrectly named " guttural r," and Faulmier uses r for Arabic ghain. See § 513. 543. We adopt Mr. Ellis's two key words betro^ren ('g) and koni(/e (V, 2O for the spi- rants of g. He adopts an additional character (a tailed i) for Spanish j. — Universal Writing, &c. p. 6'. 544. The follofmng notations may be compared. Properly as the q character (meaning the form of Pitman and Ellis, is formed on c, the V should be formed on g with the same appendage. huch, ich, tage, tdglich. c Q G 2(» Ellis X 9 s g Lepsius Z' / i i MiiUer 'h 'y 'h y Rapp X X / ■i i. 545. We follow Eapp, Bohtlingk, i^ogren, Gastren, Matushik, S^unic', and Poklukar, in adopting the character 'J' for the initial of the English year, Belg. jaar, and German jahr; Latin jugd", Ital. jugo (and giogo,) Spanish yugo, Gothic juk, Ger. joch, Angl. geoc, joe, Eng. yoke, = JOC. J is used with its historic value in the Etujlish alphabets of Hart, 1851; R. R., Phonotypic J. 1846, p. 160; and the Rev. W. M. Reynolds, (Pre- sident of Wittenberg College, Ohio,) 1846.* 546. The surd aspirate "J occurs in the English hue, hew; — ^yh of its discoverer Ellis. 546a. Nasal J^ occurs in Jakutish. Bohtlingk's letter is j with a horizontal line through the top. We have heard it in Cherokee. * He says (Lit. Record of Penna. College, Vol. 1, p. 48,) — " The letters 0, Q, x and T are rejected, the first three as superfluoas, and the T on ticeount ofitt unsettled power in English as well as in other languages." Here an author, by following Lepsins's Bole III, p. 82, rejects y and adopts/, whilst Lepsius does the reverse — thus demonstrating that the < rale' which was unphiloeophic is also impracticable, and therefore no Rvle. See § 167, end of the note. 116 ANALYTIC ORTHOORAFUY. fh ,'^> .T<» FAUCALS. ^ - Qq, Q, 2r(2.) 547. ^^ /at^ca7« toe 77i«a7i certain consonants of which the type is the oriental Q qa/f the 2l8t letter of the Arabic alphabet, and (p) the 19th of the Hebrew. Qaf is a kind of posterior cai/, made behind the palatal veil, and therefore incapable of nasality. Guided by description, we pronounced them correctly (except ain) before hearing them in nature. 548. The surd aspirate of Q IB the aeyenih Arahic letter ^7uz. Richardson says ''it is generated by a gentle vibration in the throat." This removes it from Greek (x) and Ger- man c7i. Its letter would be 'Q, but as this implies a smooth form, it is better to indicate the vibration by (J, or still better QT. 549. The sonant o/Q is the 19th letter ^ghain' of the Arabic alphabet, and the third 'gimd' (= Gijmfl,) of the Hebrew. We indicate it by 2 (not 2 with a straight base,) from its similarity to Q. The mark of vibration would be an advantage, and should a lenis form occur, its sign would be '2.^" Richardson (Arabic Diet.) says correctly, that it is " articulated in the throat with a vibration producing a sound like that given to r by the Northumbrians, or the noise made in gargling. ... It seems to bear the same relation to kh as b to p." It is not the German g in regen, § 542. 550. We cite Armenian examples of Q*, ST, — which, though identic with the Arabic equi- valents, they seem to have a dialectic variation, as we have heard Armenian ghain re- placed with Ellenic ghamma, § 542. The letters are purposely varied for comparison, here and in the next paragraph. danda^k, a cymbal. qb"'q B'ntfel, a neigh. qelc, tha mind. (J^a'tf , a crucifix. 551. As independent p{.^, i\.0, c{.;(, can be formed without air from the lungs (§ 446,) so in the Chinook of Oregon, q.(.'q is similarly treated, according to the pronunciation of Dr. J. K. Townsend, which we acquired. But Mr. Hale makes the sound t;f1, in which he is probably wrong, because all agree that the Chinook sound is a very difficult one to pro- nounce, whilst Hale's is an easy combination. Moreover, the effect upon the ear is not unlike that in the word for thigh given in § 448, which we learnt in nature. In the fol- * Mr. Hale notices a sound which may be a variety of this, in the Patagonian language. It is formed in the innermost part of the mouth, which opens a little, the tip of the tongue being applied to the lower gum. The tonant of Q seems to occur in Berber, and dialeotically in Arabic— judging from the paper of F. W. Newman, Esq., in the Philol. Soo. Proceedings, 1848, Vol. 1, p. 137. ANALTTIO ORTHOGRAPHY. 117 lowing examples (in which allowance must be made for two 'personal equations') the vowels are normal, and the diphthong as in out. hevq^^'qeq^qe, grandmother. Q|.'qavQ|.qavBQJ.'q, ycZfow.* LARTNGALS. r I. 8 h T letiia. jfi G, 7, 8, aspirate. 8 552. The laryngal contact pertains to the larynx, and we adopt the term, in preference to glottal, because this is commonly made to include the faucals or pharyngals. But the faucals of Lepsius are our laryngals. 553. Many deny that A is a consonant, because * it is not made by contact or interrup- tion.' But when the breath is impelled through an aperture which obstructs it, there is interruption, and if we vary the impulse we can make English oo and lo with the same aperture. 554. Tlie walla of the glottis can close, thus forming a consonant contact; and as the glottal fissure (§ 148) is the narrowest part of the breathing tube, it is the seat of the deep- est point of interruption, and of h. 555. The spiritua lenia (') has been described in § 115a, but authors are not agreed about it. Some make it the Hebrew aleph, and Arabic hamza, about which opinions differ also. Max Miiller says (Languages of the Seat of War, p. xxvii.-viii.,) — " We can more easily perceive what is meant by the spiritus lenis inherent in every unaspirated initial vowel, if we pronounce blacking and black ink. ..in black ink, the i is ushered in by the spiritus lenis. This spiritus lenis is the Hamzeh of the Arabs. . . . Its sound is produced by the opening of the larynx, but there is no previous effort of closing the larynx which alone could be said to give it an explosive character." a. This describes the spiritus lenis as understood by moderns, hut the hamza is nothing like it. 556. Ellis gives the spiritus lenis as occurring between ao in a,orta, being "the slight effort made when any vowel sound is uttered," whilst in the hamza — "the effort of * This Boand is probably identic with that described by the late Rev. Emmanuel Naxera, a Mexican ecclesiastic, as found in the Othomi language of that country. " K simplex vel duplex est. Duplex Hispauo-Mexicani gram- matioi cc castanuela$ Tooant, quia ejus sonus similis est stridorl & simi& facto, nuoes frangenti. Litteris cc, qq, vel. qh ooulis pingitur. T, aliquandi etikm sonitu effertur." (§ 448.) — Am. Phil. Trans. Vol. Y, new series, p. %254, 1837. This Tshinook fauoal may be the Hottentot guttural clack, described by Thunberg as " the most difficult uf all, and performed quite low down in the throat, with the very root of the tongue." v^ » 118 ANALTTIO ORTHOGRAPHY. enunciation and separation of the following vowel from preceding sounds is more distinct. An exaggeration of this produces a kind of bleat, which is the true Arabic gain." — Univ. Writing, p. 5" below. 507. Lepsius aaya — "By closing the throat and then opening it to pronounce a vowel, we produce the slight explosive sound which in the Eastern languages is marked separately, but not in the European, except in the Greek. We perceive it distinctly between two vowels which, following each other, are pronounced separately, as in go 'over." Here the hamza is correctly described, and the English e£fect improperly referred to it. 558. We do not think it necessary to represent the initial eflfect of at ('a<, or better — ''at, or caat, with whispered a,) as distinguished from hat, unless the glottis is closed — and we do not mean the epiglottis, which cannot act in speech. 559. We deem the effect in black "ink, a oita, go "over, Fr. le "heros, as a separation akin to diaeresis (§168,) or an accentual difference without separation, as in zoophyte, neophyte, zoology, neology. 560. Hiatus (*) is a break or pause commonly caused by dropping an intermediate element and not closing the remainder, the word and each of its constituents retaining their proper length, as in saying o *orae (not a orae,) for a horse, or a "orse. See Ellis, Essentials, p. 41. It would occur in zo-ophyte, if the least pause were made, and avoiding hamza. 561. Such a hiatus has been attributed to the name Hawai'i, as compared with the earlier New Zealand word hawaiA;i. But whilst one traveller called our attention to this ' hiatus,' two others pronounced this word (as they believed) in the native mode, with a genuine hamza. (§568.) Wm. Ellis (Polynesian Researches, vol. 4, cb. 2,) does not mention anything of the kind in giving the pronunciation of Hausan, but in his appendix on the language, he speaks of "a peculiar break" distinguishing o'u (/) from ou (you,) this being, as it seems, a diphthong beginning with true o. 562. We do not adopt tlie two dots of §§ 227, 306, to indicate hiatus, because they are used for an etymologic "and not for a phonetic purpose; and because we prefer a sign more like that used for the (hamza, § 568,) closure of the glottis — although hiatus does not belong to any contact. 563. The sign () represents the slight phase, whether aspirate, independent (§446,) or even vocal, at the close of abrupt syllables, as in top*, tub', or tubh. 564. The sign (') indicates the opposite phase to ', where the breath is not allowed to escape after tap" (the lips remaining closed,) as in Chinese. This inconvenient notation is preferred to (') because this is used to contradict aspiration like that of s, i>, not the AMALTTIC ORTHOORAPHT. 119 Ct. false 'aspiration' like that of p in 'haphazard/ vhioh is no more aspirate than the p in up stain. 565. H, h, is the common English and German h, in the syllables held, hat, hast, hose. It is unknown to French, Italian, Spanish, Russian, and EUenic. -. ; , ; 566. j/ji ^, w for the eighth Hebrew letter hheth (='/i€J<>,) and the sixth (hha) of the Arabic alphabet. We adopt the Greek •^ inverted (but of a better form than these,) which is nearly the Ethiopic and Amharic letter of probably the same Arabic hJta. 567. ^ (h) is commonly called an emphatic A, and often represented by hh. As heard by us, it is an enforced, somewhat close h, with a tendency to scrape along the throat, and consequently, it is not a pulmonic aspirate. But S^ufiic' probably describes a differ- ent element, for he compares it to the open coughing of an ox, which differs from h as warm or pulmonic breath differs from it.*" The glottis would be opened for such an ele- ment beyond the normal position, so as to render more lung exertion necessary, to give it body. The pulmonic breath is often used in the continuous portion of a cough. Should these two varieties be found to exist in speech, they will run (from the closer to the more open direction) 'ft, h, h. 567a. The Florentine aspirate in casa, misericordia, chi, we have casually heard, and believe it to be -ft, and also the Spanish j, x, before a, o, «, as in jabon (soap, = fa bon,) and the geographical name San Juan (=: ean-fi van,) in English — sin vv5n, which a Chinese would accept for 'crooked mountain.' , 568. Bamza is a closure of the glottis, which we indicate hy > . It occurs as a cutting off of the breath at the beginning of a cough, ( > h, or >4>) during laughter, and when the breath is held in lifting a heavy weight, or in leaping. It is found in Wyandot (§ 486) and Chippeway. 569. Rapp considers thespintus lenis a closure, and writes it (1, 84) with y. He cites a South German negation (1,166; 2,267,) with which we are vernacularly familiar, as "haipia," doubling the sign to shorten the (nasal) vowel.f We would write it (with h nasal "also, h^B, '~ > b,", both vowels being short, the first accented.J It has several vowel- less forms which he writes hmmyrn, hnnypiy &c., (*mm'>m,''nn'>'n, 'g,g/>g, or ' f!''>f.) 570. The Arabic and Hebrew ain Volney regarded as a vowel modification, using a marked a (e, o) for it, the sound being forii*ed with a varying vowel aperture. The vowel is heard with a simultaneous faucal scrape, which may be regarded as a sufficient * See Ellis, Essentials, p. 40, § 5, 6 f The oorresponding i/et, which Rapp writes " hmhm"' is rather m* lum', the second syllable accented. In Eng- lish a single long m is sometimes used fur i/e», as cited in Medhurst's Chinese Dictionary. "l Not having examined Ellenic with a view to deteot hamsa, we have no settled opinion in regard to the an- cient spiritus lenis. 120 ANALYTIC OBTHOGRAPHT. interruption to make a modified liquid; and the vowel and scraping effect being 8imul- tancous, they cannot be represented by a consonant character preceding a vowel one, as , H a. We propose a minute < below the vowel character. 571. Tlie Sanscrit visarga (:) is a final "strong aspirate" (Sir W. J.) which becomes a in languages which do not admit it, as the Hebrew final oi Jonah, Jeremiah, is either pro- nounced 8, or supprest. The Sanscrit sound was probably h pronounced with the mouth not sufficiently open, causing the breath to strike along the fauces and palate, thereby re- ceiving a modification suggestive of % and «. We represent it by the figure 6, which is sufficiently like one of the German forms of capital h, whilst it is equally suggestive of a. 672. The following ayatema of notation have been proposed or used for the members of this, and of the preceding contact. FAUOALS. Q Hebrew, Volney, Lepsius, Richardson, Mttller, Paulmier, Eichhoff, Ellis, S'unic', Riggs, * A peculiar form on this basis. q k q k' K q 'k Q a X t kh 'A kh K' X h LARTNGALS. 2 y i gh 7t r' 9 g g h n h h • • h h H h h h n h* U h • h* h' Ht h* a' a' «« •h ••a, &c. 8 .iw^' ** The Arahio let ter. f With a short vertical medial line. LARTNOO-FAUCAL. Q q. > > 573. In the Waco of Texas, the entire surface from the glottis to the Q position, form a contact which is opened suddenly and independent of the lungs, upon a vowel conforma- tion, producing a clack or smack like that which accompanies the separation of the closed palms when wet with soap and water. The preceding closure bears some resemblance to the incipient act of swallowing. We describe it from our mode of producing it, and we were said to be the first person with whom it was not vernacular, who had acquired it.* * We pronoanced Arabic q, qh, and gh, and Welsh tt, rh, as ascertained from descriptions, before they bad been heard from natives, but we did not recognise hamza from the descriptions, althongh we were familiar with the phenomenon. • * AKALTTIO ORTHOORAFHT. 121 574. The/ollomng toords are Waco (ve'co,) the r being the vibrant European element. The word for noee (ti'sa > ',) is (except the first vowel,) whispered. Oitic lb r, T, r h L 3 J»'J asper 6 •1? -fr, - lenis 7 . • - asper 8 ''v "1 "r/rr -" - '3 'J. 5 h,J lenis 1' m D - n. n^ r asper 2' "i 'n - !.?§>«' - lenis 8' *« - - - • • asper 4' «m a lenis 5' b d - d. i. Q - asper 6' 'b&a - 3 f> 3, J 2 '0 2 lenis r P t - t. K c qq ► asper 8' 'pfiJ- SOS} K f q 'c 'q - §490 *^': »Si-»*4 •^■i.^^/f mn 122 ANALYTIC ORTHOGRAPHT. »1.' . .■,>■-! . >1 *::f -f'-A t,>t. .,•-» ■ I M •*•>'*■•■.".' :) ■i.:i^ -'«,;ui-,,»-'i •! ' ..-. ri^'.lV' CHAPTER XVI. I'l , > EXAMPLES. The diffioultiei attending the oonitniotion of a phonetic alphabet are so great, that those who have not spent manj months over the task, can nave no adoqunte conoeptinn of them. After the invention of au alphabet which seems theo- retically perfect, the luckless inventor too frequently finds, that when practically applied it will not realise his expect»> tions. Even should it work tolerably well, the difficult question arises how far to employ it properly. Phonetic spelling is more difficult in English than it would be in any other language, though if the Irish or Scotch pronunciation were adopted, or even that of the laboring classes in the agricultural districts, the task would be comparatively easy. — Phono- typtc Journal, 1846, p. 156. In expressing the sounds of a new language . . . the missionary should be guided entirely by ear, without paving any regard to etTmological considerations, which are too apt to mislead even the most accomplished scholar. Max Muller, p. XX. ... we feel how essential it is, in a first attempt to fix a spoken language, that the writer should not be swayed by any hasty et^ to decipher it. Id. ical theories, xxxi. The missionary should give a true transcript of a ipoken language, and leave it to others § 678. Some languages are readily written, even by children, and it is difli( u '■ j J J 1 m m n r p r T r 8 2 3 rf 3 r J 1 a t tt u 1 V V % COLLOQUIAL ENGLISH. 600. The fdlovoing example is transliterated from one given by Ellis, (Essentials, p. 104,) being his translation of a passage from Pott's Etymologic Researches, and printed here without any perversion of tJte powers of the Latin alphabet. 601. For convenience of printing in this version we use J for the more anterior j , b for a, r undotted (§ 501,) U for U, and j, v, for the coalescents; and although J, j, are identic, J is preferred in certain positions when printed, to increase the distinctness. 602. The vritten and printed representation of the sounds of language by means a« ritn end printed rep^iaenteran ol wa sAvndj at UfGVidj bAJ min3 of characters which are insufficient both in kind and number, and which must therefore OE CA'xactTa, "vvir ar insBft'rant, boi in cAJnd and n^'mbx, and v vitr mBst atxfor be combined or modified if we would give a graphical syrabolisation of bi cBmbAJnd at [§ 403, 7, 8,] modafAJd if vi vud gie b Grificl simbl'iaeran afc the phonetic elements with only some degree of exactness and convenience, has been, from OB fonetlc ilamants via onli SBm diGxi aE eGsictnes and CBnEinjans, hjis bin, frum all time, for nations as well as individuals, linguisiical students not nl tAJm,fnT n^rana aa vel a? indaEidJual? [indaEidjal] It/GVisticl »tjudnts nbt excepted, one of the most necessary and one of the most difficult of problems, and has £C8iptad, VBU Qh QB most nesasari and VBn aE av most dificalt oe prbblama, and hxa consequently scarcely ever been happily solved. Let this teach us that the invention cDnsicvantli sce'rHli ehv bin hipali salEd. let ais titr us OBt oBinEgnran of writing, the greatest and most important invention which the human mind has QB TAJtij', QB Gretast n most impn-'rtnt inE^nran vvitraBJJU man mAJnd[m8ond]hA3 ever made, and which, as it indeed almost exceeds its strength, has been often iEu raed, and "vvitf, jl3 it indid nlmost ecsid its strtj-? [stren??] h^a bin ri'Tn and not unjustly attributed to the gods; like the organism of a state, at once [ofn] and n&t BndjB'stlt itn bratad tB as Gn-d?; Iajc ob orGanum aE b stet, it vBns simple and complex, is not the work of individuals but of centuries perhaps of Bt'mpl an ctiraplecs, is not db vbtc aE indaEi'djuals hot aE sint"} j Bxia paxhips aE thousands of years. ' '"" "* lATsnda aE Jir^. 128 ANALYTIC ORTHOGRAPHY. y': 603. This upecimen has suggested to ua the probability that three kinds of r will confuse many writers, so that practically *r' initial, and 'r' will be sufficient; but if *r' is re- stricted to a single tap of the tongue (§ 501,) it should have a superior dot — neglected here, from the difficulty of printing it. 604. We at first wrote 'students' with 'a,' then excluded it, and we changed the second vowel of * individual' from t to a, although we use i in our own speech, from heterotypic influences, as we believe. We think the cases rare, where the same vowel occurs thrice in the same word. It seems contrary to the laws of English. 605. We had written 'representation' without a vowel in -tion, — and 'invention' with it, but finding a vowel aperture to be made, we wrote 9, which is correctly placed in pro- blem, convenience, greatest, exactness, — but is it not 8? in human, and difficult. We think not. Nevertheless, a difficulty in discriminating them may require b to be used for both, in the one case (a) dotting it beneath. At first we assigned b to organism, then a. 006. The words 'therefore' and 'scarcely,' in which the open e of th^re is shortened (as indicated by the acute accentual,) without closing to s — may really require the use of 'e,'§ 385-90. 607. FAVST, ' ' ' ' ' 'translated by Emma Stanwick.' Read n, i, e, o,\ long, unless otherwise marked; and t), TJ, I,', short. Thou full-orb'd moon ! Would thou wcrt gazing now, QAV ful-nrbd' mu'n ! vud oa'v vurt GesiG, uav. For the last time upon my troubled brow! f :«* fnr[§403]aB last tAjm vpb'n mi tr^'bld brAv! * t , * v- Beside this desk, at midnight, seated here, bisAjd ais disc', it mi'dnadt, sited hir, , , .., c. Oft have I watch'd to hail thy soothing beam ; ■ , ^^ . !. nft hiE sd vQ'trt tB hel aae siiair brm; Then, pensive friend, thou cam'st, my soul to cheer; aen, pinsit frind', qav cemst', mi sol tu trlr; ;.: Shedding o'ei book and scrolls thy silv'ry gleam. i fidir or bifcs nd scroU Dae si'Ieti Glim. O that I could, in thy beloved light, o aet OB cu'd, in aae' bilB'tad" lae't, Now wander freely on some Alpine height; nA'v vandr frill bn sym i'lpm hsBt; Could I round mountain caves with spirits ride, cud Aj TAvnd mA'vntan ceEs via spirats rAJd, 608. 609. ,'..^^1- ANALYTIC ORTHOGRAPUY. 129 se e- In thy mild radiance o'er the meadows glide, in aaj majld'" redjans o..r* av midas clajd, And purged from knowledge-fumes my strength renew, and pu'rdjd frvm nb'lidj-fjuma mi sixin. rinju', Bathing ray spirit in thy healing dew. beaif nil gpi'rat m qaj LiUr dju'. *'or' in line 6. GERMAN. 610. TJie iiexi is the original of the preceding example, which we retransliterate into German from Rapp's phonetic version (4, 92.) We follow Rapp's pronunciation, except that he uses a alone, for our a and a. Had the phonetic version been our own, w^e would have put 'mondan' in the first line, and 'Jetstan' in the second. The syllables without vowels are our own. Read u, a, short. s&hest du voller monden-scbein, o aest du[tu?]fblar monden-fAJn, might you look full moonshine Zum letzten mal auf meine pein, tsu ..m k tsten mal Avf mAJne pAJn, for the last time on my pain, Den ich so manche mitter-nacht den iq so mAnqe mitr-nAct [nA;ft] that I 80 often midnight An diesen pult heran-gewacht; vn diaam pult hcrA'n-OeBAct; at this desk here watched; dann Uber bUohern und papier, dAn y~br by'qrn und[unt?]pApir, then over books and paper, TrUb-sel'gcr freund erschienst du[tu?]mir! try"b-3el..Sar fr&jnd arrinst du mir! sad friend shine you to me. Ach konnt' ich doch auf berges-hohen AC cH~nt" iQ doc Avf btrsas-hnn could I but on mountain-height In dcincm lieben lichte gehen, m dAJnam liben li'qte cen in your loved light go Urn bergcs hSble mit geistern schweben, um bergas ha"le mit GAJstrn pBebn, round mountain caves with spirits hover Auf wiesen in deinem diimmer weben, A'vf iiaen m dAJnam d^mr Beb^n, over meadows in your radiance float Yon allem wissens-qualm entladen fon A'lam iisens-c^Alm cntladsn from all knowledge-vapor unburdened In deinem thau gesund mich baden! in dAjnam tA'"v. Gesu'nd miQ bad*n ! in your dew salubrious, nie bathe. • I WESTERWALDIAN. 612. TJie following is the first verse of a popular poem in the German dialect of the Westerwald district on the east side of the Rhein. It is given in K. Ch. L. Schmidt's Westerwaldisches Idiotikon, (Hadamar, 1800,) under the title — Das Hotzel-Mous-Lied, oder Lob der Hotzeln. A hotzel {hufzel in Pennsylvania,) is a dried apple, pear, or peach, 130 ANALYTIC ORTHOGRAPHY. .i 1 f especially if dried entire, and moua in their cooked condition. The first line of the origi- nal stands — Nu ha n' eich all mein Lebclang — where n' seems to be a fulcrum to prevent the concurrence of two vowels. The i of ich ( /) will be observed to be diphthongal, as in English ; and, in fact, most of the shades of English pronunciation are present in the idiotic forms of German and its cognates. 613. Da"s hotsl mus. nu hb -n- AJQ a"1 now have I all raw i my le'balAf life-long nAVt b^sars a"s naught better as hotsln 'hootals' g£sa, eaten, 'b5n ..Q ter en, het'.. [hti] da 'ber ['^or] ajq crAYc, when (wann) I of-it none had, then were (subjnct.) I sick, 'b5n ir ot reqt [rest] 'bait 'bisa if (tcewn) you it! right would know. oarapta mur..n an SAvrorAvt * grated carrot an' saurkraut A. .A, as A'c no'c ^5*"^ ibas [ipas] GUS is eke(a«c7t) yet something (etwa») good doc A'ct -n- AJq dAf crA" bR DAVt though regard I that quite (gar) as naught, an (sa hotsl mus. an' eat dried-fruil i mush. FRENCH. 614. The following table shows the discrepancy of opinion among the French, upon the value of their vowels when compared with English standards. The first column contains the French examples, and the others the words supposed to contain the English equivalents. Le Brethon, Bolmar, Value, Picot, PantoUon. patte pat fat add at • • • • pate pall arm far father arm b^tte bet fate gate fate ale bete bear where get there dare hotte hot not no, nor nor, over .... hote hope more nor « t< old 616. The older alphabets are not worth quoting. In the Miscellaneous Works of Wm. Marsden, F. R. S., there is a paper On a Conventional Roman Alphabet, where a is pro- ANALYTIC ORTHOGRAPHY. 181 gi- icb of the ins Its. m. re- posed for the English and French — fall, male; a — sad, far; a — manner; e — It. vero, Ger. lesen, Fr. cher; e — It. nello, Ger. bett; e — Fr. pres; i — Fr. long i; i — the same short; (a correct feature;) i — sit. It. piccolo, Fr. quitter; 6 — glow; o — motive, (a correct view;) 6 — not. It. dotto; u — Ger. and Ital. ; u — but, "In high German it is denoted by ii in fur." — au, out. 616. TJie example following is nearly a transliteration of that of Mr. Ellis (Univ. Writ., p. 21,) whose pauethnic notation we consider the best among the several modes proposed by him, and which the want of type alone prevents us from quoting. He indicates long vowels by a repetition of their character, which makes the sign of quantity heterogeneous (now 0, now e,) and too conspicuous ; nevertheless his palaeotype admits of a high degree of minuteness. 617. We use Jiere tJw small capital i for the long sound, and the dotted i for the short one. The elided e is sometimes represented by two dots. The e of de, le, se, we write with the vowel of up (perhaps incorrectly,) and using (u) for it. For convenience in printing, we use o for the long and o for the short sound, §412. A period point before an initial indicates a capital letter — capitals, however, are no part of language. Calypso ne pouvait se CA'lipso" n.. puis s.. elle se trouvait malheurcuse el sa" truLs mA^lu'ra^s lea nymphes qui la servaient le ni,f ci 1a serEe gazons fleuris dont ga"3o< flu^ri dat d^bria d'ua navire qui d «" nA^Eir ci dcartdes i^k ecA^rte" sa" debri rainea TXms flottant flO^'tA. 618,619. consoler du depart d'Ulysse. DaDS su duuleur, CO "so"le dy" depA'r d.. .y"lis. dA" sa" dular '^^^'^ d'etre immortelle. Sa grotte ne resoDoait plus de son chant: detr imortel. sa orot' nvT reso^ne ply" d so^ fA,": n'osaient lui parler. Elle ae promenait aouvcnt scale sur les nOse ly"i pArle. el sa" promne sula, sa^l sy^r Is son ile.... Tout-a-coup ell appcr^ut so.n il.... bancs de bA; d gouveruail, tut a" CU el Ics le A persy rameurs mis en pieces, rA^raar mis a" pies, un m&t des cordages des de GUEernA^il) <"***" a, rax de cordA^j un printemps eternel bordait u prjL"tA"3 eternel bo^rde Tcnait de fairo naufrage, des Ene d fsr no^frA'j, de et Ik sur le sable, un e 1a" sy"r 1b sA^b^l, b, sur la c6te sy"r Ia" co't. Translation, in French orthography, from two French Treatises on English. 620. Calypso koud not bi konforted for thi dipartieur ov Youlysses. In heur grif shi filt [faound heurs61f] eunhappy ete (§378) biing immortal. Heur groto no longher rizaounded ouith heur song. Thi nymphs hou served h§r derd not spik tou heur. Shi of^n ouakt alone on the flaouri (§129) teurf, ouith houitch 6n iternal spring covered heur ailand On & seudden shi persivd thi fraghm^nts of e v<^s8el that had djeust binn 132 ANALTTIO ORTHOORAPHT. T^kM, To6n b6ntcbgs broken in pissSs, on skattgrdd hir end th6r on thi sand, 6 reuddSr, e mast 6nd kord^dje floting on thi shore. — P. Y. de Seprh. 621 . The next ia based on Rapp'a example (3, 141-2,) from Moli^re's Tartuffe, act 1, sc. 6, the orthography of the original being our own. Instniit par son gtrfon qui dans tout rimitait A.Btry"i pAr so, OArso, ci dA, tu 1 imite et de son indigence et de ce qn'il ^tait, 3 d 80, A,dijA,8 e do 8 c il Stf, je lai fesait des dons maia areo modestie j« ly"i &« de do, me ^*"'> albc modesti il me voulait toujours en rendre nne partie. il ma Lule trjfur a, rA,dr y^n pA'rti. o'est trop, me disait-il, o'est trop de la moiti^, B e tro mB diaetil, 8 e tro d Ia myAtiS, je ne m^rite pas de vnns faire piti4. JB n merit pA" da WJ fer pitiS et qnand je refusais de le vouloir reprendre, e CA, j» rfy'ae dv 1 Eul^Ar rprA,dr auz paavrea, k mes yenz, il allait le r^pandre. po"Er, a" me ju" il Ale 1b r6pA,dr me le fit retirer mv I fi rtire tout semble y proap^rer. tu SA,bl i prospSre. enfin le oiel ches moi i,fA, 1b 8jel re" mvA et depuis oe terns Ik e dBpy'i s tA, Ia" 622. Boih Bapp and Ellis write French nasal in "on the basis of the vowel of e&&, the Polbh e, in pje.to (five,) which is concurrent with i., "in Wyandot. Here the practice of Mr. Ellis is based upon French opinion rather than upon his own ear, as he has informed us. 623. Most of (he suooeeding examples were taken before we had distinguished a from b, and the open u frcm the close u, so that the one may often stand for the other. CHEROKEE. 624. The Lord's Frayer, the native version. Bead n in /all, i in pit, b dose and shorty in up. 'noitnta " ofilB.latt- »hehi' « oalB,cti5e{ju; ' ceseti' • tets&txiB.i'. onr-father (in) heaven who dwellat honored be thy name, ' tsiOB,vijuhi' • cesB.' ' vrcftnSnu-cm'. " &fint " Slnhi " vintflTlijBta " hitinnte-sOB/i' thon-king being let-appear. here let be done thy will ANALYTIC ORTHOGRAPHY. 133 «r, 8C. the I of U8. '« nahsci'ja '* calti "lati' '• tsin'ca'Ij,stiha. " nltatnte'cvlsti/ '• xica'ljstaJV.ti' as it is (in) heaTcn done. daily our food " sci'u/sv "cnhi' " i-ca. "ticesciB~siovo"n&/ "tescviu'cua' "nahscija " tsiticajn'tsinnelm', givo U8 this day. aud forgivo us our debts (") as we forgivo '" tsntsituci', ^'&U "cle'sti' *»u~tac5'lijstiji' 'o vitbscljatinu.'stanti.oi', those vho owe us, and not in the way of trial lead us, " sciju-taliscesticvo'scini' "uja' "cesB/i'. «« tsti'stselicahje'nt)' " tscctj/vtju-hi "Cssb/V but deliver us from (the) evil that is. for thine king (^) being.is " t8a"llini*citi' "ali »»cesB,i' " e-tsdB,/c"vtiju' «ale "cssb/ " nioahiltt/i' "iiasci' strong and the-being honored and the-being always so ** vinia"lsta. (") let it be. ' 625. This is the pronunciation and rendering of the Eev. Stephen Foreman, a native. The final of », ", &c. is not a coalescent. The accent in ' may be erroneous — see ". The final of " is probably i. It seems ('•, ",) that flat t, c, do not associate with e. 626. The whispered vowel in ", ", ", should probably be omitted as an error of the alphabet (§ 589,) as in No. «. The vowel b in »» is nasal, and whispered, being between surds. The final syllable of " means and, like Latin -qye. 627. The final syllable of •• would be omitted if not followed by a stop; and if " ended the sentence (the verb in the three last letters of ** being implied, or given in another place,) it would have a final accented i. This peculiarity appears in » and ". In ■< the final hjend' means ybr; and in ^ tsn means tJiose. 628. We have taken but one liberty with our manuscript, namely, in making » correspond with " on the authority of the version in Gouraud. We had written " '^naskija," failing to catch the h heard in " (if it was sounded,) and also the flat sound of cay. At that period we used Je, and a dot above for the short quantity, a good enough mode, but difficult to print. itt, WYANDOT (=Vt)NDbT.) ' 629. The Lord's Prayer, the notation the same as the preceding. scvB">isti 'ja'r6,nJa>aj6 » i;g8ta'CE>'; * tuji hti * de • f£finda>-'; ' tu-tave our father heaven in tbou-inhabitest righteous thy name. let-it-come • saxo.njameh. » avatTivaJTit' '» de " saTivb^f*' " o'm^tsa>aj£ " tiju'hti '* de your heaven. let-it-be-aceoinplisht thy purpose earth on like-os 18 134 ANALYTIC ORTHOGRAPHY. " jar6,nJ,a>aje. '• tava'no.t "da>atemi,ntaje '* macja>atanditS,hcvi. " sB8B>aclij6^r,h,e^ heaven in. give-to-us evcry-day ... our-Biiatenance. forget-thou *> ds " sBTiiB>ac6.ndih " tijov&rreha> " du "nj,o;mB'>B" " o-cirijaVaco^ndih. nj,o;mB>B own sBTijB>ac6^ndih " tijov&rreha> " du our breaking thy laws as we do * our own law breakers. " tavb'hsarit "to>omB'h " di ''sti>B''h *" ta>atanJo'mB'htTatB"t " ducau'ht, lead us that way ... not to be beset (by) evil, ^^ somi'.h »de **jaT6nja>aje '* ds " ja-vi'hrTa>' *' tu'ndi' " du ta'rai>*. thine ... (the) heaven and (the) power and-likcwiso the glory. 630. This version was composed in our presence (we writing it down) by the intelligent native chief of the Wyandots, J. M. Armstrong. Wyandot is an Iroquois language, and thp three first words of the precedin«» version correspond with the four — "Songwaniha ne karonghyage tighsideron," as given in the Mohawk 'Common Prayer,' London. 1787, p. 53. Zeisberger gives garochia as the word for Jieaven, in the Onondago dialect. No. *• *> ^^ &c., have the common h. The o seems to be always nasal, and in number " it is probably erroneous. 631. The elements in the language are — i, i, e, e, i, b, a, t>, o^ u, u, — v, m, (no other labials,) — n, d, t, t, (no ?,) — x smooth, s, (no z,) r, j, — j, G, c, c, ;^, h, >, besides nasal vowels, u is used for a short sound without discriminating it from u, (§ 623.) 632. u>ucjeTt~he duVunda>''. u'>unda>' ju'>UJ,e>' vo'ti^jo hB"nda"rio'. it is straight the-arrow. arrow. §486. ground-squirrel in-a-hole lives. Gja> ara, Niagara, probably from cja>acd> , broken. scn'> nta, head. cveujn>nta, cicada. jU5>', pigeon. tsa>andu^sce, Sandusky (=at the waters.) ilicesi, Allegany. ajnda>', bow. hntrro, rakoon. jentso, fish. cv6se!>-', fowl. h6ta>-', ear, which some may consider akin to o^Ci o'roc* nj?)teru>', my friend. ^^ ^^^, ' NADACO. 633. Orw of the peculiaritiea of the Nadaco or ' An-a-dah-has' has been alluded to in §448. Another is the occurrence of the vowel u or n, the Latin consonant v, and the allied coalescent 'v; also, i, i, j, j. We heard a man call a finger-ring nace''8embe''-ca~s8*',* whilst his wife called it nacesembetrahasB, with an additional syllable ha. See the word for finger. The vowel of add occurs here, and a final vowel is often whispered, as in eight of the following examples. * Although we use (') for short accented syllables, and Q) for long ones, the accentual leaning towards the co-accented consonant, yet when we use Q'^ together, as in this place, the syllable is to be considered as made of ce^ and not of £"s, — and 4'» might occur also. ANALYTIC ORTUOORAPUY. 186 dasbtp crown trohotd, hair tsahatau, forehead tanoadaus (not tar-) cJieek birtu, ear tsahav, eye trabehetaT, brow 800, nose tameso, jaw adetp tongue ate"'t9, warm vacoho, cJiin natseo, neck behedavso, shoulder nansh, clavicle cor, Jieart co's, breast tsoto, nipple sentd'hp, wrist eeco, palm sembeto, Jinger sembssas, thumb basoboto, leg bided, knee nahatoh, ankle ha' 80, foot nahcuha'v, sole nastsoto, heel nauoto, toe nati, wo7nan tanaois, leech tA'nat', gryllus etfi't', toad Cabs on, coat oantaso, leggins Conavta/co, hatchet nvc^oabriva, pipe nacimpi, beads tapit', fan vai, shoe ea^bav, Aotwe cantaibf, mirror cooe, icater vatot', ground aco^hoto, cold ha'hat', good. KANSA (=C5nst).) 634. The vowel Y, French u, is found here, although very rare in the aboriginal lan- guages of North America. In our examples we add (in parentheses) the Dacota equiva- lents, but placing Riggs' diacrits after instead of over the letters, as g* (which is compared with ghain,) s', English sh; h*, *a deep surd guttural;' c', Eng. teJi; z' Eng. dzh; n, as in English sing, and French bon, the two being confounded after eighteen years study by a number of missionaries. Probably both sounds occur, as in Kansa. ear. nota (nog'e) eye, iftate^/e" (is'ta) brow, Jftahifaba (is'tah'e, eye-ridge.) mouth, iha (i) tongue, jisse* (c'ez'i) nose, pah* (pog'e) nostril, poru're (p6g*e-oh*d6- ka, nose-hole.) * Pronounce each s, and the / forehead, plessss (it6) fan, tcIlajB (ic'adu) nonoba (c'a^^duhupa, c'andi, tobacco.) mohf (minna) mB.rtjea'v leggins, hy^/ba (hu?iska) shirt, 5sc5'sc6ucjdo,cuda {on- h'doh'da.) pipe, knife, vmrm CHiPPEWAT. (otfibv^, pi. otnbv^G.) 635. ne^nepatrintc^W, the mole, being inic for onic, an arm, his arm (ninic my arm, cinic thy arm,) ne.patr, wrong, left, opposite; ne- its reduplication, for both arms; e si', a noun suffix of the animate gender. The mole then, in the view of a Chippeway, is the 186 ANALYTIC ORTHOGRAPDY. animal with reversed arnm, the right one being apparently on ihe left shoulder, because the palms, instead of facing each other, are exterior. 630. psjiGoGBji, the horse. For pcjiGOfCBjI; from psji'o one, Srcoj hoof, nail; the siuijle-Jioo/ed, or solidungular animal, this being its zoological characteristic, and one which very few Europeans would have observed. How few, for example, who have seen the gnu (=black,) and the camel, can tell whether the feet are solid or cleft. The Chippe- ways name an elephant, not from its trunk, but from its straight or columnar legs; and a sheep from its ' ugly hair,' the wool striking their attention unfavourably. In Bishop Baraga's Dictionary of the Otshipwe Language, Cincinnati, 1853, the word for Jiorae (bebe- jij.oganji,) has an initial reduplication, like that for mole. In Choctaw, a horse is I'soba, from iBi deer, hu'lba resemblance. In Nadaco, it is the Mexican Spanish cubajo, which varies to cavaru* (Eng. w, trilled r,) in Waco. Similarly, in Penobscot (here t for tuc'v means river, compare Aroostook,) the English name with its article, appear as D.hb.s; and ahahsB, and a buffalo as babulo. 637. mm (min, i being used for I,) huckleberry, pi. minen, mi^n^,s, thorn-apple. Datura stramonium; es, dimin. mini' a round sore, mini's an island, mirimi'n, apple, (great berry,) written mishimin by Baraga, miti'Gvnb, a bow, because difficult, (ovnb,) to draw or bend; nin vnb, nivab, I see; vabefc, to-morrow (the time of being light,) viGivnm, a hut, from sheltering, in Lenape — vicvbm. , ' » ■ • » /r • , ,t 638. The musTcalcmge or great pike of the lakes, is in Chippeway morcinojs., from mor- great, (compare Mich-igan, Missi-sippi,) cmo js', pike, and cinu'rciji is any long-snouted animal, as a hog. Compare j>i<7 Kn^peak, pike. vB'ju'rc, muskrat, Bjarci', mud. afcibvGcnn, (place of artichokes,) Cheboygan, the orthography of which is French. mBrciG, swamp, whence Maskegon. min'-sibi' great river, atntamo^, the red squirrel, Sciurus hudsonius, because it descends trees head foremost. ase'nB^Go, grey squirrel, that sticks fast, or close, (to a limb.) 639. Tfie poll/synthetic structure of the Vesperian languages is widely spread. In Aztec, according to Humboldt (Vues des Cordilleres, p. 316,) a kiss is tetennamiquilizUi, and pain is tetlayhiouiltiliztli. Condamine (Pinkerton, xiv. 225) thus speaks of the Tam^oa of South America. ''The language of this people is indescribably difficult, and their enunciation still more extraordinary than their language. They draw their breath in speaking in such a manner that the sound of scarcely one vowel can be distinguished. They have words which, to describe, and then but imperfectly, would require at least nine or ten syllables, though, as pronounced by them, they seem to consist of but three or four * These forms are sufficiently like the West African Gr^bo c^boso (horse,) to suggest an identity of origin. But this is from cd (to die,) in this manner. The peculiarity of the white race iu Africa is to die in a short time, hence c6bd dying kind, is the word for a white man ; so is lizard, so that a horse is considered the ' white man's lizard ANALYTIC ORTUOGHAPHY. »T Poeftarorinrournar, signifies the number three in this tongue : happily for those who have transactions with them, their' arithmetic goes no farther."* ' 640. Pitchlin, the intelligent chief of the Choctaws, gave us the etymology of the Choctaw (=trt)'tftf) word fo'nobi (iron-wood, Ostrya viroinica.) It is for io'norto.bi (with all the syllables short, the third with the secondary accent,) contracted from iononrtubi, that with which kill buffalo, (as a club, arrow, &c.,) their bows being made of this wood, io'ner huffitlo; ut that with which; o'bi kill. GREEK. 641. m2*nin a>jde ^e\k pS|lsia;de'o &Z»lls*6s . .> . v ^ . 6'vlomi[nsn, hs] myrf.. &|;KajdJ8 I aloe" ^jS^Scc. — Tliad, I. 1,2. The xext is from 1 Corinthians xiii. 1, 2, being the passage transliterated by Mr. Ellis. We preserve the accents, t9, and ;f. 642. can tajs Glossajs ton anSro'pon lalo caj ton afOflon, ilGapsn de" mS' e'Kp, GicSna ;falc6"s e^on, e'cy'mbtllon alula^on.— caj e&'n e'xo pro'pStejun, caj fjdo ta, mysts'ria pu'uta, caj pasan ten Gnosin; caj can £'^0 pasfm tgn pistin hu'ste o'rg me'>istanrjn, uGapsn di ms e'jfo, 6 v«?6n sjmi. ITALIAN. . , . THE lord's i'rayer. t 643. padre nostrw" ce eel nel trls'll, sia santificata>~ II n«»me tiiw", hifan, II ren},*"' tiiw"; sia fatta Ifi L(u''Ia7nta tiiS, ca>me in trlsloi" ccwsi In tirru; dattrl [give thou] oddji II nostra p^ne cvw'tldl^nw", € rlmettl a na»I I no'strl debltl siccwme nwl II [for them, accus."] rlmettlamo; a no'strl debltcoVl, e n«>"n tf„ [us] Indiirre In tentutslcone ma liberutrl [-trl, us] d.. al male, t cwei sia. {amen.) ■■' -s— , . ... ♦ LATIN. TBZ lord's FRATER. 644. pater noster cvl es in cojlis; safctificetur nomen tuu,. a'dveniat regnu, tiiu,. fiat voluntas tua, eicut in cojlo, et in terra, pane, nostru, supe'rsu'bstantiale, [cvotldianu J da nobis hodie. et dimitte nobis deblta nostra, sicut et nos dimittimus debltorlbus nostrls. *"The sounds of the 'Tinn^ language can scarcely be expressed by the English alphabet, and several of them are absolutely unpronounceable by an Englishman. In my attempts to form a vocabulary, I had great difficulty in distinguishing several words from one another which had dissimilar sounds to the native ear, and were widely different in their signification. A Dog-rib or Athabascan appears, to one unaccustomed to hear the language, to be stuttering. [§551.] Some of the sounds must have a strong resemblance to the Hottentot cluck, and palatal and guttural syllables abound in the language. Vocabularies of this tongue cannot be greatly depended upon, as no two people will agree on the orthography." — Richardson, Arctic Searching Expedition, chap. xiii. f This word cannot be spelt with the English alphabet, although every element is Euglish, the vowel being that of odd, as in Kansa (=cO'n3&.) % As pronounced by Mr. P. L. Rosteri, instructor in Italian at Florence. : - m< 188 ANALTTIO ORTHOORAPnT. ot no nus inducus In tuntutl6ne,, edd libdra nos a mdlu. nu, tuu, est r^gnQ, lit imperin,, ct mugnlfTcontlu, In Hcmplt^rno. — Gouraud, pi. 12. 045. Our variutiona from the Latin text are due to the inconsistonclos uf Latin ortho- graphy in the use of Q, M, U, X, — and E both as a coalcscent and a vowel. In several cases wo mark length 'by position,' where the vowel is naturally short, by (). We omit the coalescent dots, and write aj, oj, for 00,00. VIRGIL. 646. //} the following example, the first and fourth feet of the first line must have no accent, because the verse has the rhythm of time, the ear being informed by the accent of the fifth and sixth feet, that the measure is hexameter. Vulnus, at the end of the third line, has its time made up by the consonant at the beginning of the next line, or by a comma point. Aeneid, I, 84, 35, 36, 220. consp^ct they {iho vies e sc&rcc were v^la da bant lajt sailTng in U 81CU laj lys cum jun.. whea jii to white- ajt^r no h\ prajcipu e now chiefly plus the pious frOm siot I et spu mas capt wilTOB their nu"* Servians sub wftys h(kld|ing her ajn^ as nti/c rone as the tell shore sails ni6tal uns m out tn th^ ajre ru parting the p^ctore woiind in re acrls 6 fate of the altu, high sea ^bant — witters, vulnus — mimbranQe — . rbntej— I. 220. active, &o. 647. The false * hexameters' of Southey, Longfellow, and others, together with our accentual music, crush the rhythmic sense which Latin verse should have fostered, and gives us the barbarous relish for the rhythm of noise which rustics exhibit when they think their step in the dance should be heard as well as seen and felt; — the dance (the ancient chorus) being the only rhythm of time we are acquainted with. 648. The last Latin line, therefore, strikes the modern ear as a five-foot measure of English amphibrachs — , ,. now chiefly the pious en^as, the fate of orontes and also of amyc and ly'cus or like the next, in English dactyls — the normal form as recognised in our music now ' the active bemoans him— chiefly the rontes and pious e also of neas, the fate of the amyc and l^cus be active o- moans him. ANALYTIC OBTnOORAPHY. IW GREBO (§ 351f«.) G49. The LonVs Prayer is given hero from the dictation of a native, the trannlation being furnished by tlio Right Reverend John Payne, KpiHcopal Missionary at Cavalla. Nasal and stopt vowels (§ 350) are very common. Vowels unmarked, as to quantity, aro Bhort, and especially so when stopt. Wo have probably not marked all the stopt vowels. a biia mb," nu, n« our Father tbuu ho art di jQ, na/ thoro heaven, thy mo. bo mi' mo/ va' ci, they m\M\ muko theo their king, for tcni, o' nl-di' ne, ju, h,J,i as they do-thore it heaven, give us be' p6/ amo,' h,v,t,'so'f co a do thou put us forgiveoesa nj,e ne. namo *wro bi cii ■ -fnp, nj,6, ' bo ' let it have huliiiess, men JU, na, "wro be' nu,f-di' cunf)' thy mind (will) let it be-dono-hcro world amo, nJ/.na.Jedo day bone, cucvi' a ta ieni mt), in for our conduct wicked ni'no/ e'ne,diba'di, no, this food, and a nl.' wo do a wo p6 put nJ,o,no, na.e. lead amo/ us cmo,' m5, for thou nJ,6,ne o' nl/ amo who-they they do us tudo ' tu ' a tld^ temptation its way teji ' CI, ' art king bone, ciicvi ' conduct wicked Je ne, and a -.,, their glory c&'mo, have tr time bie' all. Its account as mo.' hvtso't Ji; no na, unto forgiveness also; and not be ha amo (you) must tako us mO ve'-te-J^' ,X no' mb minD/' thou art-able-things, and thou shalt amen. Jidi ' , in. ca but ca devil from, de' bie' things all 650. In tJie following examples, when the languages are not noted by parentheses, as in (Armenian,) we have taken them directly from the dictation of natives. * This • IB perhaps nearer to ooze than to awe, § 418. f The penult vowel is more open tb -i it, or between this and eight. § 801a. ^ te, things; j for. 140 AKALTTIO ORTHOORAPHT. NUMERALS. 651. 652. 653. 654. /. 655. Mandic. ' Danish. English. Saxon. ^ German. ' sjtn ^ in, en Wan, v5nf ' 9jn ^ ^ Ajns ''tfb^jr " ' * tu, to- ■ .*tt^-7 •- - ^t:b^ * ^ "^ tS bAJ ' ^ 7rir ^ tn, tfe, -I ^ in 'drAJ ^ »drAJ * florlr * f ira, fife * for *fa-jara , *fir »fim 'fern »fAJL , 'fa'jba • ' fy-nf •"sees "sees 'sics " sisa "aces '' sJa" ,;: ^ sy% siE ' :• ,.• ^ setn ^ si'bana : ^ aiban * a'vhta* ' ota, ote 8et » a;cta » A-ct » nia ® ni, nl •nAJn i ® ni;Kana ' na'jn " tia. "ti,tJ. "ten. '- "tAJna. ' - " tBen. 656. 657. 658. 659. 660. German. Ilemi^i. BeJ/ian. '"• Valais. '"' Fa/ai8. (Westerwald.) V (Leukerbad.) (Sitten.) ^ ens M-en ' "" ^ en ; '': * Ajs ^aJ8 ^ ^ ts'be ^i:hh ' ' ' =^t;bet ^"^; " '^ ts b^ ; '^ ts bej nr3 'dvAJ =>(!« "^ "tn *fejr *ur - *fir * fin * f iri ' fimpf, fa^nf ' EAJf . »fA-jf » fifi, ftmfi ' fufi ' sees • 368 ■ ' 36S, sea ' seen « se'cfJl ^si'ba ^ 3eLn ^ seLen, seEen 's^P^^ 1%. . ^ si'bnl * o6t ««!•) "adt '' "act «&;cti «a;ctt|l '•noj ' neon ' neocn • nipi » nini " tsena. ^Um. "tin. : " tsa;cni. " tsA-^tnl. a? ■"•: J.; * The slight & was denied by the speaker. f Pronunoiatioc of Mr. Kean, Princesses' Theatre, London, 1859. j As we have used different notations at different times, we are uncertaiQ whether we used 'w' in our manu- script of this example, with its German or English power, and two grammars leave the question unsettled. JVaai (handsome)=frflj; Mi«(out)=8Jt; mi7(ow1)=b31; Ami (whey)=h8j ; Aomjo (aout)=ho"v; Aooi(hay)=hoj. The Belgian ut we have heard iu English; and in Swedish, nej (nay)=:nEJ. § Feminine plural, se'oro. • ,,,v ; II Accusative singular, a^tu',. ANALYTIC ORTHOGRAPHY. 141 6(51. 662. 663. 664. 665. Suahian. Suabian. Pennsylvania. Rtutian. Ilfyrlan. (Schwartzwald.) (Hohenzol. Hecb.) ^ 6.9S * ' 0,9 ^ ens, e,ns . Udi'n * Je'dan Ms'boa ,,^ '^ ts'boa 2 ts'be =*dEa ^ di.5", dia " trill, tri Hru "• trAJ " tri Hri * fira, fire *riri *fir * tretire ' tre'tri, tretirl »fAJ.f,.a" " ^ ftvjXi :. ' ftnf ' PJ^tJ * pet, pst' ® se'csd, 7 siba ® se'csi ip. " sees « ttStj * fSSt ^ si'bana, -e ^ si'bani, . , ^ St ba ^ sj'em ^ se'dam «axta 'ta:ti " 5ct * Lo'sjem * o'sam • ' nAJne ® nAJni " nAJn '•d36EJat3 » de'ift " tsene . "tseni. " tsea. "dj^sjatj. ^0 de'sef. 666. 667. 668. 669. 670. Dalmatian. French. Savoy. Savoy. at Sion. ^ Je'dan 1 b; ^0. ^ u"n, y"n 'o. '^ dta '^ du ««o. "^du ''du «du Um •'' trvA" ^ tre, tra Mrej nre * tfe'tM ♦cA-fr * CA-'t ♦ catro * catro 'P^t. ;; 'si;c ;^- "fA: "^sl.rc "Bl, " res T: "sis ''." "fi " sejs « BIS ^ se'dam ' set' (set') ^ SA- ^ ret' ^sat' ® osam ' y"i*t * VI « Vlt »vtt » d^LSt "nn^fsisi. » nA" ® nu "nu ^" de'ssi^t. ^»dls. " dA". ^°dis.' »° dj». "'Compare Allomanic (Bodensee, Aarau,) hast, hat; tuaB,/oot; iiaga, to fly ; liiags, to look ; liiogt, looked ; i'bal, evil. 19 142 ANALYTIC ORTHOGRAPHY. 671. 672. 678. 674. 675. Ll/OHS. MarwUla. Narbonne. Beamats, Spanish. (Provensal.) (of Pau.) U- ' -:»:■■- -■' r*' ^ 'jon - MafjJu/" » yn, yon ^y-n ^ uno '^du '^ dos, dus « dus ''dy-8 *d6s ' trAJ 8 tres 'tres "tres , Hres, tres * catro *catr6 *catre * cv^tre * cv&,tro Maj * BiA;, eef * sie/c ; i. 'fife ■ ; '.i.. »7i/co ■ " SAl «8ie « sles "rej ■ ' - *sej8 ^ 'SA ■ ' set, BE ' sU ^8€t ■: ■.■'; ^slite . > ^ "vr ''j/e,.lUs' « bejt ' vejt "otfo ® nH * nav, no • nov •na'v ® nuite, nue'te '' dJH". *° dis's, dss. "dets, "dutse. "dEts. " dl£7. 676. 677. 678. 679. 680. ' Catalonia n. Portuguese. Wallachian. Armenian, (^rnteniau.) ^ un ^ u,m ^ iinu * DlfiG ^mi ^ dos '^ diief ^ddj 2jer,GU' '^jirGu- =» tres • » tr^f Mrej ]-\l ' jeriic " Jirec * cv^trs * cvAtr5 *p^tru * tfu5rs, tfvors * tiO^'rs ^^ , "* ri/cve ^siTco Mrintf ^hlfc ■ = 'l . ■ .. . 'hi-nc .., «SIS " ssjp "rSse ,^ " 'bJets, ^ets "Lets ■ ' !. ^ set ^ ssts 'fs'pt}? A ^ jotan, -tn ^Jotn ,, , « bujt « vit6 :. '^P*^ ^ r^ * iitan, ■ tn ,,^ ' ut, utn ® noT « n ote * nove »inan -.•:' »inan ^» de'V. '°dSf. " d&sn. "dYsn. * " iiDBprssetf £, where spr is for iii^er. '^■'r-V_;-* * tJ ANALYTIC ORTHOGRAPHY. wm 681. Turkish. Mr ici 'y-tf ' dairt ' ber Hunffarinn. I edj" - '' cetta' ' harom, o? * ned}" ■^Ht 683. Albanian. df tre" catr 684. Ellen If. 685. Arab lie. ' VAhfit, — d. ^ i?nAJn ' 7A''lA"/A't * Umrix, tissni * A^rbA^H " 'qA'msit t S na aio ' trill pe nds « alto' ^jedi ' Bfcls «hot * 'hejt * n Jolts « djart* ^ftat* »t€t ® icse, ecsl ^ epta ' ' octo, octo, oeto « sIttAt ^ SA'bAt " 7AmA'nJAt,*$ ' ■ 'CVB .' "cilfints *' " nand' ' en a a, enja » tisrt " tl9. '» act. ' "as^ca. '■' >» A'rrA-t. < 686. Chaldee. ' haa ** tren ~ 687. (Striae. ^hao • 2 trm ^ 688. Coptic* '^ sn^'v 689. Welsh. ^ un, in. ' ^ d&j 690. Irish. ^ heBn ^do Mla7a * arba ' 8tl67o *arb{) ^ * f emt *ft6u • * pidvur Hri * ci'hi^r < ' hamra ' hamrS » tiu ' pimp * cuiG, cud} "I'fta ' faua - ^ tmanJa «ift6 ^rkuo ' tmatijo ** sou ^ re fit - * rmin « "vvecj ' V • 8vt7 ■ V-' "re ^fftct " bc^t « tira ^» a'sra. « tt'fS Vpsit te "> mst. ' nAV 10 deG. ® ne ^»ds. J-' * This ^ ticin ^ hahsB * cajeli * uiscli (c'h) «j^jBch(c'h) ^ tS^tB , * satecu. * »j6hto. " ujeli «**'> II This ,v seeius to He between y and u. ANALYTIC ORTHOGRAPnY. 701. Wyandot. * SCBt 2 tendt' ' ff/hc * ndb;fc ' uvi'i * uva-j ' tsut&xe > " 10 asi.h, 8£,h 702. Comanche, ^ simmus ^ « vha " '^ puha^t' ^ * vb'^tsuzt * <* manuchf (c'h) » •^ n^"vazt' « ' tatsuzf ^ ' na bev5;ft8U;ft' <*> * " si'.vbnBvubnut' ' ^° si- v&nnhut' ^^ 703. Niulaco. vi'stsi bith (fh) daha'v dj6ve desecat' dafci bi'sich da'vssc ivesec' bmaJB 704. Waco. ^ tfe'&3 ^ Vltf ' ta-v * tacvitf '*iscveto«*»"> * ciahio ' ciua^vitr ■^ ciata'vh <""^ ' tsiesci'nte ^° ClTlOVbbbD 705. Lenapr. ^ cvu'ti * nim ' nu'/u' * nAva * pAlenB;^ c * cvB'tar ' nirar ° percu'/c' >» telon 705 a. (^Lenape.*) ngutti niscba nacba newo palenach guttasch niscbasch chasch pescbkonk tellen 706. Chippeway. * psji'G, byiG ' nisvi' * nivi'n * iia,nB/n ® ifGodvnsvi' ^ nijvasvi'f ' nirvnsvi'f ® rb'vGBsvi', jrt) r ^° mitasvi' 707. Peiwbacol. * pesBc '^ nif ^ nabs *jeuh * palenBscTT ' nec'vdB's ^ tBmb5.,v,BS * nsa.sBC ® nolivi i^mdala 708. (^Pas.iamaquo(hIi/,)X ' niCBt '^ tabu ' 818 *ui6 ' nnn ° ca"mat8 ^ iloiGB'tlBC * oGme'ltfe * e8cv6n5,d£C " tilsn 709. Poleicutemi. ^ ngot ^ nif '' nsvt * nje o ^ nrtrnun " nGotvti'tso ^ novu'c * SVlVt80 9 10 raca mBt^'tso. * This is Zeisbergcr's version, taken in Pennsylvania in the last century; ours is from a resident of Texas. Zeisberger did not recognise the vowels of up, at, like lose who first wrote English (§585, 587,) and when the Dclawares have their men of letters, these may imitate the English orthoepists, by assuming that Zeisberger's spelling was strictly phonetic, and that it ought to influence modern speech. f This resemblance is unusual. Baraga gives Kven in Ghippeway as nijwatuwi, and eight as nishwatswi. Keat- ing, in Long's Expedition, 1824, (whose vowels we transliterate,) gives seven ninjuassoe, (he knew the French nasals, so that n represents our nasal sign,) and eight nishwassu. Six is formed on one, (Lenkpe, &c.,) ««t;ett on ftro, and e>V7A< on titto/o,) the second in the dialect of Macao, where the abrupt t may not have been observed except in the one case. But in our notes, the breath is indicated as escaping in the word for six. In the Canton word tor /our, we marked the vowel as made with the jaw open and the lips close, which would indicate a kind of French u based on the vowel of it. In this notation we used a sign like ^-^ for lips open; ^-^ lips close; i—>jaw open; i—ijato close; tj lips closed and jaw open, &c. ANALYTIC ORTHOGRAPUY. 147 720. Chinese. 1 JAt '^JI ' SA"m * sr »m 'luc » tf At' "pAt 10 CJA-=V SAp. 721. 722 Jtfa%.t (//atcojV ^s^tu ' acahl ^ diivS '^ ariia Miga ' acoru * ampad * abil » lima = » arima ^ anam " a6n6 ^lujut ^aoto »dmpan " « avD'ru " ssmbilam "aiEa 1° sablas. '» umi. 723. 724. Tonga-X » taha Grebo. (W. Africa.) »d6 ' iia ^sn. Molu Ma. *ia *hi. * nlma * m, Am "onS *rali.d6(5.J) 'fita ' ml38&.'(5.2) »Ealu « b6hX.bchX. * * * « hiEa " sigdo " hofOfulu. " pu^ t From the dictation of a Hollander speaking the language. X In Wallis I. the same, except i tkht, 2 liiii j in F&tuna, the same, except > ta«t, ^ lu^, » limi, » Jvi, '» ckuoiofnin. 1^ ANALYTIC OKTUOGRAPUY. CORRECTIONS AND ADDITIONS. § 181a. In some languages, j'i, id, eg, are used indi£ferently, and as we employ (' ') for sonant and surd, it may be necessary to have a p, t, c, /, or surmounted by (^) to indicate this indifference. In Baraga's Otchipwe Dictionary, it is directed that words not found under p, t, k, are to be looked for under b, d, g,. and the reverse. § 369, above € Suab., 'c Coptic' may be inserted. -, .^. § 624, 12th word, the vowel after lis not nasal (as marked) but whispered. § 379, note. At Covent Garden Theatre we heard pass, glass, man, with the vowel of fat lengthened, and passed, flaunt, can't. Mr. Kean, at the Princesses' Theatre, used the vowel of fat in France. Whilst the foregoing pages show the extent to which the Latin and Greek alphabets may be used, they exhibit at the same time a number of undesirable forms, which may be avoided by selections from the various types (whether in use or rejected,) published in Mr. Isaac Pitman's Phonetic Journal, at Bath, England. The rejected letters amount to 110, of which about one-half are capitals. All of these are accessible in long primer, and most of them are in perfect harmony with the Roman alphabet. On the other hand, the letters of American origin are in the aggregate badly formed, and cut without taste or skill. The fact that our own illustrations have been taken from about seventy languages or dialects— of which a somewhat minute notation has been made — renders it obvious that the alphabet of any single language must require a much less complicated symbolisation. ■•'4*fe^^'f;