LIBRARY NrVERSfFf Of CAUPOfcfltA. ;. SAN DIEGO EX BIBL10THECA ,. CAR. I. TAB ORI S. STANDARD ALPHABET. STANDARD ALPHABET FOR REDUCING UNWRITTEN LANGUAGES AND FOREIGN GRAPHIC SYSTEMS TO A m C. R. LEPSIUS, D. PH. & D.D. PROF. AT THE UNIVERSITY, AND MEMBER OF THE ROYAL ACADEMY, BERLIN. RECOMMENDED FOR ADOPTION BY THE CHURCH MISSIONARY SOCIETY. SECOND EDITION. LONDON. WlttAjAMS &NORGATE, HENRIETTA STREET, COVENT GARDEN. BERLIN, W. HERTZ, BEHRENSTRASSE 7. 1863. ADVERTISEMENT TO THE FIRST EDITION. THE need of a fixed system of orthography induced several of the missionary societies of London, a few years ago, to agree upon "Rules of reducing unwritten languages to alphabetical writing in Roman characters." These Rules, though imperfect, have been already applied with success to several African lan- guages. The societies were assisted in this work by the late Professor Lee of Cambridge, by Mr. Norris of London, and by Professor Lepsius of Berlin; but feeling it to be necessary for the establishment of any Standard system that an alphabet should be presented in a more complete form, and that the scientific principles should be explained upon which it was constructed, Professor Lepsius, at their request, kindly undertook this work, and has furnished the following admirable treatise, which will prove, it is hoped, an invaluable help to missionaries. The clear and scientific exhibition of vocal sounds which it contains will relieve Missionaries from many of their first difficulties in studying a foreign language, and will spare future translators much painful uncertainty respecting the powers of the letters which they employ. It has therefore been adopted by the Church Missionary Society as A STANDARD ALPHABET. It is not expected that those who have already adopted a different system of orthography will at once conform to all the a 1 ! IV ADVERTISEMENT. recommendations of Professor Lepsius, and in some minor points the genius of a language may possibly require a departure from the general standard. But such exceptions need not annul the standard character of this alphabet as one to which all parties may refer. Attention is particularly directed to the ob- servations of Professor Lepsius upon this point at page 23 (45). Founts of letters and matrixes are provided for printing ac- cording to this alphabet, and though its adoption may thus involve in the first instance some trouble and expense, these will be counterbalanced by the great and permanent advantage of a fixed orthography. The object of this treatise concerns not only missionaries, but also the interests of the natives whose language is to be reduced to writing. It is most desirable that a nation should be furnished with an alphabet combining simplicity and precision to the utmost degree in which they are attainable. The art of reading will be thus greatly facilitated, and the natives will themselves teach one another to read and write without the perpetual aid of European teachers. In illustration of this remark, we may refer to the following instances: In West Africa the Vei tribe invented a syllabic alphabet, in which every sign had its fixed sound, and the people taught one another to write without the aid of European teachers or the knowledge of European alphabets. Similar instances of natives teaching one another to read and write by a syllabic alphabet have occurred among the Indians in America. In New Zealand a very simple alphabet was carefully prepared by Professor Lee, and many tribes learned to read and write by the help of instructed natives before they were visited by Europeans. ADVERTISEMENT. V In respect of Africa it is especially important to take every step which may facilitate the mutual instruction, and supersede the labours of European teachers. In this way only can we hope for the Evangelization of that vast continent. It is a matter of much satisfaction , that in this, as in other instances, science lends its aid to the Christian zeal of mis- sionaries for communicating to mankind the highest benefits; and the work is commended under this aspect to the blessing of Almighty God for the furtherance of the Kingdom of Christ among: the nations of the earth. H. VENN, B.D. HON. SEC. J. CHAPMAN , B.D. SEC , Late Missionary in South India, and Prin- cipal of the Syrian College, Travancore ,,. . Missionary H. STRA1TH, HON. LAY SEC. C. GRAHAM, LAY SEC., Late Persian Interpreter to the Coni- niander-in-Chief in India. Church Society. HAVING been concerned in the preparation of the Rules, &c., referred to above, which have been successfully employed in our West African languages where the want of a uniform system was especially felt, we express our cordial approval of this treatise, in which Professor Lepsius clearly explains the scientific principles upon which a standard alphabet must be constructed, and renders it, in its complete form, capable of the most extensive application. JOHN BEECH AM, D.D. SEC. ] Wesleyan ELIJAH HOOLE, SEC., \Missionary Formerly Missionary in South India, j Society. a2 VIII ADVERTISEMENT. guistische Alphabet" niedergelegten Grundsatzen der Ortho- graphic und \vird demgemafs Hire Missionare fur deren lin- guistische Arbciten instruiren. Namens der Deputation: Insp. WALLMANN. [Barmen, 30 th July 1855. The Committee of the Mis- sionary Society of the Rhine declares hereby its assent to the principles of Orthography laid down by Professor Lep- sius in his treatise on the Standard Alphabet, and will give directions accordingly to its Missionaries for their linguistic labours. For the Committee: Inspector Wallmann.] Calw, Wiirtemberg, den 29. Oct. 1855. Der Unterzeichnete mill's, auch abgesehen von dem imisieh- tigen Fleii's, womit dieses Standard Alphabet eutvvorfen ist, schon um des harmonischen Zusammenvvirkens willen, dringend wiinschen, dafs wenigstens auf diesem Theile des Missionsge- bietes Uniformitat /u Stande komme, und schlieist sich daruin demselben mit Freuden an. Dr. CHR. G. BARTII, Yorstand des Calwer Verlags-Vereins. [Calw, Wiirtemberg, 29 th October 1855. The Undersigned, besides acknowledging the care and completeness of the views, upon which this Standard Alphabet is founded , cannot but earnestly desire for the sake of harmonious cooperation, that Uniformity may be attained at least in this part of the Mis- sionary field, and therefore begs to give it his cordial assent. Dr. Chr. G. Earth, Director of the Calw Publishing Union.] ADVERTISEMENT. IX Evangelische Missions-Gese 1 Ischaft zu Basel. Au-szug aus doin Protokoll vom 9. Nov. 1855. ,,Die Committee der evangelischeu Missions-Gesellschaft hat ,,in Anerkennung der grofsen Wichtigkeit ubereinstimmen- ,,der Grundsatze bei Feststellung des Alphabets bisher nicht 5,geschriebener Sprachen besonders auf dem Africanischen ,,Sprachgebiet beschlossen, das von Hrn. Prof. Dr. Lepsius in ,,Berlin aufgestellte System der Orthographic zu adoptiren und 5,den in ihrem Dienst stehenden Missionaren dasselbe zu all- w mahliger Einfiihrung zu empfehlen." Namens der Committee: JOSENHANS, Inspector. [Evangelical Missionary Society at Basle. Extract from protocol of 9. November 1855. r The Committee of the Evangelical Missionary Society, acknowledging the great im- portance of uniform principles in fixing the Alphabet of pre- viously unwritten languages, particularly among the African races, has resolved to adopt the system of orthography pro- posed by Prof Dr. Lepsius of Berlin, and to recommend it to the Missionaries employed by this Society for gradual in- troduction." For the Committee: Josenhans, Inspector.] American Board of Com missioners for Foreign Missions. Missionary House, Boston, June 20, 1856. The Secretaries of the American Board of Commissioners for Foreign Missions have examined Dr. Lepsius's "Standard X ADVERTISEMENT. Alphabet for reducing unwritten languages and foreign graphic systems to a uniform orthography in European letters," and regard it as an advance upon the practice of Missionaries of this Board heretofore in reducing languages to writing; and we will gladly do what we can to secure its general adoption. R.ANDERSON,) S. B. TREAT, [Secretaries. S. L. POMROY, 1 XI ADVERTISEMENT TO THE SECOND EDITION. A SECOND Edition of the English Version of the General Linguistic or Standard Alphabet is now offered to the public. It has undergone complete and careful revision; several minor points heretofore left undecided, have been settled by the light of experience and in accordance with the judgment of many distinguished scholars; and the Second Part, which exhibits synthetically its applicability to particular languages, has been much expanded and enriched. The Church Missionary Society adopted the Standard Al- phabet more than five years ago on conviction that it was theoretically the best. It appeared to be . founded on clear scientific principles consistently carried out, and therefore simple capable of easy comprehension and unlimited applica- tion. The interval has not disappointed their hopes. It has been applied successfully to at least fourteen African 1 and seven Asiatic 2 languages, and the test of experience enables the Society to commend it now, with far more confidence than when it first appeared , to all classes of students of languages, whether altogether unwritten , or rendered too often practically inaccessible by the cumbrous local alphabets in which their literature is embedded. 1 Aongla, Bornu, Ewe, Fula, Gd, Hausa, Herero, Ibo, Masai, Nama, ian, JYpe, Oji, Zulu. 2 Turkmenian, Kurd, Kafari, Pasto, Sindhi, Hakka- Chinese, Korean. XH ADVERTISEMENT. This Alphabet is available for two very different classes of readers the European student, and the uncritical Native. For the former, whether traveller, missionary, or philologist, the whole apparatus of diacritical marks will be needed, and more especially for such works as Grammars, Praxes, Dictionaries, &c., where the object is technical and linguistic. For the latter, many of the diacritical marks may be dispensed with, or will gradually drop off of themselves; and the Alphabet readily admits of such simplification, while preserving all its essential principles in their integrity. The Standard Alphabet may be regarded as already lord of the domain of Africa, where it has had to compete with but few conflicting systems of any importance, either European or native. In Asia, the difficulties are greater, but here too its advance is encouraging. To refer more especially to India, not only is the possibility and expedience of "Romanizing" still far from conceded by many oriental scholars, who are naturally attached to Alphabets which they have mastered at the cost of so much labour, but ever since the opening of that great country to Europeans, attempts have been made, with more or less of system, to represent Hindu words and names in Roman letters. There is however only one scheme that can profess to compete with that of Prof. Lepsius. Consider- able currency has been obtained in India for the system first propounded by Sir W. Jones, afterwards adopted with con- siderable modifications by Prof. Forbes and the late Prof. Wilson, and recently advocated with much earnestness by Sir C. Trevelyan and Prof. Monier Williams. The oriental student is much indebted to Sir C. Trevelyan for his weighty argu- ments in favour of the application of a Roman Alphabet to the languages of the East, which have never been met by a satisfactory answer; but the particular system, which he re- ADVERTISEMENT. XIII commends, is far from perfect. This is not the place to enter into the objections which present themselves, but the many changes which this alphabet has successively undergone at the hands of Prof. Wilson and others, and the further changes still suggested by different patrons of this alphabet, show a want of complete confidence in it. The system of Prof. Lepsius is the result of many years' patient study as well as exten- sive practical experience ; it is based on philological principles and a careful analysis of all the sounds possible to the organs of the human voice; full advantage has been taken of the researches of Sanskrit Grammarians; and, wherever available, the most widely received symbols have been adopted and incorporated into it. These are no slight claims to atten- tion and cannot be overlooked by candid scholars. It was under the conviction of the great services already per- formed by this Alphabet, and its capabilities for much further usefulness, that the Committee of the Church Missionary So- ciety commissioned the writer of these lines to visit Berlin last autumn, and came to the following resolutions in consequence: Extract from Proceedings of Committee. "The Rev. W. Knight reported his recent conference at Berlin, in company with the Rev. J. F. Schon, with Prof. Lepsius, respecting a final revision of the Standard Alphabet, and explained that the object had now been satisfactorily ac- complished by the preparation of a second Edition of the work, and that it was now desirable to take steps for the promotion of its general adoption by circulating a large Edition among the Society's and other Missionaries, and also amongst learned men likely to take an interest in the question, and by furnishing Founts of Types to such of the Mission Printing Presses as may be ready to employ them. XIV ADVERTISEMENT. Resolved 1. That an English Edition of 1500 copies of the Standard Alphabet be printed at Berlin under the di- rection of Prof. Lepsius, for circulation as suggested. 2. That the Secretary be authorized to urge upon the Missionaries of the Society the adoption of this system of orthography, and that as soon as there shall appear in any of the Society's Mission Presses in India or elsewhere a readiness to adopt and employ the system, the Committee will be ready to consider of the grant of a Fount of Types for such press." Church Missionary House, London. Aug. 26, 1861. W. KNIGHT, Sec. C. M. S. XV CONTENTS. Advertisements p. in Introduction to the Second Edition - 1 First Part. Historical and Systematical Development -21 The scientific Object of the Standard Alphabet -23 The practical Object of the Standard Alphabet - 26 What has been done by Science for the Solution of the Problem up to the present time - 30 What has been done by the Missionary Societies for the Solution of the Question - 39 The System proposed - 46 A. The System of Vowels - 46 B. The System of Consonants - 59 Second Part. Collection of Alphabets reduced to the Standard Alphabet - 85 General Division of languages - 87 LITERARY LANGUAGES P- 91-265 A. GENDER LANGUAGES - 91-208 I. Japhetic (Indogermanic) languages - 91 173 Sanskrit p. 91 Modern Persian p. 130 Pull - 96 Armenian - 132 Old Prakrit - 97 Kurd - 136 Hindi - 98 Ossetian - 138 Hindustani, - 100 Lituanian - 141 Sindhl - 103 Old Slovenian - 145 Gujaratl - 106 Serbian - 151 Marsthl - 108 Russian - 154 Panjabl or Sikh - 110 Cheskian (Bohemian) - 157 Nipali -Ill Polish -160 Bangali - 112 Serbian - 162 Ur'iya - 113 Rumanian (Walachian) - 164 Pasto or A/yan -114 Old Icelandic -169 Old Baktrian (Zend) - 1 1 7 Welsh (Kimri) -171 Old Persian (Cuneiform) .... - 124 XVI CONTENTS. II. Semitic languages p. 173 193 Hebrew p. 173 Getez (Ethiopic) p. 188 Syrian - 182 Amharic - 191 Arabic - 184 III. Hamitic languages p. 193208 Old Egyptian p. 193 To-maSey (Mrdiy) p. 205 Coptic ' - 200 Haftsa - 207 Beja (BiSari) - 202 Nama (Namayua) - 208 Qalla - 204 B. NO-GENDER LANGUAGES P. 209-300 I. Asiatic languages - 209 255 1. Talaric languages - 209 231 Manju p. 209 Madyaric (Hungarian) p. 220 Sarr a- Mongolian 212 Mordvinian - 221 Buryetic - 213 Livonian - 223 Yakutic .214 Tamil (Tamulian) - 224 Turkish - 215 Malayalam - 228 Turkmenian -218 Tidu -229 Kazak (Kirghiz) - 218 Karnataka (Kanarese) - 230 Samoyedic dialects - 219 Telugu (Teliiiga) - 231 2. Monosyllabic languages p. 232 243 !Kwan-hwa (Mandarinic) ... p. 232 Thai (Siamese) p. 237 Hok-lo - 235 Kamboja - 241 Hak-ka - 235 Mranma (Burmese) - 242 3. Isolated languages p. 244 255 Y'ukagiric p. 244 Georgian p. 251 Caucau (Chukchic) - 244 Tus - 253 Japanese - 245 Albanian - 254 Tibetan - 249 II. Polynesian or Malayan (Oceanic) languages .. p. 255 265 Malayan p. 255 New Zealand p. 264 Batak and Mandailih - 257 Raro-Tonga - 264 Javanese - 259 Gambier - 265 Dayak -260 Tahiti -265 Makassar - 261 Marquesas - 265 Bugis -263 Sandwich -265 Eastern Polynesian languages - 264 CONTENTS. XVII ILLITERATE LANGUAGES p- 266-300 III. Australian or Papuan languages p. 266 268 South Australian (Adelaide) . p. 266 Mare (Loyalty Islands) p. 267 Aimatom (New Hebrides; - 266 Viti (Fiji Islands). - 267 IV. African languages p. 268 288 tiuiiheli (Ki-SuaJieli) p. 268 Akra or Gd p. 280 Makua (Mozambique) - 269 Tyi'(Otyi) or Akwapim . . . . - 281 Tsuaiui (Se-t*it>-tna) - 270 Timne - 282 Kafir -271 Vei -283 Zulu - 271 Susu - 284 /tOsa (Ma-ffOsa) - 273 Mandihga - 285 Herero (0-Tyi-Iferero) - 273 Wolof - 285 Fernando Po - 275 Fula - 286 Ibo - 275 Kdnuri (Bornit) - 287 Yoruba - 276 Konyara (Dar Fur) - 288 Ewe - 279 Nuba - 288 V. American languages pag. 289 300 Indian langu. of North America p. 289 Tsalagi (Chiroki) p. 293 Greenlandic - 289 Dakota - 297 Massatsuset - 29 L Otomi - 298 Irokwois - 292 Ketstta - 298 Muskoki - 292 Kiriri - 299 Tsahto - 293 General Table of languages p. 301 308 Postscript - 309 315 Introduction to the Second Edition. FIVE years have elapsed since the first edition of the Standard Alphabet was published. It has during that period enjoyed a wide circulation principally owing to the recommendation of the Committee of the Church Missionary Society, and the progress it has made leaves no doubt on our mind that it will ere long be universally adopted in all Missionary literature. An intimate relation exists between linguistic science and Missionary labours. The latter, especially in new and hitherto unwritten languages, supply the former chiefly by means of Translations, Vocabularies, Grammars, and Specimens with rich , and in most cases the only, materials for further investi- gation and comparison. When we consider this close relation, we are led to expect that by degrees science also will employ our system more and more extensively, the fundamental prin- ciples of which have hitherto remained uncontraverted. This Alphabet has as yet been more frequently applied to African languages than to any others. The reason for this is obvious. No attempts, or but very insignificant ones, have been made to reduce them to writing: at all events none such as could have stood in the way of the general introduction A of an Alphabet otherwise acknowledged to be in every respect suitable. l 1 The following works on African languages, in which the Standard Al- phabet has been adopted, have come to the knowledge of the author: 1854. S. W. Kolle (Church Miss. Soc.): a) African Native Literature, or Proverbs, Tales, Fables, and Historical Fragments in the Kdnuri or Bornu Language, and a Kdnuri- English Vocabulary. London. Church Miss. House. b) Grammar of the Bornu or Kdnuri Language. London. Church Miss. House. 1856. J. B. Schlegel (North German Miss. Soc. at Bremen): Agbalg ke me devio osrQ hlele. Aong la- Primer , printed for the Ger- man and Foreign School-book-Society at Caltc. Stuttgart. Steinkopf. 1857. J. F. Schon (Ch. Miss. Soc.): a) Fdrawd letdftn mdgdna Hatisa ho Mdkoyi mdgdnan gaskia etc. [Haiisa Primer], Berlin. Print. Unger. b) The Gospel according to St. Matthew, translated into Hans a. Printed for the British and Foreign Bible Society. London. Watts. J. Erhardt (Ch. Miss. Soc.): Vocabulary of the Enguduk Iloigob, as spoken by the M as ai- tribes in East-Africa. Ludwigsburg (Wurteinberg). F. Riehm. H. Tindall (Wesl. Miss.): A Grammar of the Namaqua -H ottenlot Language. Cape-town. (Sold by Trubner.) J. C. Wallmann (Inspector of the Berlin Miss. Soc.): Die Formenlehre der Namaquasprache, ein Beitrag zur Siidafrihani- scken Linguiilik. Berlin. (Published by W. Hertz.) Hugo Hahn (Rhenish Miss. Soc. at Barmen): Grundzvge einer Grammatik des Merer 6 im icesllichen Africa, nebst einem Worlerbuche. Berlin. (Published by W. Hertz.) J. B. Schlegel (North German Miss. Soc. at Bremen): Schlussel zur Ewe-Sprache mil Worlersammlung , nebst einer Samtn- lung von Sprichtcortern und einigen Fabeln der Eingebornen. Stuttgart. Steinkopf. J. Zimmermann (Basel Miss. Soc.): \ a) Genesi alo Mose klenkleh tcolo le; ye Gd iciemo h mli. The first book of Moses in the Akra (Gd} Language. (London. Watts.) Par more difficult is the application of a new Orthography, even of the most perfect kind, to Asiatic languages, especially b) Daniel gbalo Is, ye Gd wiemo lg mli. The book of Daniel in the Akra (Gd) Language. London. Watts. c) Johane kg Juda toodsi Is, etc. The Epistles of John and Jude, and the Revelation of St. John the Divine, in the Akra (Ga) Lan- guage. (London. Watts.) 2d Ed. 1861. Basel. d) Spruchbuch des Calwer Verlagsvereins. Stuttgart. G. Christaller (Basel Miss. Soc.) : Kirchengebet und Katechismus (Anhang zu Earth's Biblischer Geschichte). Stuttgart. 1858. J. F. Schon (Ch. Miss. Soc.): Labari nagari kdmmdda anrubutasi dagd Lukas. The Gospel according to St. Luke, translated into Hausa. Printed for the British and Foreign Bible Society. London. Watts. A. Steinhauser (Basel Miss. Soc.): Kanemo- Wolo. Primer of the Ga Language. Stuttgart. Steinkopf. A. Steinhauser & J. Zimmerniann: . Gesangbuch, 168 Lieder in Akra. Stuttgart. J. Zimmermann (Basel Miss. Soc.): a) A Grammatical Sketch of the Akra or Ga Language, and some Speci- mens of it from the mouth of the natives. Stuttgart. Steinkopf. b) Bofoi le Asadsi etc. The Acts of the Apostles translated into the Akra Language. (London. Watts.) 2d Ed. 1861. Basel. J. B. Schlegel (North German Miss. Soc. at Bremen): a) Jesu Krislo etc. History of Jesus , from the raising of Lazarus from the death to the day of Pentecost, with the Epistles and the Reve- lation of St. John the Divine, in the We Language. Stuttgart. Steinkopf. b) Mawu-agbaba me nija veve tewe blaato vo eve. Dr. Barth's two times fifty two Bible Stories etc,, translated into the Ewe Lan- guage as spoken in Ahlg, Slave Coast, W.Afr. Stuttgart. Steinkopf. 1859. J. F. Schon (Ch. Miss. Soc.): Letafin Musa Nabiu. The second book of Moses called Exodus, trans- lated into Ha lisa. London. Printed for the British and Foreign Bible Society. J. C. Taylor (Native 'Clergy man, Ch. Miss. Soe.): Isuama-Ibo Katekism, translated from Dr. Watts' s first Catechism. (London). Watts. A2 to those which have possessed a settled native system of writ- ing, and which, through either their literary or their practical C. L. Reichardt (Ch. Miss. Soc.): a) Primer in the Fulah Language. Berlin. Utiger. h) Three original Fulah pieces in Arabic letters, in Latin transcription, and in English translation. Berlin. Unger. Lewis Grout (American Board): The Isizulu. A Grammar of the Zulu Language, accomp. with a his- torical introduction, also with an appendix. Natal. J C. Buchanan. Published by May & Davis, Pietermaritzburg; sold by Trubner. London. J. Ziminerinann (Basel Miss. Soc.): a) Jesaya gbalo le ye Gd wiemo la mli. The book of the Prophet Isaiah, in the Ahra Language. Printed for the British and Foreign Bible Society. Basel. C. Schultze. b) Bofo hrohkroh Paulo \e wolo ni enma eyamadse Romafoi le; yi' Gd wiemo Is mli. The Epistle of St. Paul the Apostle to the Romans, in the Akra (Ga) Language. (London. Watts.) c) /. Kor. II. Pelri. d) 8 Wandlabellen. G. Christaller (Basel Miss. Soc.): a) Olyi kefikaii nhoma. Otji Primer, together with a Collection of Scripture Passages. Basel. C. Schultze. b) Yen aicurade ne agyenkwa Yesu Krislo ho asgmpa no, wo Olyi kasa mu. The Gospel of our Lord and Saviour Jesus Christ, in the Olji Language, as spoken in Akuapem, Gold Coast, \V. Afr. (London). Watts. c) Asomafo no nneyge ho asgm a Luka kyergic mae, wo Otyi kasa mu. The Acts of the Apostles by St. Luke, in the Otji Language. Printed for the British and Foreign Bible Society. Basel. C. Schnltze. d) Krislofo asafo a WQWQ Akuapem Tyi-dyom-fihoma. Hymnbook for the Christian Church in the Akuapem country in the Otji Lan- guage. Stuttgart. Steinkopf. L. Grout (American Board): Imisebenzi Yabatunywa : i kumselwe ngabafandisi ba semerika ngokwa 'mazulu. (Acts of the Apostles translated by the Teachers of America among the Zulus.) Emsunduzi. J. Buchanan. H. Hahn and F. Rath (Rhen. Miss. Soc.): Omahungi oa embo ra Yehova omukuru mu Olyiherero. (Tales of the word of Jehova in the Herero Language.) Capetown. S.Solomon. importance, have long since afforded to Europeans occasion to express by Roman Letters the sounds which they contain. 1 1860. 1861. J. F. Schon (Ch. Miss. Soc.): Oku Ibo. Grammatical Elements of the Ibo Language. London. Watts. (1861.) J. Chr. Taylor (Native Clergyman, Ch. Miss. Soc.): a) Akukwo ekpere Isuama-Ibo. A Selection from the Book of Common Prayer, translated into Ibo. London. Watts. b) Isnama -I b o Primer , by the Rev. Sam. Crowther, revised and en- larged by the Rev. J. Chr. Taylor. London. Watts. c) Isuama-Ibo Sermon, preached at Trinity Church, Kissy road, Free- town, Febr. 17, 1859. London. Watts. Sam. Crowther (Native Clergyman, Ch. Miss. Soc.): Nupe Primer. London. Watts. G. Chris taller (Basel Miss. Soc.): a) Nhyontobea ama Tyi-dyom-nhoma ne Ga-lala-wolo no. Tune Book to the Otji and Akra Hymn-books. Compiled by G. Auer. Basel. C. Schultze. b) Seeks Wandlabellen. Basel. J. Ziminermann (Basel Miss. Soc.): a) Exodus, in the Akra Language. b) SddSi kpakpai edse. lg. The four Gospels in the Akra Language. K. Rudi (Basel Miss. Soc.): Biblia mli sadsi ha gbekebii fufobii. Bible Stories for little children in the Akra (Ga) Language. Basel. C. Schultze. R. Lepsius : Ingil Yesu el Messlhnilin Markosin fdyisin nagittd. The Gospel accord- ing to St. Mark, translated into the Nubian Language. Berlin. 1860. (W. Hertz.) These and some other works, which we shall mention hereafter, have ap- peared under the auspices of no less than nine Missionary Societies or kindred Institutions of England, Germany, and America; and several of them have been printed and published at the cost of the British and Foreign Bible Society. 1 With respect to the Asiatic languages the sanction of our system by the distinguished linguists of the Imper. Academy of Sciences at St. Petersburg, the natural centre for Asiatic Philology, is of decisive im- portance. We refer to the notice given in the Bullet, de la Classe Historico- philologique de 1'Acad. Imp. des sciences de St. Petersbourg, tome XIV. 1857. p. 238, Seance du 5. (17.) dec. 1856: M. Dorn en presentant le glos- saire Kourde de Mr. Lerch (Dialectes Kourmandji et Zata) pret a etre mis 6 Among these again the most difficult as well as the most im- portant are the languages of India. Experience has convinced sous presse, communique que lui et MM. Bothlingk, Kunik et Schief- ner te sont concertes atec HI. Lerch quant au choix de V alphabet le plus propre pour la publication des specimens linguistiques Kourdes. On s'est decide en favevr de I' Alphabet panlinguistique, propose par Mr. Lepsius , auquel ont concourru les linguistes de Berlin et de Londres. In consequence of this decision Mr. Peter Lerch has employed the Standard Alphabet in his work: "Forschungen iiber die Kurden und die Iranischen Nordchaldder". First part: Kurdische Texte mil Deutscher Uebersetiung. St. Petersburg. 1857. Second part: Kurdische Glossare, mil einer liter ar- hitlor. Einleitung. 1858. The alphabet has since been likewise adopted by A. Schiefner, Ueber die Sprache der Juhagiren (Bull, de 1'Acad. de St. Petersb. tome XVI. 1859. p. 242 ff.); by J. Wiedemann, Ueber die Livisc he Sprache und ihr Verhiiltnisn J der Esthnischcn (Bull. t. XVI. 1859. p. 193 ff.); by Ahlquist, Eine kurze Nachricht ilber das Wogu- litche (Bull. Hist, philol. torn. XVI, 1. 2. 1859. p. 25 ff.); by the same, Ueber das Mordvinische Verbum. Helsingforth. 1859; by Ilmiusky, Ueber die Sprache der Turhmenen 9 %> gy ( K > 9) ty, dy (t, d') tr, dz (c, /) ts, dz (t, d) Of these forms the first amplifications, ky , gy, happen to stand closely connected with their originals, the simple palatals, which can scarcely be pronounced without a distinctly audible y. Where then a simple basis is required , they need no special distinction, but may be denoted, like the simple palatals, by K and y. For the second and fourth amplifications the marks t, d' and t, d have already been proposed in the first edition. For the third however, which we thought we had before most suitably expressed by ty and dz , we now employ c and f for the reason which we have given. We write therefore the Turkish palatals K, g ky, gy. In the Sindhi alphabet there appears by the side of c, f, also a d' ( dy} , and in the Pashto by the side of c and f also t and d (= fe, dz) , which have also descended from palatal sounds. Another deviation from the first edition consists of the changing the aspirates (f, d, &c.) into their corresponding mutes with an h following (thy dh, &c.). That this change is not in contradiction to the fundamental laws of the Standard Alphabet is plain from the explanation of the aspirates we have already given ^l st Ed. p. 44). Indeed, the aspiration is not absolutely one with the consonant, but partly follows it, and can therefore be ex- pressed separately. Following the practice of Bopp and others, we previously marked the aspirates in the Sanskrit by an added Spiritus usper; we did so, because in the Devanagari the aspirates are represented by single signs. But it has also been mentioned that in the Arabic transcription of the Hin- dustani the Indian aspirates are already written separately. Still more decisive however is the practical ground that the notation of the aspirates as simple sounds, has not been 12 favourably received, on account of the inconvenience attached to it in writing and printing. We believe therefore that we shall, in this point also, meet the views of the majority of those interested, if we free the Standard Alphabet of the in- convenient hooks of the aspirates, and substitute in their places the full h; except only in the more accurate transcript of the ancient pronunciation of the Devanagari. Lastly, in the first edition of the Standard Alphabet, the choice was left open between % , #', and the new bases /, d. We have found with great satisfaction, that the most intel- ligent voices have been raised in favour of the latter signs. We therefore do not hesitate to drop the reluctantly admitted signs %, #', and to substitute for them ^, 6. Of minor im- portance is the adoption in the present edition of the angular sign *, z instead of the circular s, z, to denote the English sound sh and the French j. Whoever may prefer the latter form, may use it without hesitation. In favour of the former we find not only its use in several Slavonic languages, but also its greater clearness in writing and print, whilst the round form may easily be confounded with the Spiritus asper or lenis. We have already in the first edition protested against the use of Italics to denote the deviating classes of sounds according to which we should, for instance, have to write: "danda" instead of in conformity with our system: danda. It gives us much satisfaction to perceive that Professor Max Miiller has in his last publication , The History of Sanskrit Literature, relinquished this mode of transcription and substituted dots to express the Indian cerebrals. We must also repeat our de- cided protest against the use of ch and A, according to the English pronunciation, instead of our c and We rejoice at the accession to our view of Dr. C aid well, at least as far as the former sound is concerned, since in his last 13 work 1 he has given up the use of ch, which he had previously adopted. Our Alphabet now agrees in most points with his, as far as regards the pronunciation of the present Indian languages. Still we consider the diacritical sign " over c and j (c, /) to be unavoidably necessary in order to distinguish them everywhere from c and j as used in European languages. As ch instead of c", so also sh instead of s, violates not only the fundamental laws of all correct transcription, but also the requirements of practical applicability. The combination sh can, according to the analogy of kh, ph etc., only indicate an aspirated s, as it actually occurs in the Chinese and other languages, and may also occur in any other, by the concurrence of a final s with h following, as in English mishap, in German Grashalm. Neither can we approve of the stroke a, e, 6 in the Alphabet of Dr. Caldwell to denote long vowels, because this stroke is, as a general practice, never used in European languages to indicate the prosodic length of a vowel, but the accent of the word, as in the Greek. These and some other imperfections in the proposals of Sir Charles Trevelyan adopted by Professor Monier Will- iams 2 and the Rev. G. U. Pope 3 , should be avoided before 1 On ihe substitution of the Roman for the Indian Characters (com- municated to the Madras Literary Society by Sir Charles Trevelyan), 1859. 2 A new collection of various essays referring to this subject has just been made by Prof. Monier Williams, entitled: Original papers illustrating the history of the application of the Roman Alphabet to the languages of India. London 1859. 3 One Alphabet for all India. Madras 1859. This pamphlet, like that of Dr. Caldwell, was elicited by Sir Charles Trevelyan, when at Madras. While both essays contain strong arguments in favour of a Roman Alphabet for India, it would be incorrect to regard the authors , as advocat- ing, in its integrity, the modification of Sir W. Jones's system, adopted by Sir Charles, and promoted by Professor Williams. Amongst other divergences, for example, Mr. Pope employs the circumflex to mark the ong vowels (a, e, &c.), and the Spanish . We have noticed one of the 14 we can either wish or expect that his long continued and laudable exertions for the introduction of the Roman Alphabet will meet with complete success. Just as little can the report of the Sub-Committee of the Madras Literary Society, and auxiliary of the R. Asiatic Society , on writing Indian words in Roman Characters ( in the Madras Journal of Literature and Science, Vol. 3, New Series, Madras 1S59), exert any de- cisive influence on the solution of the main question; since Messrs. Elliott, Bayley and Norman, the gentlemen com- posing the Committee, are not even agreed among themselves, but have conveyed in three appendices their own individual proposals, which deviate in some, and partly not unimportant, points from the main report composed by Mr. Elliott. Messrs. Elliott and Bayley have even gone so far as to return -- in opposition however to Mr. Norman's protest to the ex- ploded method of denoting aspirates by k'h, g'h, ch'h, thus separating distinctly by means of a comma the mute from its aspirate, although the same is found to be so intimately con- nected that in the Devanagari it is expressed by but one letter. These writers can, in this, as in some other points, plead the example of Sir William Jones, whose great services we have always acknowledged (see below), especially with reference to the vowel-system. It has however been already pointed out, that his transcription of the consonants was very defective, principally because the physiological laws of the system of sounds were not then fully recognized. The Sub-Committee appear unfortunately to have been unacquainted most important of Dr. Caldwell's variations above-the abandonment of ch to express the first consonant of the palatal row. He has also mis- givings as to the acute accent for denoting the long vowels, and even says, "For cursive writing, I suspect, it will be found that the simple horizontal line the ordinary prosodial line of length is the easiest" (p. 27). Even Mr. Williams hesitates not only as to ch but as to sh, and appears to propose c with a diacritical mark, while he thinks that "a similar modifi- cation of sh might perhaps be introduced with advantage." (See Williams's Bdg-o-bahar, pp. xxvn-xux.) We welcome such approximations to our system. 15 with the proposals of the Standard Alphabet, which is especi- ally based upon a critical choice, according to internal laws, amongst the different systems of notation hitherto used , and has the object not only to harmonize the wants and customs of In- dian writers with European science, but also with the wants of Missionaries beyond India, embracing, as they do, all the rest of the world. They would otherwise have perhaps at- tempted to combine their Anglo-Indian standing -point with that of the Standard Alphabet. Whilst referring to this subject we venture to call attention to the introduction to the Zulu Grammar by the Rev. Lewis Grout, in which the Standard Alphabet is thoroughly re- viewed, both with regard to its intrinsic value, and its gene- ral applicability. Mr. Grout, Missionary of the American Board of Missions, has for some time (see p. 41) taken the lead in the diversified and fundamental discussions of the question of the Alphabet in American Missions, and especially at Natal. He therefore has a strong claim to be heard, and his example in applying the Standard Alphabet to his learned work on the Zulu language cannot fail to cause those who take a cordial interest in the question, to follow him with all the more confidence. A few more observations on different points in relation to transcription into Roman Characters may now follow. We form the Appellatives of the letters in various ways. In naming the explosive Consonants we put the Vowel after, as: ka, te, de, pe, qu; in naming the fricatives and liquids we put the Vowel before, as: es, ef, el, em, ei\ Others again we call, as did the Greeks, by special names: as in the English: aitch, double- you, wy, zed; the German jod, vau, zet, ypsilon; the French ach, ygrec. Lastly, we possess no common designation for sounds like s, z, % and others, or for such as do not exist in European languages; yet it is necessary to have such both in teaching, and in other cases. Under these circum- stances it seems most advisable to follow the example of the 16 Sanskrit and other syllabic languages, adopting a uniform nomenclature throughout, and pronouncing every letter with- out difference, with an a following, thus: ka, pa, la, ma, fa, tfa, za, va, ya, and so on. It is also necessary to come to a definite understanding, with regard to the Order of letters in Lexicons and similar works. It seems that a scientific arrangement can only be obtained, by keeping the Vowels and Consonants respectively by them- selves, and by arranging the latter according to the different classes of the organs, i. e. gutturals, palatals, (fee., or as as- pirates, explosives (fortes, lenes), nasals, &c. In the first Edition (p. 47) we have (where an organic arrangement was necessary) given the preference to the classes of exspiration, because in them especially the bases of similar sounds are most closely kept together. For European use we have, however, recommended the European order of letters, and now con- sider it well to recommend the same to a still greater extent. It may in fact be introduced into all lexical arrangements, as a scientific order is only needed when one has to speak in a grammar or elsewhere of a scientific classification of sounds. Lexicons in foreign characters will necessarily follow the foreign arrangement; but, applied to Roman characters, either principle of arrangement deduced from sound would uselessly separate all the homogeneous roots and hard and soft letters, which belong to one another. Classified according to the organs, n, n, n, n, and again s, s, s, .s would be widely separated from each other; classified according to ex- spiration, k would be separated from g, t from d, $ from c, from d, and so on. Within these classes of exspiration, it would also be necessary to place the bases with distin- guishing marks before the simple bases, e. g. t and t before t, n and n before , which seems not to be natural, and as to other letters such as c, j, s, ~, #, d, it would be altogether doubtful what place should be assigned to them. Besides this, any arrangement of the letters according to the organs 17 would present great difficulty to Europeans, who are ac- customed only to the Latin mode, and this difficulty would be vastly increased , when we come to the order of the letters not only at the commencement, but in the body of each word. For foreigners however, who will have under any circum- stances to relinquish their accustomed succession of letters, it is of little importance what new arrangement they may adopt, and a scientific one is of no advantage, where convenience and practical utility only are aimed at. The case would as- sume a different aspect, if the Alphabetical arrangements of European languages were as diversified as their orthographies. In this case a new and necessarily oryanical arrangement would be unavoidable. But inasmuch as all European nations use one and the same order of letters as handed down to them by the Romans, who received it from the Greeks, who again received it thousands of years ago from the Phoenicians, they possess also the right of communicating the historical arrange- ment, as well as the characters themselves, to foreign nations. To enter into detail. We shall give the precedence to let- ters without diacritical marks; the rest, when there are several of them , will be arranged according to the organs of speech. As far as the signs ' and >' are concerned, the simple curve ' is not likely to be of much lexicographical use. Should it however be called for, it also would be treated as the distin- guishing mark of a vowel , and would either be dealt with as not existing at all, or follow the simple vowel as 'a after a, 'e after e, and so on. For the Semitic sound ? two curves > have been chosen in order that for European languages, we might not be obliged to take notice of this sound at all , more especially at the beginning of words. If we were to take any letter with a diacritical mark, for in- stance q, some inconveniences would ensue. One would con- sent to write 'Abdullah, I All, but never Qabdallah, Qali; we shall therefore treat / like ' as not existing, and should B 18 take no notice of it in lexical arrangements, except where two words have no other mutual distinction, in which case we should place the word containing the * immediately after the one that does not contain it, or which contains only one curve '. Lastly, the four Greek marks may most easily be as- gociated with those Roman ones, whose fricatives they are, con- sequently d will come after rf, / after g, % after *> * after L In conformity with this plan the Alphabet with the principal diacri- tical signs employed in the different languages could be arranged as follows: a, fZ, a, a, 6, , , J , >*a; 6, #, &', b; c, < , r', <; d, fi, ?-> ^ ?'> A ' // Z, ?, Z, /; m, w'?,, ???, m; n, n, n, n, , n, w, w, w, n; o, o, o, o, 6, o, p, p, p, p, p, p; jt>, y, p, p>; ) to t or , and the character for the corresponding soft sound to z. For sh he proposes a lengthened or old-fashioned , viz. f, or an inverted j, f; while for i, , ( #, j, /), he retains the additions, although he changes their forms. The notations of the vowels also un- derwent changes. At the close, he makes an attempt to apply his system of notation to the Hebrew, and the first line of his Hebrew Lord's Prayer will give a good idea of the awkwardness of this third method of writing. It is the following (p. 209.): f' b e C"mim i q"dd"\ t e m-k". 1 Both transcriptions are placed by the side of the Arabic names in the Index (leoyraphic/ue , which forms vol. xvm. of Panckoucke's edition. '-' He discovers a difference between French ee or ez (don nee, donnez) and the simple e (arme, bonte), and finds the former again in the German eh (dehnen"), the latter in the German besser, or in the English red, head: s. p. 49 - 52. He pronounces the nasal in the German Anker as in the French Ancre, p. 59; the German z he resolves into ds, p. 83; and the Arabic yain he calls a grasseyement dur, in opposition to the grasseyement doux of the modem Greek y , p. 100. The German ch in ich he places as a soft sound by the hard sound in bucA, p. 103, etc., etc. No one of the three editions of Volney's system met with any approbation or adoption, because his proposal was ba- sed neither upon scientific nor upon practical principles, be- cause it embraced in its field of view only the Arabic alpha- bet, and because it admitted no direct application to other languages and especially not to those of India. His exertions, however, were not forgotten, as by his will he founded an annual prize to be conferred by the Institute of France. This legacy was designated : Pour le meilleur ou- rrtKje relatif a t etude philosophique des lanyues, and at the same time the wish was expressed: cPencourager tout travail tendant a donner suite et execution a ^lne methods de transcrire les langues Asiatiques en lettres Europeennes. This endowment, which was recognised by an Ordonuance of 1820, has produced many good results for the advancement of linguistic science, but it has conduced so little to the solution of the problem in question, that the French Academy finally determined to omit this sub- ject in their Programme, and only to propose exercises on comparative grammar. 1 The system of Sir W. Jones, which had 1 Compare Memoires de I'Jnstilut de France, Academic des Inscr. et Belles Lettres, toine xrv., Paris, 1845, p. 7 seqq. In the year 1835 a book appeared by A. E. Schleiermacher: De I' Influence de I'ecriture sur le language, Me- moire qui en 1828 a partatje le prix fonde par M. le comte de Volney, suiti de Grammaires Barmane et Malaie, et d'un aperqu de I'alphabet harmonique pour les langues Asiatiques qtie I'lnstilut Royal de France a couronne en 1827. The author gives in the preface p. ix seqq. a transcription of the Devanagari, the Bengali, and four Slavonic alphabets, with respect to an Alphabet har- monique, which he exhibits in the Aperqu mentioned on the title. But, as in neither place the reasons of this transcriptions are developed, and as the complete Memoir on the Alphabet harmonique has hitherto not been published, we must abstain from ofl'ering any opinion on it. The peculiar division, however, in 16 gutturales, 12 palatales, 15 sifflantes, 16 linguales, 9 labiates, 9 nasales, and 16 melees, and the 5 subdivisions oilellres simples, variees, fortes, mouillees , and aspirees, seem to indicate that the author Marts from a physiological and linguistic basis different from that which we consider the correct one. At the same time, however, the principle of using single signs for simple sounds is constantly observed. 37 proceeded upon more correct principles and upon a broader basis, was, indeed, occasionally alluded to by Volney, but never followed. No language has a system of sounds more rich and regularly developed than the Sanskrit, or expresses them so perfectly by its alphabets. The old grammarians of India did not, indeed? invent the Devanagari characters, but they brought them to the state of perfection which they now possess. With an acumen worthy of all admiration, with physiological and lin- guistic views more accurate than those of any other people, these grammarians penetrated so deeply into the relations of sounds in their own language, that we at this day may gain instruction from them, for the better understanding of the sounds of our own languages. On this account no language and no alphabet are better suited to serve, not indeed as an absolute rule, but as a starting-point for the construction of a universal linguistic alphabet, than that of ancient India. Hence it is that the late progress in the solution of the alphabet - problem has been associated in Europe, as formerly in India, with Sanskrit studies; especially since these studies were made the foundation of the new science of Comparative Philology. Here Bopp took the lead. In the earlier editions of his Sanskrit Grammar he had still employed the German compounds tsch, tschh, dsch, dschh, sch, ng, M, &c. ; but later, in his Comparative Grammar, published in 1833, he introduced single letters for all these sounds, and distinguished the va- rious classes of sounds by certain uniform diacritical marks. This orthography was soon adopted by the very numerous school of German and other linguists, and may now be regarded as the historical basis upon which, on account of its intrinsic value as well as its extensive use in science, the future superstructure must be built. H. Brockhaus 1 , Benary, Gorresio, Roth, Benfey, 1 We mention particularly his Essay Ueber den Druck Sanscritischer Werhe mil lateinischen Buckslaben, Leipzig 1841, in which he presents important considerations on the scientific advantage of printing large Sanskrit works in Latin letters. 38 Ktenzler, Lassen , Weber, and many others have adopted this principle, although, in particular instances, they have often differed among themselves as to the choice of the diacritical marks. But all these men had either the >SV//^s//vV language alone in view, or at most those of the same family. On the other hand, the Semitic scholars were equally exclu- sive, and generally retained the use of sh, kh, gh, th, i, the Polish 54 ?/, in our system? We have to answer it so much the more as this vowel appears also in many other eastern languages. The vowel u takes the middle between i and u. The phy- siological reason is that in forming the z, the lips are broad, the tongue slightly elevated and stretched out, whilst, in forming the u the lips are round and put forwards, the tongue drawn back in itself, so that in the forepart of the mouth a cavity is formed, the diameter of which is greater than its entrance and issue; which is the reason of the hollow sound of this vowel. In forming the w, the lips have the position of the u, the tongue the position of the i. So we are right in saying that u takes the middle between i and u. But it is on the other hand evident, that there must exist still another middle which has the same right to be counted as a peculiar sound. For we may form a vowel in such a manner, that the lips take the broad position of the z, and the tongue is withdrawn as in the u. This is the vowel which is called in the Slavonic languages the hard 2, the yeri BI of the Russians, which we write i. The origin however of this sound is, according to my opi- nion, not in the Slavonic, but in the Tataric languages, where we find it in the Turkish, Turkmenic, Yakutic, and other cognate languages. Here this vowel is an essential part of the so called harmonic vowel system , whose peculiar arrange- ment demanded it as a necessary complement 1 . Vestiges of 1 Without entering into the details of this question , we remark only, that the distinction of the lower and darker vowels a o M on the one side, and the sharper and clearer vowels e o w i on the other, exists in all lan- guages and manifests itself in different ways. The same opposition however is of much greater influence in those languages, where it forms, as in the Mantschu, Mongolian, Kalmuki, Turkish, Jakutic, Hungrian, Finnic and others, the basis of the "vocalic harmony." There the three "hard" vowels a o u, which are pronounced in withdrawing the tongue, correspond to the opposite "soft" vowels e o u, which are pronounced iu stretching out the tongue. The fourth "soft" vowel i would have 110 corresponding "hard" vowel, if it were not the vowel i, which is formed in drawing back the 55 it are even found in the Dravidic languages of India. Although there are intermediate vowels between a and f, as there are between a and M, a and i, a and u, all these languages have taken up only the one vowel ;/, because this is the most distinct vowel of the column and the only one they wanted for their parallelism of vowels. There is however at least one language known to me, which makes use also of a second vowel of the a i column, viz. the one which corresponds to o, as i cor- responds to u. The Ru manic or Wallachian language, as spoken north of the Danube, distinguishes a lower and a sharper vowel of the same kind, so as to till up the place left open for it in our system. We should like to keep for the sharper sound the Polish writing y. But this sign is already generally received for the semivowel of the palatals, and moreover would not be fit to form at the same time the basis for the deeper sound of the Rumanic language by taking a diacritical mark, because this mark ought to be added below where the space is occupied by the tail of the letter. No doubt, the basis of the sharper sound must be i, of the deeper sound e. They demand a common diacritical mark, different from those, which are al- ready in use for other purposes (i e e e). We regret not to find any symbol already adopted and therefore propose as a clear and convenient sign for handwriting to put the angle below i e. i- L, The vowel -pyramid of our system takes in consequence the following form, where the new vowels i e may be put with the same propriety on the e i side or on the o u side. e e o o or e o e i i u u t u i u Lips : broad broad round round broad round broad round Tongue -. forward back forward back forward forward back back tongue as far as the palatal point in the middle of the hard roof of the palate. Now there is a regular correspondence between a o u i and e o u i. 56 In most of the European languages the "soft" vowels of the middle have supplanted the "hard" ones. In the Rumanic language the contrary has taken place, and in the Slavonic also no o or u has been received, but only /, as the following comparison shows. German. Rumanic. Polish. Russian. a a a a e, e o o e e o e - o e,e - o i u u i i u i i u i i u In the languages however of the Turks and Jakuts the pa- rallelism of the "vowel -harmony" has called forth the two forms between i and w, viz. i and u. The deeper sound e has not been received; it would have been the most perfect cor- responding "hard" vowel to the "soft" one e. But it has been in this respect supplanted by its nearest neighbour , which in reality is neither "hard" nor ,,soft", or both together. The parallelism became (a) hard \aoui . -. P e o u i instead of soft: e o u t e g u i or according to our pyramidal arrangement the Turkish and Yakutic vowels are f "a a e [e] o o instead of e e o o i i u u i i u u It would be interesting to know, if there is no cognate lan- guage, where this last form of vowel harmony has been de- veloped. Similar, but not to be confounded, is another formation of vowels which one might call Gutturalisation. In no lan- guage, as far as 1 know, this formation has been independently organized as a peculiar part of the vowel system. But it has got a secondary influence in the Semitic languages, espe- cially in the Arabic. In forming the j, the middle tongue is lifted up to the palatal point in the middle of the hard roof of the palate; from this point forward it slopes down almost perpendicularly so as to leave a cavity between this point and the teeth. We may however also pronounce an i in lifting up the tongue only to the deeper guttural point near the velum palati, and in pressing down the whole middle part of the tongue so as to form a cavity between the concave and lengthened tongue and the roof of the mouth. This position of the tongue is very different from the shortened shape of it in forming the . The tone of the vowel becomes generally somewhat deeper then the tone of the ordinary i. The change of formation and sounding is less in the other vowels and almost none in uttering the o, be- cause also the ordinary pronunciation of this vowel is formed at the deeper guttural point. Again the vowel a, which is formed beyond the guttural point in the larynx itself, can not be pronounced purely with the same guttural motion of the tongue ; it approaches perceptibly in its pronunciation to the o. In the Arabic language this "gutturalisation" of the vowels is distinctly- heard after and in consequence of, certain con- sonants. The Arabic Orthoepists call this pronunciation the "thick" or "fat" one, the modern Grammarians use to call it the "emphatic" pronunciation. It is very naturally con- nected with the deep guttural consonants, which are with pre- dilection developed in all the Semitic languages, and besides with the four linguals, viz. % ? <}) X ? > c -(-) - - -> sometimes also with I and r; it is not connected, and could not be, with 5 and A, which are formed behind the deep guttural point in the larynx, nor with &, which, as well as K s y and all the rest, is pronounced before the same deep guttural point. But it strikes us as a peculiarity of the Semitic languages, that this gutturalisation of the vowels takes place after the four letters d(t) 3 z, which are called by the Arabs, for this very reason, the "closed" letters, meaning the cavity-letters, and E 58 by the modern Linguists conventionally, though very improperly, the "lingual" letters. They are principally distinguished from the dental letters d d s z by the circumstance that in pro- nouncing them with the forepart of the tongue, the throat is contemporaneously narrowed at the deep guttural point, as if one of these gutturals was to be uttered. This movement of the throat not only causes a somewhat different position of the forepart of the tongue from the dental position, but it imparts to them also the thick emphatic sound, which they transmit to the following vowels in gutturalising them. In the same way also I and r may be pronounced (as well as , although it is not mentioned), if they are preceded by a lin- gual consonant without an intervening vowel. In the word w sJJ! Allah the I after a and u is always and by law of the orthoepists pronounced emphatically. It is in this case no other letter then the Slavonic "hard" I, the Polish }, which in consequence is to be written in our system I. It must be granted that the "thick" pronunciation of the vowels in the Semitic languages proceeds really from the con- sonants and has therefore no linguistic value in itself, because this vocalic tone appears exclusively after the said consonants which, on their part, keep their peculiar pronunciation even if they close a word or are followed immediately by another consonant. It is evident therefore that we have also in our transcription, as in the indigenous writing, not to express the gutturalisation, either in the vowels, or in the guttural con- sonants, but only in the lingual consonants d(t) d s z and in the Slavonic _Z.] Finally, the clear vowels are further capable of a peculiar alteration, that of nasalisation. This is produced not by closing nor even by narrowing the canal of the mouth , but by simul- taneously opening the canal of the nose. There is no conso- nantal element brought into play (although the nasalisation is mostly caused by the dropping of a nasal consonant), but it 59 is an alteration entirely within the vowel. As such it has been rightly understood by the Indian Grammarians, who express the nasalisation (anusvara) by a vowel-like sign, namely, by placing a dot over the letter. For the European alphabet, we choose the*ign ~ placed over the vowel 1 , as the dot would be inconvenient in the case of the , and write , and (/ , cannot be accurately de- fined, but was probably contemporaneous with the alteration of y and J, whilst fi seems to have approached latin r> in still earlier times. 3 Instances of this tendency are generally known from the Romanic lan- guages. See also below, where the Palatals are considered. 67 lately required. There are , indeed , eight bases for the above stated nine explosive sounds, and only six for the twelve fri- cative sounds. An augmentation of the latter by introducing the Greek signs ^, /, & and J, is consequently almost unavoi- dable ; and their absolute necessity will soon be still more evident when we come to consider the Asiatic sounds in ad- dition to the European. We are thus enabled to give the following tableau of the European sounds: Alphabet of the European Consonantal System. explosive or dividual. fortis. Icnis. nasalis. k g n fricatives or continues. fortis. lenis. semivoe. x h y y 8 Z t d n 8 Z f\ Jfr p b m \r / w ancipites. r r I G u 1 1 u r a 1 e s Dentales Labiales Enlargement of the Alphabet by the Addition of the Foreign Sounds of Oriental Languages. The Asiatic languages, especially the Indian and the Arabic, possess, besides the sounds hitherto considered, others, which hardly exist at all in European languages, or at least are only fully developed in Asiatic languages, and, therefore, can only find their proper position in a more comprehensive system. In- stead of the three European classes, we must distinguish seven, which we shall now consider separately. I. THE FAUCAL CLASS. h We are accustomed to reckon h among the gutturals. It is 68 easily observed, however, that we pronounce this sound behind the guttural point, immediately at the larynx. When pronoun- ced so softly as to be vocalised, i. e. so as to imply a vowel sound produced in the larynx (as with , v, $, 2) the friction ceases to be audible, and only the vowel element is heard. This vocalised consonantal breathing, is, therefore, not pecu- liarly marked in any language, h belongs, therefore, to the unvocalised strong fricatives. Arabic 5, Hebrew N, Sanskrit ^T, Greek spiritus lenis. 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 the Italian sard 'a casa, the English go 'over, the German See- 'adler,; or even after consonants when trying to distinguish, in German, mein 'Eid (my oath) from Meineid (perjury), or Fisch-'art (fish species) from Fischart (a name), &c. We indi- cate this sound, when necessary, by the mark ', like the Greeks. Arabic , lain. The slight sound just described can be pronounced hard by a stronger explosion almost at the same point of the throat. Thus arises the sound which the Arabs write g. 1 We find it expressed by scholars generally by placing a diacritical sign over the fol- lowing vowel, a, , a, , a; sometimes below, a. This method would suppose , from the analogy of all systems of writing, that the ^ were only an indication of a change in the vowel. It is, however, a full consonant, preceding the vowel. We indicate 1 See on the accurate pronunciation of this and the other Arabic sounds my dissertation: Ueber die Aussprache und die Umschrift der Arabischen Laute, in the Transact, of the R. Acad. of Berlin. 1861. 69 it, therefore, with regard to its affinity to the soft sound, by doubling the spiritus lenis, '. Arabic f, Ka. The fricative sound corresponding to ' is not the common A, but a stronger aspiration, which requires a greater contraction of the faucal point, and is distinguished by the Arabs from the simple h. It has, therefore, been often indicated by hh. We write K and have a precedent in the writings of Fleischer (1831), Ewald (1831), Vullers (1841). The absence of any nasal sound in the faucal series is ne- cessitated by the physiological position of the faucal point, the contraction of which closes at the same time the canal of the nose. The faucal series is confined, therefore, to the four following sounds, thus represented: > ' K h. II. THE GUTTURAL CLASS. As we have already excluded the h from this class, on account of its being pronounced behind the proper guttural point, we must, to be accurate, exclude the y also, and put it in the next following class, this sound being formed in the mouth be/ore the guttural point. Again we are obliged to comprise a sound peculiar to the Semitic languages, viz. The Arabic o and Hebrew p, qaf or qof, which is formed at the posterior soft part of the palate, although this class has its place of formation a little more forward, at the point where the velum palati, joins the hard palate. We indicate this sound by the sign which the Greeks and Romans substituted for it, although it cannot be proved that they pro- nounced it exactly in the same manner, viz. q. We obtain by this addition the following complete guttural series: k q y, n; % /; r. 70 III. THE PALATAL CLASS. In passing from the guttural to the dental point, another point may be distinguished, and has been fixed by several languages, namely, the palatal point which is situated almost in the middle or on the highest point of the hard palate, and occasionally extends to the gum of the upper teeth. We dis- tinguish this class of letters from the Gutturals by a stroke put over them. A K or K h II. Gutturales. k 9 9 n x y r III. Palatales. k g n X, S, s /, z, z >j I' IV. Cerebrates. (Indicae) V. Linguales. (Arabicae) VI. Dentales. t d n d n * ?, s, 9 z, d r I r I VII. Labiates. P b m f v w Examples of Pronunciation. We follow here the vertical and not the horizontal order, because we thereby keep together all the letters, which in the different classes have the same bases. VOWELS. a engl. father, fr. dme. d ger. Mann, Hal. ballo. e fr. mere, ger. Bar. / engl. head, ger. fett. e engl. cane, vein, fr. donne. I engl. se#, fr. lit. i engl. ,M'W, fr. fil. Q engl. aW, Hal. pero. Q engl. Ao^, not. o engl. wo, fr. faux. u engl. rw/^, fr. nous. u engl. /oo^, fr. ours. o fr. beurre, coeur. engl. cunning, but. Q ger. Konig , fr. / arab. p ('din). k engl. coo, fr. cause. K old sanskr. ^. [see above p. 72.] c modern sanskr. ^, engl. ch. t sanskr. Z. engl. town, fr. faw. p engl. pine, fr. pew. 6. Lenes. arab. 5, hebr. N, gr. spir. len. '. q arab. o f^a/)- ^ engl. gold, fr. gauche. g old sanskr. ^ / modern sanskr. 5f, engl. _;'. (/ sanskr. ^. d(t) arab. -b (see below). d engl. cfear. & engl. by. c. Nasal es. n engl. singing, ger. 0/^0. ?i sanskr. "5f, ital. gnudo, fr. w sanskr. TJT. w engl. no. m engl. me. B. FRiCAxrvAE. . Fortes. /t arab. A'o). A engl. hand. % ger. .BwcA, ac^; pol. chata. s engl. s/iOt, fr. chat, ger. schon. % ger. wA, '/ve/i^. / old sanskr. ^. / mod. ind. If, pol. s'W^. s arab. o> (^ac?). s engl. sense ) fr. savoir. 6 engl. ^m, mod. gr. #eog. / engl. fine. b. Lenes. y arab. (ya'in). z fr. jeune, pol. bazant. y mod. gr. yscfvQcc. z pol. pozno. z fr. ,z?<3, engl. ^a/ 1 . (? engl. fo/, mod. gr. diipa. d arab. J=> (^aj. 2 arab. o^ f^aj (see below). c. Semivocales. T/ engl. year, fr. Bayonne, ger. /a. w engl. we. C. LlQCIDAB. r germ, and fr. dialects. r sanskr. T. r engl. very, ital. rabbia. I' ital. ^^', fr. mouille. I sanskr. 35. I engl. fow. 78 C>n '//> J.>y>'/v///'.y and Consonantal Diphthdngs. . !>/./'/to*irt' sounds which arc pronounced with njilc l>ut audilile breath. This class has been most fully Irvi'lopcd in the Sanskrit, where the fortes as well as the lencs of all classes can be aspirated in this manner. In the ancient Greek only the forte* admitted of the aspiration, and these afterwards passed into the corresponding fricatives. The aspi- ration can only follow the explosion, not accompany it through- out, as it does the friction of the fricatives. Thus, a real composition takes place. 1 If, notwithstanding this, the aspirates are represented in the Sanskrit as single letters, this is to be explained by the circumstance, that the spiritus unites itself more closely with the explosive letters than any other conso- nant, and is of so little weight, that it does not make the pre- ceding syllable long, being, properly speaking, no more than an increase of the breath necessarily inherent in every conso- nant. It is optional, therefore, either to regard the aspirates as single consonants, or as compositions with h. [We prefer now the latter. See above p. 11.] In regard to the doubling of consonants, it will readily be granted, that they ought not to be employed merely to show that the preceding vowel is short and accentuated, but only where the duplication (from the prolongation of the friction or of the moment of touching) is distinctly heard, or the double letter is justified etymologically, as originating in the assimilation of different consonants, or wherever nothing is intended, but a transcription of a foreign orthography, which makes use of double letters. 3 1 The best linguistic proof is, that no aspirate can be doubled; when a duplication is intended, the unaspirated sound is placed before the aspirate. From aka arises by reduplication not altlia, but akka. In Greek you write for the same reason iitlhj, Jit'c^os, Zi'tn<((o. - Every double consonant is pronounced with but one closing or narrowing of the organ and with the intention to unite the first half of the double consonant to the preceding syllable, the second to the following. 79 On the Application of the General Alphabet to the Alpliabet of particular Languages. It has been remarked above, that the general alphabet, when applied to particular languages, must be capable of simplifi- cation as well as of enlargement. All particular diacritical marks are unnecessary in those languages where none of the bases have a double value. "We then write the simple base without a diacritical mark e, o, s. Where two sounds belong to the same base, one only of the signs will often be wanted, espe- cially in the case of long and short vowels. If further essential differences should be shown, which are not yet represented in the general alphabet, and cannot be ex- pressed by a modification of the bases already adopted, nothing prevents the selection, or, if necessary, invention of other new diacritical signs, without deviating from the principles above developed. Among these latter cases we may reckon, for instance, the clicks of the southernmost African languages, which are formed, not by throwing out the breath, but by drawing it inward. We often produce the same clicks by the same movements of the tongue, but do not use them as articulate elements of speech. In the Hottentot language there are four clicks; in the Zulu and some other neighbouring languages to which they were transferred, only three. The first, which had been written hitherto g, is made by pressing the tip of the tongue closely upon the middle palate and withdrawing it suddenly, and from the place of its formation is to be reckoned among the cerebrals. The second (found prin- cipally in the Hottentot, but, according to Boyce 1 , also in some words of the Kafir language), arises, from placing the breadth 1 Grammar of the Kaffir Language, p. 4. He writes it qc. I myself have heard it pronounced by Zulu Kaffirs. 80 of the tongue in the palatal position, and withdrawing it with a suction. The third, generally written c, is in the same manner dental, as only the tip of the tongue smacks against the upper teeth and the gum above. The fourth is formed at the side of the tongue, by drawing in the air towards the middle of the mouth from the right or left side. It has been called lateral, therefore, and generally rendered by x. The pronunciation of these sounds becomes difficult only when they are connected with other sounds. Whilst the an- terior part of the tongue is smacking, the throat can open it- self for a g or n, so that these latter sounds are pronounced almost at the same time with the click, or immediately after it. 1 At the same time, the choice of c, connected with clicks by the Zulu Kafirs just mentioned, who in the be- ginning of 1854, sojourned for some time in Berlin. 2 We cannot, therefore, assent to Grout, who, instead of the former no- tation proposes the following : Grout, in the above-mentioned work p. 34. accepts our mode of writing the clicks, but places the three sounds which appear in connection with 81 sound, \ie may make them simply to follow, as with the diph- thongs, Thus we get the tableau: Palatals (qc) i Cerebrals (q) i ig in Dentals (c) i ig in Laterals (V) // ug tin them, not after but before the click - letters and writes ni , gii , ngj etc. and the Rev. J. L. Dohne, Missionary to the American Board, C. J. M. in his Zulu Kafir Dictionary, Cape Town, 1857, p. xxxvm, expresses himself distinctly against our proposal to write the clicks before their accompanying letters. On the other hand Wallmann and Vollmer have put the clicks after these letters, in the before-mentioned works. All are agreed, we among the rest, that the two sounds, although perfectly different, are pronounced al- most simultaneously ; and Dohne states , even in reference to the Kama and Kafir: "In the former the guttural sound begins after the tongue has clicked and continues with a peculiar and distinct force ; but this is little observable in the Kafir." I believe that I have remarked the same thing, with regard to the g, in the pronunciation of the Kafir. As far as I am aware, no one asserts that g is heard before the click. And neither does the etymological reason advanced by Dohne "that it was impossible that from the in before the root iela iihelo but (only) ihielo can be made " prove anything in favour of pronouncing ih or /, for equally little should we be justified in inferring from the fact that in the Sanskrit %f^ let'i "he licks 1 ' from f^fjjf lih and fTf ti, *' must be pronounced not as th, but as ht. It appears to us of little real importance whether one writes ig or gl , as both sounds are uttered,* as nearly as possible, simultaneously. It is however very desirable, that a majority should declare itself in favour of one or the other form, amongst linguists and Missionaries interested in the subject, to which majority the dissentient will then join themselves. Dr. Bleek (The Library of Sir George Grey, vol. I, pp. 6 and 172) men- tions a conference held in 1856 by the Rhenish Missionaries in South Africa, by which 2 of our 4 click signs, viz. / and // were adopted, and the other two, viz. / and / were exchanged for -}- and . There would be no great objection to the alterations , should a majority declare distinctly in favour of them. Meanwhile, nothing advocating the Rhenish mode of writing, ex- cept Vollmer's book, has as yet come under my notice, while our signs have already been made use of by Wallmann and Grout. Here also it ap- pears more essential that a choice should be made, than what the choice should be. The distinctive marks / and / are made according to the organic 82 The difficulty of transcription is greatest in those systems of writing which, originating in an earlier period of the language, and fully developed, have been retained unaltered, whilst the pronunciation has undergone a change, as also in those in which several reformations have left their traces. An instance of this kind has already been mentioned in speaking of the Sanskrit palatals. The differences of European orthography have mostly arisen from similar circumstances. Some such difficulties, how- ever, are presented by almost all existing alphabets which are not of modern formation. As the object of a standard tran- scription is to avoid , as much as possible , all such incongruity of sound and sign, no other course remains open in such cases than to fix upon a distinct period of the language in question, and to adapt its transcription to the different purposes of ren- dering, either the actual pronunciation, or the ancient one which had been expressed by the alphabet, and which may be deduced from it by linguistic researches. The difference is generally found to be greater in the vowels than in the consonants, the former being, in all languages, the more changeable element. The Arabs write only three vowels, but pronounce these three letters differently in different localities, according to distinct rules: in like manner, a certain number of consonants have a different pronunciation in different dialects, although in litera- ture they are expressed by means of one and the same written letter. Eli Smith and Robinson (in his work on Palestine) propose to represent the actual pronunciation in the country, classes of the cerebral and palatal clicks, and if + may perhaps appear more convenient than /', we may yet venture to say that -f resembles too closely the letter t. On the other hand, we must appear ourselves as decidedly opposed to the use of ng instead of n; and the more so, because in these languages both n and g are capable of being joined to clicks, and the reader is there- fore led to believe, that ng before a click must be either a union of n and g, or of n and g, which last frequently occur in connection in these languages. 83 and their endeavours are to be highly prized 1 ; but the linguistic scholar will prefer to follow the written system fixed by litera- ture, and to neglect the varying deviations and shades of modern pronunciation. Great difficulties are met with in transcribing the Rebreiv system of punctation, which, having only in after times been grafted upon the alphabet inherited from former ages, appears to be inconsistent with itself. In conclusion, we present the reader with a number of alpha- bets transcribed after our own system. We are aware that in many instances further researches must correct and complete our labours. We have followed the best and latest investiga- tions to which we had access in each individual language. The attempt is intended to show the easy applicability of our alpha- bet to the most different languages; and to induce scholars to follow in the same way, and eventually to correct and improve the details. 1 Compare also the excellent essay of Lane on the modern pronunciation of the Arabic vowels, inserted in the publications of the German Oriental Society. SECOND PAKT. COLLECTION OF ALPHABETS REDUCED TO THE STANDARD ALPHABET. GENEBAL DIVISION OF LANGUAGES. LITERARY LANGUAGES. A . Gender - langn ages . I. Japhetic (Indogermanic). II. Semitic. III. Hamitic. B. No -gender languages. I. Asiatic. I. Turanic or Tataric. II. Monosyllabic. III. Isolated. II. Polynesian or Malayan (Oceanic). ILLITERATE LANGUAGES. III. Australian or Papuan. IV. African. I. Primitive or South African. II. Isolated or Middle African. V. American. 89 OUR first division is in Literary and Illiterate languages. We call those languages literary, which for the most part have a system of writing and at least a beginning of literature. The illiterate languages have with few exceptions no writing. This makes of course a great difference with respect to the intro- duction of the Roman alphabet. It is far easier to introduce it among the latter nations than the former, where it has to over- come an indigenous alphabet with its characteristic features and historical claims, which it must respect even when not quite adequate to the physiological import of the respective letters. The illiterate languages offer only the difficulty of determining the true pronunciation of every sound without the important guide of an indigenous alphabet fixed by the speak- ing people themselves. The sounds once being known, the signs are easily applied. This is the reason why our explanatory remarks are more numerous in the first than in the second part. This division referring to the knowledge of writing is at the same time, generally speaking, a geographical one, since the European, Asiatic and in a great measure the Polynesian languages are literary , the Australian, African and American languages illiterate. We combine with this first division a second, referring to the use of grammatical gender. It is not accidental but very significant, that, as far as 1 know without any essential ex- ception, only the most highly civilised races the leading nations in the history of mankind distinguish throughout the genders, and that the Gender-languages are the same as those, which scientifically by linguistic reasons may be proved as descending from one original Asiatic stock. The development of peculiar forms for the grammatical genders proves a comparatively higher consciousness of the two sexes; and the distinction not only of the masculine and feminine, as in the Semitic and Hamitic G 90 languages, but also of the feminine and neuter gender, exclusively expressed in the Japhetic branch, is only a further step in the same direction. The formation of genders has appeared to me so characteristic of the three principal branches, that I thought it (1844) a sufficient reason, to ascribe all the African nonsernitic languages, which distinguish the genders, to the Hamitic branch, viz. besides the old Egyptian and the Coptic the Beja language of the Bishari (whose ancestors were the Ethiopians ofMeroe), the Dankali, Somali, Galla and other neighbouring languages, al those of the Libyan tribes between the Egyptian Oases and the Canarian Islands, including the Hausa farther on to the south, and even the widely distant languages of the miserably reduced Hottentots and Bushmen, whose immi- gration into their actual seats is still a curious problem, con- sidering the absolute diversity of their language from all their northern neighbours and at the same time its traces of a certain affinity with the Egyptian language. If we are not yet able to prove the affinity also of all no- gender languages to the former and to one another, although their original relationship is inseparable from the propagation of the one human race, it would certainly be too hasty an as- sertion, to say that we never should be able to do so. It seems however unquestionable, that the three great branches of gender- languages were not only in the past the depositaries and the organs of the historical progress of human civilisation, but that to them, and particularly to the youngest branch of them, the Japhetic, belong also the future hopes of the world. All the other languages are in decline and seem to have henceforth but a local existence. The geographical division seems therefore the most appropriate for them, and we prefer it for our purpose to the other, which might be based upon the different forma- tions and features of language. 91 LITEEAEY LANGUAGES, GENDER LANGUAGES, SANSKRIT. in r etc. w 5; ^ %r %r w T 35 The Virama v, indicates that no vowel is pronounced. a u i I u u r f I I d I etc. ai di au du ar dr a a i I u u r f I I d Z etc. e ai d au ar dr Ancient pronunciation. k y n K y n t d n t d n p b m Modern pronunciation. k y n c 3 n t d n t d n p b m h * - K tf *(j() y E / s r I f d s i f d* X V p B h x kh yh t y ch jh s r I th dh s I th dh x V ph bh s G2 JAPHETIC LANGUAGES. Specimen. II ^ II i (Beginning of the Rigveda ed. Aufrecht.) Old pronunciation. Aynim ilai purduhitd, yaghdsya daivdm rtvigd, hdutard ratnaddtamd. aynis purvai&ir rsiBii' idyau nutandir utd^ sd daiva dihd vaksati. agnind rayim asnavat, pdusam aivd divai-divai, yasdsd virdvattamd. agnail ya yagndm advard visvdtas pariBur d#i, sd id daivdisu gaftati. agnir hduta kavikratus satyds kitrdsravastamas daivdu daivdiBir a gamut. ydd angd ddsusai tvdm, dgnai, badrd karisydsi, tdvdit tat satydm, ahgiras. Modern pronunciation. Agnim ilepurohitd, yajhasya deram rtvijd, hotdrd ratnadhdtamd. agnis purvebhir rsibhir idyo nutanair uta, sa devd eha vaksati. agnind rayim asnavat, posam cca dive-dive, i/f/xasd vlravattamd. agne! yd yajnam adhvard visvatas paribhur asi, sa id dHve$u gachati. SANSKRIT. 93 agnir hotd kavikratus satyas fttrasravastamas devo devebhir a gamat. yad anga ddtuse tvam, agne, bhadrd karisyasi, tavet tat salyam, angiras. Remarks. We distinguish an ancient and a modern pronunciation of Sanskrit. Just as the Romanic nations pronounce the old Ro- man alphabet in a different way from the old Romans them- selves, and the modern Greeks the old Greek alphabet dif- ferently from the ancient Greeks, by adapting to the written ancient language the gradually changed pronunciation of the living modern language : likewise the Brahmans of to day do not pronounce the Sanskrit in the same way as the old Brahmans of that time when the Devanagarl writing was settled , but ac- cording to the sounds of the now living Indian languages. The linguistic rules of Panini and his scholars are only adapted to the old pronunciation, which happily we are able, in following the instruction of the old Grammarians, to determine better than that of any other ancient language. A real intelligence of this language and its harmonic organism of sounds is not possible without knowing the true ancient pronunciation, and considering the eminent importance of the Sanskrit for the comparison of languages, it seems indispensable for scientific linguistic purposes to approach also in transcribing the Deva- nagari as near as possible to the ancient pronunciation. The euphonic rules respecting the letters ^T W 1 ^TT etc. become absurd, if we suppose for them the modern pronunciation c j e ar etc. instead of K g ai ar etc. The case, however, is different, when the transcription aims at more practical purposes and must therefore have regard to the actual pronunciation of the Indians. With this view we have added the second scheme. The ancient pronunciation of the Devanagari letters has been 94 JAPHETIC LANGUAGES. discussed elsewhere by the author l , and his views seem to have been appreciated for the most part. Some of them have been mentioned above. We repeat here the results in a few words. ^5 / and "*/are simple vowels and can therefore not be expressed by ri and li: their value is that of a cerebral / or / vocalized by the inherent sound of e. The Anusvara, which enters instead of a nasal dropped at the end of words in pausa or before other con- sonants is a vocalic change of nazalization, and is to be indicated as such by a diacritical sign over the vowel, not by adding any consonantal letter. 1 "^h" ^f ^?TF are diphthongs. The Visarya belongs not to one but to all local classes of consonants ; it would therefore lead to mistakes, if we were to take h for its basis in our transcription. It was so weak a sound, that in the Devana- gari it was not represented by a full consonantic letter, but by two dots (). Whe should retain the same indication if it had not already another European signification. A slight modi- fication ( ) may suit our purpose. The sounds called Gihcamuliya and Upadmdnlya would correspond to a very week % and /, but, as they were of so fugitive and variable a nature, that in the De- vanagari they were, like the Visarga, only indicated, not sub- stantially written by full letters, it seems advisable not to go farther in our transcription; we keep the Devanugari indication by ?. With regard to the palatals ^ WT K y n , we have spoken above. The palatal fricative H" has conserved in some regions its original sound , but its transition into the actual sound * seems to have soon taken place; we add the sign jjr only in brackets. The letters s and r better keep in linguistic works the cerebral point, although there is no dental s and / in the Devanagari. The solution of Af, (j etc. into M, yh etc. is against the appre- hension of the ancient grammarians , who treat those letters as simple ones. The modern pronunciation has not abandoned the simpleness of the vowels / and /. The diphthongs T[ and ^h" are turned 1 f'alaographie ah Mittel fiir die Sprachforsclwng , zundchst am Sanskrit nachgeuneen. Berlin. 1834. SANSKRIT. 95 into simple vowels e and 6. The palatals ^ and ^ have been re- solved into the compound sounds ts and dz, which, considering their etymology, we write c r and ; (see the Introduction p. 9). If and *T sorrespond actually to the Polish letters s and sz; we write them accordingly s and s (see p. 71). "?Chas lost its cerebral nature, and we write it r without the dot so much the more as in other modern Indian languages there has been introduced a real ce- rebral r by the side of r, this latter being still written with the Devanagari sign T- With respect to the Aspirates, we follow the Hindustani writing, which resolves them into kh, ph, etc. We have already mentioned, that this solution into two letters is not against our physiological principles (see Introd. p. 11). We maintain the decided reprobation of the use of the letters cA, chh, sh, f instead of our c, ch, s, s, as incompatible with any sound principle of transcription. It is evident that we have to resolve the Devanagari ligatures, including "^T M, into their component letters. With respect to the separation of the single words , we have to follow, against the Devanagari custom, the European principle, that every grammatically separated word is to be separately written in the latin transcription. This is effectuated without difficulty in the cases where consonants are to be separated from consonants or from vowels. With regard to the eras is of vowels between two words, we should resolve them simply into their component parts and leave it, as we do in latin poetry, to the reader to pronounce them according to the Sanskrit rules. We write therefore f^qitiT^ tat'divdsld with three words tafd aiva dstd or, after the modern pronunciation, tathd eva dsid. We think it not necessary to indicate the crasis by an apostrophe, as it has been proposed, considering the frequency of the case, and the destination of the apostrophe in European writing to indicate the elision of a letter. We prefer to make use of the common sign of diaeresis for the rare cases, where in Sanskrit the hiatus is demanded. 96 JAPHETIC LANGUAGES. PALI. M:HI CL si Gi d ;y e N Ml etc. n e c pi M a a e e oo W W a a etc. h t d n t d n p b m lr kh gh ch jh th dh th dh ph bh Specimen. Namo Tassa, Bhagavato, Arahato, Sammd, Sambuddhassa ! 1. Namas sitvdna Sambuddhd, susuddhd, suddhavd scyd; Maha- vansan pavakkhdmi, nanundnddhikarikd 2. Poi'dnehi katdpeso, atlvitthdrito kvact, ativakvaci sdkhitto, aneka punaruttako ; 3. Vajjitd tehi dosehi, sukhaggahanadhdrand, pasddasdvegakard, sutitoca updgatd, 4. Pasddafanake thane, tathdsdvegakdrakc^anayantdpasddanca, sdveganca, sundtha td. (Mahavanso ed. Tournour, ch. I, 1 4.) Remarks. The Pali is one of the older Prakrit languages, which, together with Buddhism, has been extended beyond India, principally to Ceylon, Birma and Siam. In these countries the Pali is still used by the Buddhists for their religious books, where it is written in the different indigenous characters. The character which we have represented here is that of Siam. The palatal and cerebral sibilants, as well as the vowels r and / have disappeared; the cerebral / has become a dental / 1 . ^Tand ^T have been dropped. PALI. OLD PRAKRIT. 97 T! Tt %%T etc. OLD PRAKRIT. 17 ad %T * T ^3f i 'I u u d I etc. ai au k g n h kh gh $ n - y ch jh t d n - th dh t d n s r I th dh p b m V ph bh Specimen. I f!T (Prabddha Candrodaya, beginning of Act. IV.) Maitrl: Sudd me Muditde sdasddo jadhd mahd Bhailavl gd-sanasdbbhamddo bhaavadle Vinhubhatrle parittddd piasahl Saddhetti; td ukkanthidena hiaena piasahl kohl pekikhassd. Remarks. In the Indian literature the different dialects of the popular language are called Prakrit in contradistinction to the Sanskrit as the purer literary language of the higher classes. It ap- pears in the dramatic works by the side of the Sanskrit and is written likewise with Devanagari letters. It has lost the same sounds as the Pali, and moreover the /. The letters n and n only occur in conjunction with the letters of their own class. 98 JAPHETIC LANGUAGES. HINDI. ^ ^ z , as / is an assibilation of g. There is indeed physiologically very little difference between a cerebral sand y, zh and y/i, the friction on the tip of the tongue, erected at the cerebral point, causing almost unavoidably a slight vibration of the tongue, and re- minding by it of the letter r, A perfect analogy to it is the o physiological proximity of y and r of the Arab. (cf. sL, yazah^ of which the French have made razzia), the slight friction of the y at the guttural point causing likewise very easily a vibration of the soft palate. It would therefore be more con- sistent with the genius of the language, to write those two letters z and zh\ but it seems nevertheless advisable to prefer the hitherto usual transcription of r and r/t, so much the more as already in the Hindustani writing the arabic characters V and pv have taken their basis from r, not from 3 d. The traders and in general the lower class of natives, write and print the Hindi very frequently in a character called Kaithi, which is an imperfect imitation of the Devanagari. 100 JAPHETIC LANGUAGES. HINDUSTANI. In Arabic or Persian words. f'T 8 t z 15? j? ^ !&(#) o - - tf f O t t > [ > f & \^\ ! -> ! o ^ LA L5 *f^,:: 3 ? O l | etc. - ' - - - - - (jo J (jto 0,0, '"* ^ ) LT J> ^ $ ^ J ;} vv r * W a a h t K e 6 k g h - kh gh x y i I u u c jf s y ch jh % d a I etc. t d - r th dh rh ai au - - - - i s d z t d n 8 r I th dh e d z p b m 10 ph bh f Specimen. oc Matth. 2, 1. 2. HINDUSTANI. 101 Jab Herodis bddsdh ke waqt yahudiah ke Bait-laKam me lisa paidd hud , to dekho kal majusio ne purab se Aursallm tnS dke. Kahd kih kahd hai wuh jo yahudio kd bddsdh paidd hud, kih hamne purab me itskd sitdrah dekhd aur usse sijdah karne kd de hat. Remarks. The Hindustani or Urdu is still more generally understood through all India than the Hindi. It is a mixture of Hindu with Arabic and Persian. The Mohammedan conquerors, pe- netrating into India since the 11 th century, carried with them their language and writing. The latter was received by the conquered population; but from the language only a number of words was inserted into the Indian language ; the grammar, although mutilated, remained Indian, and likewise the system of sounds. It was therefore necessary to introduce for the In- dian sounds new letters into the Arabic alphabet, principally for the cerebral sounds t d r, which were expressed by the dental bases with the addition of four dots ^ ^ J or other diacritical marks. The letters n n n and s, which already in the Hindi were of little use, dropped entirely. IT turned into U* s. *T and ^ were written or ^ and ~. The anusvara . was expressed by Q n. The merely Arabic letters .' q % y z d(t) d s z Q d z f were still written in the Arabic and Persian words , but seldom preserved their original value. *- and are not pronounced at all; ^ K is not distin- guished from A; (j& s and ^ # sound like s # 3 t t & % 5 u^- ^ ^^=- ) O .:. 3 J P ^^ LT ^ . \ . ... - 1 ^ o ' L - O 1 Jadhl Isu (jlsa) yahuda (yahudah) je Betlahama (Bait-laKama) mS Heroda je patiMha je dlhani me jay 6 ta disu majusani ubhirande khd Yirusdlemi (Aursallma) m& act cidti ta. Yahudyani jfo pdtisdhu Jo Jay 6 dhe, so kithe dhe? chd kdni ta una jo tdro ubhirande me disl huna khe pujana dyd dhyu. Remarks. The Sindhi, the language of the province of Sindh on the lower Indus, differs in essential points from the Hindi and is an old independant Prakrit language. There is a great number of different Sindhl alphabets, a survey of which SINDHI. 105 Capt. G. Stack gives in his "Grammar of the Sindhl language", Bombay. 1849. p. 3 8. They originate all from the Deva- nagari, are none of them widely different, and are all incomplete inasmuch as they do not distinguish all the sounds in writing, which are distinguished in speaking. The European scholars have therefore preferred to make use of the most generally known and most complete alphabets, especially of the De- vanagari for the Hindus, and of the Arabic for the Muham- medans. Others, as the Missionaries A. Burn and A. Matchett have given the preference to the Punjabi alphabet. Capt. Stack in his numerous writings uses the Devanagari. Dr. Trumpp in his SindhJ Readingbook * uses besides the Devanagari also the Arabic letters. The four sounds, which we have written g J d b are peculiar to the SindhL Their pronunciation is that of the letters g j d b uttered with a certain stress in prolonging and somewhat strengthening the contact of the closed organ, as if one tried to double the sound in the beginning of a word gga, dja or gga, dda, bba. The letter ^ or T is described as sounding like ddy\ but I conceive that we have it to do here with the old pure palatal ' K q _% y <](t) o s z d z f are generally expressed in modern prints by the Indian letters ^f^TsTTfT^T^^^rsrsr^ with dots below. GUJARATI. (t) etc. (T Ul a a e 6 i I n n d a I etc. & ^l H k g n h kh gh n i ^ ?5 **l ^ } ^ t y ch jh /^jT , *t i S ? ^ 9 G) i th dh *l R. ^ J. U t d n s r I th dh H y rv^ p b m - V ph bh GUJARAT I. 107 Specimen. , i and I are pronounced as in the other languages c jf ch jh, but before the deeper vowels a, Q a Q a a Q Rl q o 24 er a d etc. ai au h kh gh 1 y ch jh s th dh s r I th dh - V ph bh Remarks. The language of the province Uriya, the maritime country south of Bangal , approaches much to the Bang all, but with a greater share of Arabic words. The sounds are almost the same, but the pronunciation is said to be in general somewhat harsher, and the cerebral r is wanting entirely. The peculiar character of this country is often used there even in writing Sanskrit. 114 JAPHETIC LANGUAGES. PAS TO or AFT AN. i or t T L$t J? ' L5| 3' 2 a a i i at au u u 1 1 C C L5 t k g n x r c j - fl ( "V A "*/* V \ \ A f **\ a / y # tf - V ' 6' Z ? '/ *. 4 T < rf n 8 r I p b m V K h s d z Q d z / o ... Specimen. Matth. 2, 1. 2. PASTO. H5 Pah waqt cih lisa da yahudiyyah pah bait laHam kse yah zamdn da Herodls sdh zowalai wah^ ndydh majusdn lah masriqa nah pah Aursallm waraylah, pustanah e (ye) wakrah cih da ya- huddmt bddsdh , cih zowalai dai } da cartah dai? lah de sababa, cih e (ye) storai pah masriq kse mu lldalai dai au muz rdyll i/, cih sijdah e (ye) wakru. Remarks. The language of the Afghans is, in accordance with the geographical position of their country, a middle limb between the Sanskritic or Arian and the Persian or Eranian languages. They use the Persian characters with a few modifications. Besides the usual vowels this language has an obtuse vowel nearest approaching to a, which we write in consequence a (see above p. 49). It has in common with the Sanskritic lan- guages the cerebral row; there is at least no difference of opinion with regard to the letters t and (Z, whilst, according to some writers, the letters n and r differ in sound from the Indian n and r. Slight deviations however appear sometimes to an ear not accustomed to physiological apprehension greater than they are, or result from unessential circumstances. Con- sidering, therefore, that in most Indian languages the four cerebrals t d n r have been developed together, and that even in Afghan writing all the four characters have been likewise characterized by one and the same little circle added to the corresponding dental letters, we do not hesitate to follow those who recognise the cerebral nature of the Pasto n and r. Out of the original Palatals two new sounds have been formed besides c and J, viz. t and d, as in the Mahrdtht, to which we refer (see above p. 109). This latter did not receive new signs for those sounds, the Pasto on the contrary added one new sign ^, expressing by it both t and d. Only in modern times there has been introduced by the learned Pasto scholar JAPHETIC LANGUAGES. Dr. Trumpp a second sign r for d besides #, remedying thus an obvious defect of the Pasto Arabic alphabet. There are two other sounds peculiar to this language which we must consider. They are represented by the characters (Jt and _.;.. Their pronunciation differs essentially in dif- ferent parts of the country. In the western Afghanistan it approaches very near to the pronunciation of U* s and $ :\ in the eastern portion, for instance in Peshawer, to that of and y. According to Dr. Trumpp these letters are derived at least partly from Palatals another part seem to proceed from original Cerebrals and as the Afghans them- selves have taken the bases of their signs from u^ s and j z , we propose to take the same bases s and z in adding the palatal line 6^ and L Should it be desirable to indicate the eastern pronunciation specially we should take as bases % and y and add the palatal line j(, /. The Semitic letters which we have separated from the rest, are used only in Arabic or Persian words. We regret that the transcription of Capt. Raverty in his last Afghanic publi- cation (1860) has deviated so far both from any sound prin- ciple and from practical suitableness. OLD BAKTRIAN (ZEND). 117 OLD BAKTRIAN (ZEND). Vowels. Consonants. MM C JL& 5 (xrev v-t. 3 3 > y tf* * ti 1 \ ] 00 & ^ ^^S, I *u eb Semivowels. v ^ - _je^- G ^ b ^ ii Ligaturen. J^O C^K-? ^^ Original pronunciation. a a * d a k K h 9 9' y n n - - - - i '7 e e K - - 9 - - n - 1 1 y - u u a s z t f 6 d d d n - s *y o S r r y(ii) <) P P ~ b 1) - mm - - v v' s& s a// Later (Persian) pronunciation. a a e e an k x h 9 - y n (n*) - - - - i -I e e c - - J - - n - I I S Z y - u it 06 a s y t tf(tf) d d d n - s z r (r) y w P f - b v - m (m) - - A. o st ah JAPHETIC LANGUAGES. Specimen. o o Veudidad, first fargard OLD BAKTRIAN (ZEND). 119 Mraod Ahuro mazda spitamdi ZaraQustrdi: Azem daddm, npitama ZaraQustra, aso rdmo-dditlm noiS kudad sditlm; yedi zl azem noid daidydm, spitama ZaraQustra, aso rdmo-dditlm noid kudad-sditim , vJspo anhus axtwd airyanem vaejo frdsnawdd. [A*6 rdmo-dditlm noid aojo rdmistdni, paoirtin bitttn, dad ahe paitydrem, mas md rawa satidm haitnn.\ PaoirJm axanhdmca soiQrandnica vahistem frdQveresem, azem yd A/turd mazda: Airya- nem vaejo vanhuya ddityayd. Aad ahe paitydrem frdkerentad Ani'6 mainyus pouru-mahrko, azimca yim raoiditem zydmca daewo-ddtem. Dana awaQra mdnho zayana, dya hdmina, \hapta henti hdmino mdnha, panca zayana askare;] taeca henti sareta- dpo, sareta-zemo, sareta-unvarayd; acfa zimahe maidim, acfa zimahe zared'aem, ada zydscid pairi-pataiti , ada fraestem v&iyna- ndm. Bitim asanhdmca soiQrandmca vahistem frdQveresem azem yd Ahuro mazda: Gdum yim Suydo-sayanem. Aad ahe paityd- rem frdkerentad Anrd mainyus pouru - mahrko , skaitim yam ga- waca dayaca pouru -mahrkem. Remarks. We call Old Baktrian, as others before us, the language of the Avesta (Zendavesta), the sacred books of the Eranian nations, especially the Baktrians and the Persians. These books of the Zoroastrian religion first originated in Baktria, in the vicinity of northern India, and are the principal witness of an old Baktriau civilisation, of which we know but little beyond. It was probably not before the time of the empire of the Achae- menides that they were introduced from the east to the west of Eran, and particularly amongst the Persians. The language still approaches so nearly to the Indian Sanskrit, that it was principally by the comparison with this language that Bur- nouf and Bopp were first enabled to decipher the Zend lan- guage. The Zend writing has the same origin as all the other phonetic writings, including even the Devanagari, with the 120 JAPHETIC LANGUAGES. exception of the Persian cuneiform characters. We are of the opinion, that the Zend alphabet existed already in the original country of the Avesta in the same perfection and completeness as we know it actually, or even higher, but that it may have undergone several changes when introduced into Persia and brought in contact with other cognate alphabets of the western countries. It approaches most nearly to the Pehlevi writing. We take this character of the Persian inscriptions and of the Persian handwriting, of the time of the Sassanides, not as the origin, but as the reduction of the Zend character, answering to the poor and partly semitized system of sounds, which at that time prevailed in the Persian language. Both writings underwent apparently the same alterations in their common signs for several centuries till about A. D. GOO and then at- tained essentially that same state which we still find in our Zend, Parsi and Huzvaresh manuscripts. In the mean time also the original pronunciation of the old Baktrian alphabet was altered, since its migration into Persia, in conformance to the altered sounds of the Persian language, as they prevailed already in the time of the Achaemenides and still more in subsequent centuries, just as the pronunciation of the Devanagari letters approached more and more nearly to that of the modern Indian languages. The right apprehension of the old Baktrian sounds is traditionally preserved only in the alphabe- tical lists, which were faithfully, but, owing to the ignorance of the writers, incorrectly, copied from one manuscript into the other and thus handed down to us in a tolerably comprehensible state. The arrangement of the original sounds as above stated, is principally the result of the comparison of those ancient alphabets. It ought, according to our opinion, to be followed in every linguistic publication on the Zend language and might even do good services in a critical revision of our actual text. The vowel system is the most developed of all the ancient languages we know, not excepting even the Devanagari. OLD BAKTRIAN (ZEND). 121 There are two characters for each of the vowels a i u e e o a, one for the short and another for the long ones. The semi- vowels ii and , y and w, were, conformably with their signs, reckoned not as consonants, but as vowels. All the explosive consonants, r included, had both a simple and an aspirated form. Two nasal sounds are lost even in the alphabetical lists, where they are represented by the repeated sign of the simple n \ . The pairs of corresponding sibilants are given as stated above. The letter -C 5 ' was, as a palatal consonant, different from the semivowel ii, y, in the same way as & v from w. The letter r had two signs, the latter of which was formed by adding to the simple ? r the upper stroke ^(J>)to indicate the aspiration (as in K, f, p'). fc* was an aspirated v>. The later pronunciation changed the aspirates K f p g r B into the corresponding fricatives % Q(s) f y y. In consequence of this change the sign _l> disappeared entirely from the ma- nuscripts, and ^ became an almost arbitrary variation of S^. We retain the writing d, although the actual pronun- ciation seems to be not quite clear. The letter <^ cf escaped the assibilation; but it lost the aspiration and was pronounced like d\ we keep however the hook to distinguish it from 4 d etymologically. The aspirate & I) was softened to v (perhaps to the German iv see p. 75); we write it v in putting a dot beneath, only to distinguish it from the afterwards identical initial v {} The aspiration of n was lost , as that of n already previously ; m is mostly dissolved into hm. The palatal sibilants ou, s and S z took almost entirely the pronunciation of s and -0; we write them s and z to indicate their palatal origin. On the contrary *^ s assumes very often, in the mouth of the Parsis, the pronunciation t3 s, and still more particularly z that of z. There are linguistical reasons why we should not, in this case, follow them, but adhere, in our transcription, to the old sounding, although the usage of European scholars would be in favour of z for eO . The pronunciation of **O s as I 122 JAPHETIC LANGUAGES. * caused its confusion with l$p , and >*O , originally c, is softened to y and used almost as identical with -C . This latter circumstance will perhaps justify our transcription of ^*O by f, with the basis of y, according to the actual pro- nunciation, and with the addition of our diacritical sign of assibilation v , to denote its former sound. The letter -O^ y is (like 1} u) always used in the beginning of words, Si y (like w) in the middle; it is nevertheless necessary, in our transcription, to distinguish them both, one being considered as a consonant, the other as a vowel. We therefore write "O y with the palatal line. The aspiration of J / dropped, and then - /' became identical with / r. On the contrary the aspira- tion of t* v increased and gave rise to the later pronunciation _.', and ultimately to % alone, thus producing a confusion with G* %. We write the ^ in virtue of its etymology _%. There is no indication of its ever having been pronounced explosively as q or kh. The first of the three ligatures ^HJ, 6^', was in later times pronounced s, and therefore occasionally confounded with *>. A glance at these alterations shows, that the principal dif- ference between the old and the later pronunciation consists in the disappearance of the aspirations, which were peculiar to the Baktrian throat, and which either dropped without any compensation, or changed the explosive sounds into fri- cative. With respect to the vowels there is a general influence of the western languages to be observed in the less decided distinc- tion between the long and short vowels. This is the reason of many confusions, and explains, how the letter tt) e had been hitherto taken for e and as almost identical with {O e, and why moreover and f, *> and ?, or even i and y, > and^> are frequently used in the MSS. for one another. The most striking change however took place in the letter J& , which originally and still in the alphabetical lists, represented the OLD BAKTRIAN (ZEND). 123 anusvara of a, viz cf, corresponding to ^? &> but which afterwards was used as a separate consonantal nasal after the vowels and before certain consonants. The fact is, that the difference between a and a was obliterated ; the sign be- came therefore disposable and was employed to express to a cer- tain extent what is called (though not rightly) by Sanskrit scholars the substituted" anusvara, whilst ^ was reserved for the w ne- cessary" anusvara. This, now almost constant, use of .w obliges us in spite of the evident misunderstanding which gave rise to this use, to look in our transcription for a corresponding consonantal sign, and there could hardly be a more conve- nient one than n , the more so as it is already employed by Bur- nouf. Simultaneously with this corruption, the entire designation of the anusvara sound, which existed in the Baktrian, but not in the Persian language, fell into confusion, and the peculiar expressions, which, according to the alphabetical lists, must have originally existed, were either exchanged for jg, n or w, or they disappeared altogether lengthening only the remaining simple vowel. The vowel sign S^ seems to us composed not of **** and , but of **** and > , and to have been originally the diphthong du, which was afterwards changed into a. We do not state at length the reasons for the assertions as given above; they will be found discussed in a special disser- tation read before the Berlin Academy. In the specimen, we have made use only of the later pro- nunciation, because our manuscripts contain it to so large an extent, that we cannot substitute the original sounds without altering the actual state of the text. 12 124 JAPHETIC LANGUAGES. OLD PERSIAN (CUNEIFORM). - T E TE M *^l - T * YT >T^ =< = Foreign characters of doubtful meaning: ^ a ka,i Ku /,(.'), a Ca,t,(u) - 8 a, i, u tj ta y i fu Ga, i, M M, O p ,,-, u - - i au ff'jt tf<> - fla,', u y;, >, () - %i Sa,i,(u) Ta,i fa da,i W \ rr m ET X> Tfc W f! ,<- A "TE yT ET Tt \ TT 5 fir r<> T 'm t< m >W \ TT ^< TIT X> W rf *\ f, W ET TE TT T<> \ TT ^< m X> KT f f r<> A fr <=< ,< m A ^ \ T KT ff Y<> \ 'T'T c < rff A ?r ff -TfT W ^ S " << ET 'T fir TE \ ft ff ^< Irf \ Sif'Mlrf A mET^fif ^Trl A OLD PERSIAN (CUNEIFORM). 125 fEr^Tfr *WTff A ^'MlrM TfTSTffX'fif T -W r< \ Tff ET ff X> Tff El >M -< TfM if TT rn A fr ff 3< ?? ff ^< A, fr f? 3< ?f ff 3< M TT r<> \ ft rrtET ETT ff ,< A ^TE r<- >W \ M K ff ^< f f r<> !ff \ ! m >M >M \ tif 'M TTT ffrT TTT \ tif 'M <*< r<> \ M \ tTiT KT ff X> A Hff A 'M -< TTT A 'M M tff X> fff A *M X> ff X- Tff X- W ff X> TfT \ Tff < f < \ fff ff * A ~~~~ } 3 ' fcy <.. (J" [) ^5 f'J f 1yD f v^ lo O O LT J ^ .Ji*' ^_^^ ^^*^ '^^ o 6 V > v_j __ 3 Arabic letters Persian sounds. with their Persian pronunciation. ad - ' - I h /(') K,(Ji) \ e o kg - \ x y ?c^) i l u u c J - 8 Z y ai au di ui - - - - - - d(f) s() z(z) J(0) t d n s z r 1 B(s) d(z) p b m f 'W Specimen. ) ,0 } 1 11 I * 5 1 - -9 j 1 -^ MODERN PERSIAN. T"- Firdusi , Book of the kings (ed. J. Mohl, t. II, p. 4.). Zi-mobed bed-in gune ddrlm ydd, liem ez goft i an plr ^ dihqdn-nizad: k-ez dn-pes cundn kerd Kd'us rdi, ki der pddisdhl bi-jumbed zi-jdi, ez Iran bi-sud td be-Turdn u-Cm. yuder kerd ez dn-pes be-Mekran-zemin, zi-Mekrdn sud drdste cun >arus: ber-dmed dem I ndi u-buq u-km. bi-pedireft her mihterl bdz u-sdw } nekerd dzmun gdw bd sir i taw. cundn hem gurdzdn be-Berber sudend, jfihdnjui bd tdj u-efsei* sudend. Remarks. It is well known that the modern Persian language has admitted a great many Arabic words as well as the Arabic character. It uses consequently all the Arabic letters in the Semitic part of the language, but only in writing. In the language spoken the purely Semitic letters f ,> ^ -b Jo u^> (j& are pronounced like the Persian sounds \ ^ j u* j , and even the letters ^ and , which once had their peculiar pro- nunciation as # and r f I w It. n m jl_ Specimen. pJ ltUL - kp tfb? i uiutib , n* nui/iui&.g* np jhnl^Kbu Y ul '^*c/^/ > ^/ > ^ O^inLiiA nub iiilini li p j n P'uifiifjfutlilt aiuluin l^uitT ihiurL^p : C) ^uiaiun uiiriui^ui in filial" ^ p-h-fifrLu n[i'ft> ty l^t'9b ^"'"'i nnnt-ir uibnilli [|^"^'^l-, np u //_ 2/'/ l Y/'/ 1 ^- nuiJt^Luijti nn 'A Y/num* uinJifiijI^. L. jbut ijfe J'u^ui uiiLbhinf uLuuiL ui&hi qJinuiL, ninf, . mjiii in /"'^ utu qnp luitHrtf^ L. ifibfcgf' [fb& "L'tb [1/"^Z'^"' ^/^^ '/" jniyy niliij [i Jon tiLnlruiLn. Eznik, Refatatio haeres. (Petermann, Gramm. Armen. p. 44.) ARMENIAN. 133 Minj jyev pnau er inj, asyen , woj yerkinK yen wo] yerkir yev woj ail inj araratK wor Kyerkins kam Kyerkri, Zruan womn anun er, wor fargmani ba%t kam pafK. Ezhazar am Hast arar zi fyeryevs wordi mi liniti nema, worum anun Ormizd, wor zyer- kins yev zyerkir yev zamyenain wor i nosa afnite. Yev hyet hazar ami Hast afnyeloK esksau atyel ezmtau, ase: Ogut inj its Hastes zor afnyem, yev liniti ind wordi Ormizd, yefe i zur inj canaityemf Yev minj dyer na zais Corker , Ormizd yev Arhmen Kyyetan f'argandi mor iureant. Remarks. We are told that the actual Armenian alphabet was made up in the 5 th century by the learned grammarian Mesrob. The figures of the letters are taken from the Greek figures of that time, as their uncial forms show; the alphabetical order fixed by the numerical value of the letters, is likewise that of the Greek, but occasionally interrupted by those purely Arme- nian letters which were unknown to the Greek. The ac- cordance of the letters common to both alphabets, as well as the orthography of the ancient proper names, and to a certain extent even the etymological comparison of the cognate In- dian and Eranian languages, with which the Armenian is closely connected, prove that the actual pronunciation of a large part of this people differs in some points widely from the ancient pronunciation. We speak of the Turkish Arme- nians, better known than others in Europe by their country- men in the Mechitarist convents of Venice and Vienna. The principal alteration of the old pronunciation consists in their pronouncing the ancient tenues as mediae and vice versa. In Armenia proper, however, and the surrounding provinces this change has not taken place. This has been carefully verified by the author examining personally the pronunciation of a learned 134 JAPHETIC LANGUAGES. Armenian from Tiflis. ' We can not, in consequence, hesitate to abandon the usual system taught, according to the Turkish pronunciation, by European Grammarians, and to follow that of the indigenous Armenians. There we find the letters l[ "' "[,ktp distinctly pronounced without any aspiration as real dry tenues' 2 like those of the Hungarian, of several German dialects, of the Sanskrit and other languages; '/ >t /', (j d 6, are our common mediae and . fi ^, K f p\ the true aspirates, pro- nounced as the so called tenues of northern Germany, France, England and others, with a sensible breathing from the lungs. The pronunciation of the two palatal classes is more difficult. There is no aspiration heard in _ and #, although they cor- respond evidently to the aspirates of the other classes. Only the stronger closing of the organ is the same as in the aspi- rates, whilst, in opening the organ, the aspiration turns into a slight breathing z or 2, as if one would pronounce ttz and ttz. We write therefore c and t to indicate the double value of the first element. The c and j are pronounced nearly as in churcli and in join', but Z d and t are hardly discernible, the one being pronounced as cfc, the other as tz. The tongue takes in both palatal classes its full palatal position, in the first more behind, near the soft palate, in the second more foreward above the teeth. The letter i_ is now pronounced every where as the arabic /, although it is proved, that it was in many cases formerly, and still in the 5 th Ct., a kind of soft and more palatal Z, distinguished from the stronger and more guttural , the value of which is now that of our common I. The i_ originated from I was therefore in former 1 Cf. the Armenian Grammar published in Armenian by Bogratuni (Venice 1852), and Petermann, who states the same in the Memoirs of the Berlin Academy 1860. p. 82. * See my exposition about the true character of the tenues, mediae and axpiratae in my treatise on the Arabic sounds and their transcription in the Memoirs of the Berlin Academy 1861. p. 105 sqq. ARMENIAN. 135 times written ^ or f_[ , and took its name ^j^ T^ un > not from the qu,u, yat, but from the ^ti linn. Hence the mistake, that in several Armenian grammars the letter t/, is reckoned amongst the liquids. It would perhaps be recommendable to linguists to distinguish the even in the transcription from the original i_ by adding the original hook (q_ = /, - [K]h k g n x r K g n y * 3 t d n s z s z r I d 1 p b m j V w au ai eu ei Specimen. Jdiki bl ci hebl, yau Aldh be, yau arewdnci be. Arts teKndini. RoJ yeke beri dri qdfelnd se keiye, sd>ue keiye rd kaut, sebah weri$t, ami areye^ aundike drdi mesdhidi cini; ay a sdu>e newete pa'ttte, aunaike mydne sd>ue yau lu>e dntei zere, sie mesdhe, drdi iverdi. Arewdnci werist, yau cua gerdute^ dai lu>eru; erzia lu>e tepiste, lu>e bermdi. Lu>e vd ke drewdn&ra: te me vera de, ez tueri keinai pasdde Misri wazewa. Arewdnci vd ke: ez yau merdumu drewdncia, ti meri citdu ktinai pasdde Misri waze'n'i? Lu>e vd ke: ti me mekse^ ez tueri ioazen*a , eke me newaiste keinai pasdde Mjsri, ti me verd de. Arewdnci vd ke: ti meri sudnd budne. Lu'e drewdnciri sudnd wend, arewdnci lu>e verd dai. P. Lerch, Forschungen fiber die Kurden. I, p 88. Remarks. The language of the Kurds in the mountanous countries of Kurdistan and Laristau seems to be divided into 5 chief dia- lects, these of Zazd, Kurmdnji, Kelhuri, Gurdni and Luri. KURD. 137 The alphabet given above is that of the Zazd dialect, as fixed by P. Lerch in his Researches on the Kurds. The Kur- mdnji has no n nor d'. The Kurds have no peculiar writing nor any literature. There exist only some modern specimens of their language written in Persian letters. Mr. Lerch therefore employed in his work the Standard alphabet (see above, p. G). The vowel * is not our ', but our vague vowel e approaching to i. Instead of Mr. Lerch's t and d we have now proposed to write c"and J. There are in the Kurd language many Arabic words, which, if written in Arabic or Persian writing, keep all their Arabic letters. But, as the people have no written literature, we think, that Mr. Lerch was right in transcribing only the sounds that they really pronounce, amongst which we see also the Semitic letters q and K. 138 JAPHETIC LANGUAGES. (a) (2) - S S E T I A N. Georgian characters. a - - - 6 6 d - A' u o 6 - ? B - ? a 9 - 6 O5 D 3 * 9 o* I a a p - I g I n K> g K - c> J c - t> d t - t> d t n P' b P m y r w Specimen. Dialect of Jalgusidse. fid keti de arwiti mideg; siydeg went de nom; ertewent de meligad', ^vent bar dew , kwid arwiti mideg, afteder bast'il; jul ne bonti radt> ma^en abort; erne niwaj ma%en ne %este, kwid ma% niwajfst>eni ne ^esginten ; erne ne ma baftaw ma% filewzeren^ fele ferwezinken ma% fidbilfzey; emencme dew u meligad , ti% erne isdirad mikk>aqm&n Amin. oo o y Digorian dialect. Ma% Jide keci de arwitiy miedeg; qedoz node de nan; artauode de p'at'oxjindde; uode de bare, arwiy miedeg kud, zan%ay bgl OSSETIAN. 139 lioteder; ma% dol taruniy tu^ey rddPe ma^an dboniy; ama %aldrkane niej^aste ma^an, mapper kud %aldrkandn nie yasKinten; ama ne ma fardduyunkane; falfayerwdzunkane ma% fudbuluzey ; oy tu%ey dawon ey paPa^jindde^ tu%e ama sPurjindde mukbdgey mukk'dgma; fauod. Tagaurian dialect. Ma% fid keti de arwitiy mideg; siydeg uod de nom; ertauod de paPa^dindd; nod de bar, arwiy mideg kud, zu%il auteder; maj% did oninen rddt> ma^en dbon; erne %aldrken ne %aste ma%en, ma^Per kud ^aldrkenem ne ^ast'inten; erne ne ma ferdduinken ; fele ferwezinken ma% fidbilizey; ay ti^ey key is dewon paPa^di- ndd, ti^ erne st'irdindd mikhdgey mikk'dyme; fauod. The Lords prayer. Sjogren, Osset. Sprachl. p. 32. Remarks. The Ossetes have uo alphabet of their own ; but their s}'stem of sounds approaches so much to the Georgian or Grusinian, that the Ossete Jalgusidse, as well as afterwards Dr. Rosen, both found it convenient to employ the Georgian alphabet to represent the Ossetian language, for which it was necessary to add but a few consonantal signs. It was by far more difficult for Sjogren to adapt to the same language the Russian alpha- bet. This eminent scholar distinguishes three dialects, two northern, the Digorian and the Tagaurian, and one southern, which he met with in the writings of Jalgusidse. We use above the Georgian alphabet , to connect our transcription with former labours. The letters J 3 $ (Jc> g K) have been added by Jalgusidse, to denote the softened sounds rising from the gutturals k> g k before e i i. Sjogren doubts without reason about the letter ', which is even demanded to complete the system and which may exist too in the two other dialects without having been remarked. He explains erroneously the letter 2? as g' or gh y since it corresponds, as Rosen rightly E2 140 JAPHETIC LANGUAGES. states, to the sound of our /. The special designation of n by Sjogren seems to be less essential , since neither Jalgusidse nor Rosen have made this distinction. We keep however the letter i/, though not distinguished by Rosen from i. The Di- gorian dialect, according to Sjogren has t' d' instead of k ) in order to utter the appertaining vowel. Hence our transcription k>, t', p> etc. The aspirates of the third column have but a slight aspiration, and we write them so much the more without the spiritus asper (K f etc.), as their opposition to the first column is already indicated there. As to the vowels Sjogren makes very nice distinctions, especially of the different e sounds, of which he states three gradations, observing, that the most open e (his ce) approaches very nearly to the English a in hat, fat. It will certainly be found convenient to reckon his second gra- dation, the short ?, and his third gradation, the long e, both being nearer to i than a>, as one class in our transcription. We get therefore two classes, e & and e t,' All Ell H1I Oil Zllf 4J (A ^ t) a k g x - e 0(6) c j s z 'i i u t d _ e o t d n s z t(j) u p b m (f) ^ e(ai) n r 1' (K g j[) ja je i()i} ju }e jo J ll t j$ an d jo. It might be desirable and would certainly be more consistent with the genius of Slavonic language to write these compound vowels by single caracters as it is done in the native writing. The division into two signs, however, is preferable by more than one reason, and is not against the rules of our standard alphabet. We have thought it convenient to write this slight /-sound not by a full consonantic y, but by.;', re- presenting as if it was half a y, although this sign j does not occur in our general alphabet. This designation offers besides the advantage of being in concordance with the almost ge- neral mode of writing adopted by most Slavonic nations writing with Roman letters, as well as by most of the respective lin- guists. We have therefore preferred it to any other designation, which we might have invented. In a similar way also the shor- OLD SLOVENIAN. 149 tened and indefinite vowel e of other languages, behind or be- tween consonants, has taken in Slovenian either the palatal or guttural form and has become a surd yet perceptible i or u (', u mutescens). These two sounds h. % and 2 u are in most cases remains of corresponding fuller vowels and were no doubt for- merly more distinctly pronounced than they are actually. There can be no doubt that we must follow the original writing in transcribing them by separate characters. We have chosen for that purpose the two signs i and w, adopted already in the Rumanian alphabet, as well as by several linguists. The sign of brevity ^ is the less objectionable, as the distinction of long and short simple vowels is little known to the Slavonic languages. The later pronunciation of the Old Slovenian drop- ped the u more completely than the i; yet, for etymological reasons, u most also be written even in cases, when it seems it was not any more pronounced. The distinction between guttural and palatal pronunciation applies also to many of the Slovenian consonants, accord- ing to their combinations with either a hard or soft vowel. The palatal modification, however, as in the case of the vowels, is also in this case the predominant, and the only expressed in writing. The dental consonants, by this modifi- cation, are generally assibilated; the liquids n I r and in foreign words also the gutturals k g % become palatalised or softened, in which case they are marked by some diacritic sign. The character K however is not by itself a sign of this palatal modification, but represents still a real vowel, which only bestowes on those of the preceding consonants, which are capable of it, the palatalised form. It would be, therefore, in- exact to replace in our transcription the i by the palatal line added to the preceding consonant. The vowel li is unknown to the Glagolitic alphabet. It is evidently formed by a combination of the Glagolitic *f- a with the Cyrillian b I and represents therefore the diphthong fa or 150 JAPHETIC LANGUAGES. at. But since the diphthong la or ja had been expressed from the beginning by the sign M (analogous to the combinations je ji ju je jo) and since Miklosich has proved that the character '.t could not in many cases where it is employed represent a combination beginning with j, we do not doubt, even without referring to other reasons, that the value of t was originally oz", afterwards m ft, II C 3 ja jc jn jo jy fa $ - aj ej nj oj yj n 6 31 (*) B .1* H Latin writing. a k g h e o c dz - s z i M c _ r t d n s z ja je ji jo ju c dj aj ej ij oj uj p b m (f) v Ij nj Standard alphabet. a k g h e o C *? J S Z i u t _ r t d n s z ja je ji jo ju J - - aj ej ij oj uj p b m (f) v I' n Hard vowels: a o u r Soft vowels: e i 152 JAPHETIC LANGUAGES. Specimen. 6o>Ke! HJIII rpMii, H.I* cc .5 oi.ii,; i mpece, HJI y^apa Mope y opcroee? liniiiii rpMn, HUHI* ce ae.Mwiba rapece, ' y^apa Mope y operoBe, 6.iaro CBemiiine.ii.ii: Ilemap H ceenni HMKO.I.-I. Ceemii loean H ceemii If.nija II ca m.iMia CBemH llaiinir.iiija: HbH^f ,H.I;I.',SI 6.iaHcena Mapnja, POHH cyae HHS 6njejio jmi^e. National songs of the Serbians ed. by Vuk Stefanovit Karajii. Mili boze! cuda velikoya! Hi grmi-, if se zemla trese It udara more u bregove? Niti grmi, nit se zemla treae, Nif udara more u breyove, Vec dijele blago svetiteli: Sveti Petar i sveti Nikola, Sveti Jovan i sveti Ilija, I sa nima sveti Pantelija ; Nim' dolazi blazena Marija, Roni suze niz bijelo litse. Remarks. The Serbians, as far as they belong to the Greek church, write their language in the Cyrillian character, the Roman Catholics in Roman letters. This language, to which the Khorvatian (Croatian) is nearly allied , has neither the vowel /, SERBIAN. 153 nor the two Old Slavonian surd vowels b I and T w, of which the former appears only in combination with the characters n and /, indicating their soft pronunciation, which we render by the pa- latal line, n /', as in all the other Slavonic languages. About the palatal modification of the vowels we refer to the Old Slovenian. In Serbian the letter p is often used as vowel and forms syl- lables, although it is not distinguished in writing from the consonant p. We think it the more necessary to mark the vowel r by the little circle underneath, as it even occurs be- fore other vowels, for instance grote umro in three syllables. In writing c s z for 4 ill JK we have the advantage to be in concordance with the latin alphabet used already in the country. The letters tS and t) express very nearly the same sounds as c and dz in the Polish alphabet, where we refer to for our transcription by c and j. The signs y ( /^VK , j) and f> (j) have been added by Vuk Stefanovic. The pronunciation of X comes nearest to h, as it is written in the latin alphabet; but in many words of Serbian dialects it is either omitted or replaced by other consonants. 154 JAPHETIC LANGUAGES. RUSSIAN. * i BI y H V II C H> ail eii oil yii >i K r - x r($) i - III VK i( m ) * ii C 3 ii 6 M (*) *) t t w Palatalised vowels : Hard vowels : a o u i Soft vowels: e e i m, (o) II. II. lib Ob 3b .11, (nb 6b nib pb Bb) f - t d n s z p b m (f) v If (0) Palatalised consonants t' c( n s z' I (p b m t 1 ' r) p .1 r I Specimen. Bo AHH Tli Bbiiu.io OTt Kocapfl Aflrycxa noBO.i1;nio, o,i,'li.iaTb nepenncb no Bceit .H-M.I i,. Cin nopcinicb obi.m nopitan in, npait.ienie liiii|iiiiiiii Ciipi'io. II iiiiiii.iii id- 1. iiniici,iit;i i I.DI. i,a>K r T,i.iii BT, onoii ropo,i,'b. Houie.Tb i ,n,,i,r ii l ro.nr.ien, ns'b Fopn^a Ila- sapcra, B'b ly^eio. Bt ropo^'b ^aiui^OBij, ici.ti.iiinon.iii Bno.iecM'B, noToaiy HTO OUT. 01.1.11, us* ,^o.>ia 11 po r i,a ^aBii^OBa, e'b Mapicro, ofipyHCHiioio psiy H>CHIO. KoTopaa ni,i.!;i Bi> CwTiiocTb nx* xaMi., n. -ic i \ n n. in npe^ifl po,i,nTb eii. II po- in.iji < i, in. -i CBoero ii*-j)itriui..-i . ii cne.ieiia.ia ero, n iin.ntvi.ii.ia rro B-b )n-. in : noTOuy HTO lie MI.I.H. HMT. M-fecxa BI. rocTiimini^ti. Ev. Luc. 2, 1 7. RUSSIAN. 155 Vo dni tje vislo ot Kesarja Avgusta povelenije, sdetat' pere- pis po vsej zemle. Sija perepis bila pervaja v pravlenije Kfirinija Sirijfju. 1 posli vsje vpisfvatsja^ kazdij v svoj gorod. Posot tak:c i Josif iz Galilcji, iz goroda Nazareta, v Judjeju, v gorod Davidov, nazivajemij Vifl'ejem, potomu cto on bit iz doma i roda Davidova, vpisatsja s Marijeju, obruconnoju jeniu zenoju, katoraja bila beremenna. V bitnost' ji% tarn, nastupilo vremja rodit jej. 1 rodila sina svojeyo perventa, i spel'enala jeyo, i potozita jeyo v jasfi; potomu cto nje btfo Jim mesta v gostinnite. Remarks. l The pronunciation of the Russian has gradually deviated very much from the Russian orthography. It would therefore be useless to transcribe Russian words literally. The Russian vowels especially are subject to varying pronunciation, f. i. a = , /?; :> = e, e; -l, or < = e, e, o, je, je, Jo; i = , ji> o = o, a; iii or wii = ?}', fj t oj, and thus the Russian alphabet as we have given it above, only refers to the re- gular and original value of the letters, not to their many later modifications. We agree with the best Slavonic scho- lars in transcribing the Russian letters according to their pronunciation and not to their character. The two characters f, a and 'k were no doubt originally the same as the Old Slove- nian t; (a) and -I;, i. e. e or e and e. But since the pronunciation of * had in certain combinations passed into e, both signs were frequently interchanged. In order to economise our dia- critical signs we write e instead of e , keeping the diacritical point merely for , which in Russian do not form any more a distinct articulation, but only indicate the preceding consonant to be either palatalised or not. As we transcribe the palatalised consonants by adding the palatal line, we do not require any other sign for t. After gutturals the sign & does not occur and their pronunciation before the soft vowels is hardly more different from that before hard vowels than in any other language; we, therefore, do not express any pala- talisation of gutturals. The consonants of the palatal row are not capable at all of a second palatalisation, or it is here at least so week, that its existence is denied by several scholars. We may, therefore, conveniently omit it in our transcription. It has also been doubted, whether the labials and r are ever pa- latalised, and this palatalisation, if it ever exists, is at all events, as far as I can perceive, much weeker than after the dentals. It may, however, be indicated where according to the Russian orthography the sign i> is still written behind them. The guttural or hard .1 t of several Slavonic languages, which is pronounced with an energetic depression of the middle tongue and a simul- taneous raising of the behind part of the tongue at the guttural point, differs so perceptibly from our common Z, as to claim a special sign, and as the character I for this letter is already generally used in the Polish and Serbian writing, we can not hesitate to adopt it also in our transcription. The palatal line of/ 1 may be dropped. The letter r between vowels ceases to be explo- sive and becomes the soft fricative which we write/. The letters v -I- < only occur in Greek words, v with the double pronunciation either of i or v, which we render by these two respective signs. CHESKIAN (BOHEMIAN). 157 CHESKIAN (BOHEMIAN). a a e e o u(6) i i y y u ou(u) I r jajdjejiji a J y e j y v j 4/ U J yj w * (d) - ch h c - - s z J c - - - - t d n s z I r p b m (f) v n r t d a a e S o u i I i I u u I r O o ja jd je ji jl a j aj e J y *j J J U J iJ v Hard vowels: a o u y Soft vowels: e i * (d) - X h t - - 8 Z j t - - - - i d n S Z I r p b m (f) n r t' d' Tcsme se blahou nadeji, Ze se vrdtt zlate easy, Ze se ndm zas vyjasneji Ceske hory, ceske lilasij. At? jen cesky sat se nese, Muzne hdji mravy ddvne, Nade vsecko ono slavne, Pravoceske: Specimen. Transcription. Tesme se blahou nadeji, Ze se vrat'i zlate cast, Ze se ndm zas vijasneji Geske hori, ceske hlasi. At' jen cesk-t sat se nese, Muzhe hdji mravi ddvne, Nade vsetko ono slavne^ Pravoceske: 158 JAPHETIC LANGUAGES. Milujme se, nedejme se, Vybime se, napime se, Milujme, napime se, A pak vybime se. Amen, rac to Boze ddti! Oroduj za nds, svaty Vdclave, Vejvodo Ceske zeme! Milujme se etc. vybime se. Milujme se, nedejme se, Viblme se, napime se, Mihijme, napime se, A pak viblme se. Amen, rac to Boze ddti! Oroduj za nds: scat7 Vdilave, Vejcodo Ceske zeme! Milujme se etc. -- viblme se. Dokud! v nds krev otcu plyne, Hrud* zahrivd, ruce sili: Slav a ceskd nezahyne, Hlavu ztyci lev nds bily. Tak jako medvedum v lese, Neprdtelum budem hrdti, Oni budou tancovati, Az zapejem: Milujme se etc. vybime se. Do/cud' v nds krev ottu pline, Prud' zahrivd, rute sill: Sldva ceskd nezahine, Hlavu zt\cl lev nds bill. Tak jako medvedum v lese, Neprdtelum budem hrat'i, Oni budou tantovaii, Az zapejem: Milujme etc. viblme se. Hussite song. Remarks. Although the vowels y and y are in modern pronunciation scarcely distinguished from i and I, yet we have not conceived ourselves authorised to give up in our transcription this dis- tinction, which, besides being historical, is moreover still pre- served in several phonic combinations as well as in the general pronunciation of certain dialects. The single i is frequently, especially in the beginning of words, pronounced like ji, by which signs we transcribe it in this case. The long o has changed into u and is now written u, which writing we have preserved as both concordant with pronunciation and etymo- logy. The vowels r and / are in the Cheskian alphabet not -CHESKIAN (BOHEMIAN). 159 distinguished from the corresponding consonants. We dis- tinguish them, however, by our usual diacritic sign. The letter y has been regularly (except in foreign words) replaced by A both in pronunciation and writing, which change we must adopt in our transcription. Also the letter / is of foreign origin but has been introduced in a few native words. The palatalised consonants n d* we mark by the palatal line n t' d' . The palatalisation of r has in Cheskian as well as in several other Slavonic languages, passed into a slight assibilation , coming up to a combination of r and z , for which we have preserved the national transcription by r. The pa- latalisation of consonants is in Bohemian sometimes marked above the following vowel (tesme, nadeji, muzne, instead of t'esme, nad'eji^ muzne) or even sometimes not marked at all (vrdti, oni instead of vrdt'i^ ohi). We do not follow this irre- gularity. 160 JAPHETIC LANGUAGES. POLISH. a k ff - ch(h) e o cz dz sz z j e 6 c dz - - - i y u t d n s z r I e a c dz 8 Z /, -m f~ i c,cz,z dz } da - - - t d n s z c,c,cz,cz d8 } tz,ts - - - p b m (/) v } w ' ' 1 ' 7 T* 1 ' l f f ' J? f a e o >a u e o o u & ij J W U J k (9} - * k t 3 - 8 fit t d n S z (j) - - - p b m CO V n r t p b' v m f SORBIAN. 163 Remarks. The pronunciation of the vowels written by the native Sor- bians e and 6 approaches very near that of i and u , although it seems not to be identical with it. We distinguish them, therefore, also in our transcription in writing them e and o, the acute ' being against the principles of the Standard alphabet (see above p. 47). For the same reason, we must replace the signs ch and y by our % and /; but we keep the character , although it is mostly pronounced as w and has preserved its original value only dialectically. About c see above our remarks on the Polish alphabet. The letters which we have renderd by d and j are distinguished as peculiar sounds by Tecelin, Seiler, Jordan and others. Miklosich, however, does not mention them at all. The letters g and / appear only in foreign words. 164 JAPHETIC LANGUAGES. RUMANIAN (Walachian). Cyrillian letters. a K r e 'b o 4 If. i * y, T,m 4 1 2 OB r n m x - m JK C Z,3 I- B iai ei n 01 ai /fti r bi u ^ ^ ) p .1 ia ic 10 in ea oa e 'f o i i 0) i a ai ei etc. Mixed letters. t d n B,h n ni X - S Z .1 Etymological alphabet. a e o i u &=dei6il\,,3> \ a e i o u 'b = a e i o u\ i u at el etc. c,c t d tji p b st,sc 111 f RUMANIAN. 165 a e e o L i u (') Standard alphabet. k g J t d t p b m r I ai e t 01 m n e ait eii iu ou (iu) eu \L / i_ ia le to hi ca da eau em oiu st sc Specimen. Cyrillian letters. Illi a *OCT s.ic.ie aie.iea emr-a ^tMa Kecap Ar*CT, CT. cc cnpie xoarb AneacTi. cnpicoape ^nr^i c*a 4'bKirr, ( 1 ipica Kipine^. Illi Mcpijea TOII,I CT ce cnpie, *ie-Kape ^n MeTaxea ca. Ev. Luc. 2, i - 3. Mixed letters. JUi a *ost in zLiewie ane.iea, emit'a nopanka de Jia Kesap A^gi^st, s'i> se skpie toat'b ^amea. A l ieastb skpisoape intm? s'a *'bkat domnind in Sipiea kipine^. Illi mepgea tout s'b se skpie, *ie-kape in Hetatea sa. Etymological writing. Si a fost in dilele acelea , esifa porunca de la Cesar August, sd se scrie tote lumea. Acestd scrisore dntini s'a fdcutj dom* nind m Siriea Cirineu. Si merge toti sd se scrie , fie-care in cetatea sa. JAPHETIC LANGUAGES. Transcription. Si a fost in zilele acelea , esit'a porunka de la Kesar August, se se skrie toate himea. Aceaste skrisoare intiiii s'a fekut, domnind in Siriea Kirineu. Si merjea toft se se skrie, fie- kare in cetatea sa. With accents. Si a fost in zilele acelea, esit'a porunka de la Kesar August, se se skrie toate, lumea. Acedste skrisoare int'nu s'a ffkut, domnind in Siriea Kirineu. Si merjea toll se se skrie , fie- kdre in cetatea sa. Remarks. The Rumanian language which is principally spoken in the Principalities of Moldavia and Wallaehia and also in Tran- sylvania and some isolated neighbouring districts, originated in the Roman colonies of ancient Dacia, and has preserved, not- withstanding many foreign admixtures, an essential Romanic character. It was written formerly with Cyrillian letters, which, however, in modern times have been replaced by the Roman, being less heavy and more convenient for Euro- pean literary commerce. The introduction of the Roman letters is owing to the endeavours of a number of native scholars and extends already even to the newspapers. \\e may predict that it soon will be generally adopted. It is the more to be regretted, that it is just the most learned of these reformers of the Rumanian alphabet, who have encumbered this reform with unnecessary difficulties by following, instead of the phonetic principle which prevailed in the Russian -Ro- manic alphabet an etymological. Whilst most other nations justly endeavour to render their orthography as much as pos- sible accordant with the successive changes of pronunciation RUMANIAN. 1(57 and to avoid thereby such inconveniences as are most con- spicuous in the English orthography, the Rumanian scholars have generally attempted to bring back the modern language to the old Roman orthography given up long since so far as re- gards pronunciation. They write e. g. the vowel i (in their Russian alphabet /&) by five different signs, viz. d e / 6 v and the vowel e (i>) by a e i o u according to the supposed, yet often problematic, origin of these respective vowels from a latin a e i o u. Trying to introduce a doctrinal orthography of this kind into common writing, would soon cause general confusion. Wherever a rich and widely diffused literature does not protect an orthography differing from the pronunciation, a nation has no other corrective for its orthography than its innate feeling and its living pronunciation; you cannot force upon the people the result of learned researches, which they can neither ap- preciate nor understand. Others indeed go not so far; they follow the simpler principle of preserving for those sounds, for which the Latin alphabet has no particular letter, the Rus- sian signs; but thereby they surrender the homogeneity of writing and its evident advantages. We have above registered the Russian Alphabet, a mixed and an etymological Alphabet, and have selected the specimens accordingly. For linguistic works the notation of the accents of words would be valuable; we have noted them in a second specimen to show how easily they combine with our transcription. A remarkable peculiarity of the Rumanian language, espe- cially of the northern Daco -Rumanian dialect, is the for- mation (mentioned above p. 55) of an e vowel by the side of the /, which we find in the Slavonic languages. We know of no native Latin notation of these sounds, which could come under consideration; our transcription we have tried to justify at another place. * 1 Transactions of the Berlin Akadeniy. 1861. p. 151. 168 JAPHETIC LANGUAGES. The Rumanian language is distinguished by a multiplicity of vocalic combinations; it is therefore necessary to mark, systematically and for the eye, the difference of the numerous monosyllabic diphthongs from the dissyllabic combinations, con- taining the same elements as the former. The vowels i and , when following upon consonants at the end of words, almost lose the very body of sound, without disappearing entirely in pronunciation, being as it were only whispered. We consider these sounds as identical with the Old-Slovenian b and 7, to which we refer, holding the transcription there proposed I and M, which moreover is already in general use in the country itself. We retain also the analogous Rumanian notation of O the short unaccented i and u in the second place of diphthongs at, ei\ ii, #4, wf, n, aft , eu etc. (as in haicle, vomik, sw, dau), in order to distinguish these diphthongs from the dissyllabic combination of vowels ai, oi, ui (as in tain, voire, suit, audit). From the same reason we note the sernivocalic short and unaccented i, e and o in the first place of the diphthongs m ie 10 fu, ca oa with the same mark of brevity (as in farna, Mete, kforu, lumen, toate), to distinguish them from the dis- syllabic ia, ie, ea, oa (as in scrie, Galilea, ploa\ The notation c j s and z results from our general alphabet as well as / which latter has the advantage of being is use already in the country; the softened d is rendered d by some authors, and is said to exist in dialects with the pronunciation dz; but the common pronunciation no longer distinguishing d from z, we prefer to omit in transcription the difference. OLD ICELANDIC. 169 OLD ICELANDIC. a a e e o 06 i i y y u u ce 02 au ei ey ja jo jo ja ju a a e e o 06 i I u u u u e (or at) o (or 01) au ei eu id io io ia iu k g t d n h s z 0) r I P d p b m f ^ k g t . d P b ks m f r I Specimen. Hljods bid ek allar helgar kindir, meiri ok minni mogu Heimdallar: vildu at ek Valfodrs vel framtelja, fornspjoll fira pau er ek fremst um man. (Transcription.) Hlidds bid ek allar helgar kindir, meiri ok minni mogu Heimdallar: vildu at ek Valfodrs vel framtelia, fornspioll fir a Qau er ek fremst um man. Beginning of the Voluspa (Edda ed. Munch). Remarks. The old manuscripts differ from each other in their ortho- graphy, which has been reduced to fixed rules only by modern M 170 JAPHETIC LANGUAGES. scholars. We have given above the alphabet, almost as it is written by Rask. Scandinavian editors usually write ja, jo, JO) instead of the za, id, zb, of the manuscripts, even where these diphthongs have grown out of the vowel i by the in- fluence of a following a or u (nom. pi. skildir the shields, ace. pi. skioldii) nom. sg. skioldr for skioldur, gen. sg. skialdar}. Pre- mising that in all such cases the vowel i remained the pre- dominant, we propose to write these diphthongs respectively id, id, id, in order to distinguish them from ia ja, as in telia (to tell), telium (we tell), which might be written also telja, teljum. It seems to us not improbable, that the letters ce and a>, which proceed from a and o before an i of the next syllable, ought to be considered as real diphthongs, ai and 6i (like the Latin ae for at), although their actual pronunciation is that of e and />. With regard to the double e (e and though often confounded with ai and oi, yet in pronunciation differ from them slightly; we therefore write them ae and oe. The aspirated liquids n, m, r, are peculiar to the Welsh, and we prefer to put the sign of aspiration over the letter instead of placing the full h after it, the aspiration being more closely connected with the liquid letter than the composition with h would indicate. The fourth liquid letter, viz. /, has likewise, besides its simple pronunciation, an aspirated one, expressed by doubling it, II. As in this case the I becomes at the same time a palatal pronunciation , the middle of the tongue touching the palatal point on the palate, and the aspiration passing on both sides of the tongue over the eye-teeth, it might be still more exact to render this letter by T; we omit, however, the pa- latal line, to spare a diacritical sign which is not absolutely necessary, there being only one aspirated 1. The letter w is the vowel u before consonants, and the consonant w before vowels. The letter z occurs only in foreign words. HEBREW. 173 HEBREW. Former pronunciation. y x - n - i ' - h - - P - n - - 2 - I - D 3 - r - * y - s - O - - * - - s n T a D T -i S t d n s z 3 3 o p b m - - Later pronunciation. \ e n e a d * I i e e d o 6 u u u e y N - n ) h - P - n 3 - ( 1 - V V J\, V 3 3 - 3 " ^ ^r - i a - - * ^ - - 8 - - w \v - - - S S - n i 3 D 7 "i S < c? n s z - - n T - * - G d S 3 D 2 3 i |> 6 m / Specimen. ntrn vi^ni 2. : Y"i*^ nxi r I w iu-^ Tixn-nN n^n'Sx XI^T 4. :Tix~^n->i -iis vr Q^n'Sx -ION>T 3 SEMITIC LANGUAGES. av -iiS cpn'HN jop>] 5. : ^nn yen n-ixn i"a ovySg Vna^_ ovvH "ictf'l 6. : nn DV ip.b'vri anjpn?i nVS soja -siuJnVi rrn'Sx iri^T 7. J^oS o?o ^2 V"jao XT') o^-l Tna j^pn vv SITCD "luJx o^rn rai ypnS nnno ittJx nvsn 7^3 S^a^i ypnn'nx '^l~ .... . - - I .. - I. TT - - . v -: .--'.. .. ; - - - I. T T Genes. 1, 1 7. Former pronunciation. 1. Bi ra'asit barwa >ilghim 'it ha samaim ita >it ha >qr$. 2. Ua ha >arx haiatah tithu ua bithu, ua %usk >'al panai tahum ua TUY nlahim mara^apt >al panai ha maim. 3. Ua ia'amar nldhim: iahi } aur, ua iqhi >aur. 4. Ua iar>q >ilahim >it ha >aur ki taub; ua iabdil >ilahim bain ha >aur ua bain ha ^usk. 5. Ua iaqrq'q >ilahim la >aur iaum^ ua la %usk qara'a Iqilqh; ua iahi 'arb ua iahi buqr iaum } g^qd. 6. Ua iq'qmqr nlahim: iqhi rqqi> bi tquk ha mqim^ uihi mqbdil bain maim la maim. 7. Ua iq'qs >ilqhim } it ha raqi', ua iabdil bain ha maim >qsqr mi-ta^t la rqqi> ua bain ha maim >asar mi-'ql la rgqti; ua iahi kin. Later pronunciation. 1. Bere'siQ bdrd> >elohtm >eQ hassdmayim w&eti hd'dres. 2. Weh'ddres hdyeQdh Gohu tvdvohu, we^osej^ lal pene Qehom werux >clohim mera^dfe^ lal pene hammdyim. 3. Wayyo>mer >elohim: yeht >6r ivayehi >6r. 4. Wayyar* >elohim >eQhd>6r ki too; wayyavdel >elohim ben hd>6r liven lia^osej^. 5. Wayyiqrd' >elohtm ld'6r yorn^ wela^ose^ qdrd> Idildh; wayehl >erev ivayehi coqer yom >e%dd. 6. Wayyo>mer >eldhim : yeht rdqt> beQo^ hammdyim , wthi mavdfl ben mayim Idmdyim. 1. Wayya>as >l6him >eOhdrdqL' , wayyavdel ben hammayim >dser Idrdqt! uven Jtammayim 'user me'al Idrdqil; way eld Remarks. The Hebrew writing of our books and manuscripts is re- markable as being composed out of two apparently hetero- HEBREW. 175 geneous elements of a very different date and origin. The one element contains real characters, which, although their form was slightly altered after the exile, yet, as to their figure and alphabetic composition, belong no doubt to the oldest epochs of Hebrew, or perhaps even Semitic civilisation. The other element, the pointing, was added only about seven hundred years after Christ. By this addition no character of the old sacred text was altered ; it only served to fix the tra- ditional pronunciation. According to the opinion now ge- nerally received , the old pronunciation did not differ from the modern one. The ancients, it is supposed , wrote only the con- sonants , and left all the vowels to be supplied by the reader. Our opinion is, that a mere consonantal alphabet would presup- pose by far too abstract a phonic doctrine on the part of its inventors , and , even if such a systematic separation of the consonants had been possible, there would have been no reason for not inventing corresponding signs for the other separated element, viz. the vowels. Moreover the first and most neces- sary requisite of every writing is its intelligibility, which could not be attained without written signs for the principal and most expressive vowels. We therefore consider the Hebrew alphabet to have been, like all the other old Asiatic alphabets, essentially syllabic, i. e. representing by each character a full vocal syllable. With this syllabic character of the Hebrew letters it is not incompatible, that the inherent vowel is occa- sionally replaced by another pure vowel following, or eclipsed by the influence of the accent of another vowel in the same word. The two characters and i represented in the old Hebrew, as in the other Semitic -alphabets, the two vowels i and u, being the two most remote from the vowel a , and there- fore the most important in writing. They were primitive vowels in many words, in which they are still pronounced as such, as in the roots ]^ din, ~i>,uJ sur, or in the proper names VJVJf Siduti (Sidon) StfaaV) ]^jf Sinn (Siyyon) 2uov. 17G SEMITIC LANGUAGES. Placed, however, before other vowels in the beginning of a syllable they represented also the semivowels y and ?r, into which, according to their nature, they usually pass in the Greek and Latin, and almost all languages. Yet, this se- inivocalic power is only a derivative and secondary one, the primary being always and even in the pointed writing its vocalic power. One might be inclined to attribute also to the letter N a mere vocalic power in the primitive writing, con- sidering the frequent use of the N quiescens in the pointed writing, and comparing the letter yyy a in the analogous old Persian writing. But a further consideration disproves this opinion. It is as easy to prove that the letter ^yy in the name of yy yyy ^ f^> >T3: ^TT <<( Dariawus (U^TT. , Darius} has no con- sonantic element, as to prove that the letter N in TiNttJ Sand 2aovh (part, of bxu) #a>al) has no vocalic element. The Se- mitic N never takes a semivocalic power, nor is it an aspi- ration, but a slightly explosive consonant. The condition of the N quiescens is therefore totaly different from that of the or i quiescens. If the assertion is selfevident, that each system of phonetic writing must represent the most essential and prominent ele- ments of that language, for which it was originally intended, we must presume, that in the language which was first written with this Semitic alphabet , only three vowels were necessary to make it intelligible, , i and w, as in the old Egyptian, old Persian, Assyrian, and likewise, though with a separation of long and short vowels, in the Sanskrit, the Arabic and other languages. We may be almost sure, that also the Hebrew language of the time of Moses and David did not yet distin- guish all the vocal shades , represented in the pointed system of the 7 th century after Christ . but only the three principal vowels rt, i and u. But the language of our oldest Hebrew texts suppo- ses already the separation of long -and short vowels, which was indispensable for certain grammatical distinctions. Yet the old HEBREW. 177 writing knows only one i and one u and no separate a. We must therefore suppose either that this writing was invented for a condition of the Hebrew language before the time, to which our historical knowledge of it reaches, or that it was taken from another people and therefore, from the beginning, was defective in expressing the Hebrew vowels. This dis- crepancy between pronunciation and writing necessarily in- creased in the space of perhaps 2000 years, the former always advancing, the latter being stationary, till the invention of the vowel-points brought them both again together. The question now is, whether and how far our way of transcribing may be applied even to the oldest Hebrew text and its primitive pro- nunciation. We still have a certain number of unpointed texts of several other Semitic languages , the writing of which we may likewise suppose to be defective. The usefulness of a general and regular way of transcribing these texts as well as the oldest Hebrew, in a linguistical point of view, can scarcely be contested ; for a transcription ottering to the reader only unreadable consonants is certainly as inconvenient as a transcription with more or less arbitrary vowels, of which the genuine are not discernible from the hypothetical. We may therefore be justified in proposing a transcription according to our views especially in application to the oldest Hebrew, for the interpretation of which we possess more ample means. The principal thing to be done in this respect is, to separate in the language the ancient phonic elements from the later. A later origin and introduction we attribute to all the fri- cative modifications of consonants (indicated by the non inser- tion of the dayesh lenc), which are of mere phonic import without any grammatical signification. The age of the pro- nunciation indicated by the dayesh forte, which also serves to distinguish grammatical modifications, is more doubtful. Whilst in reference to the pointed writing we propose transcribing the dagesh forte ( as the Arabic tesdid) by 1 7 ( s SEMITIC LANGUAGES. duplication of the consonant, we should , as regards the old writing, prefer the transcription with the usual line of dupli- cation above the consonant. As to the vowels, the difficulty of a regular transcription is greater, owing to the defectiveness of the old writing. The gradual increase of vowels may be considered a general rule in most languages. The later origin and secondary value of the pointed vowels in Hebrew cannot be contested; this results also from the fact, that they no-where, like the old vowels > and i , distinguish different roots. The e and o vowels of the pointed system may still be traced back to their respec- tive primary vowels, and replaced by them without altering or obscuring the language itself. But it has already been ob- served, that the Hebrew language in all parts of the Old Testa- ment supposes the distinction of a short i and u as well as of a long a, although they are not separately expressed. We must in consequence supply these vowels, where they are wanted. It is another imperfection of the old writing, that it does not indicate, whether the inherent vowel a or no vowel at all is to be pronounced. We find the same defect in the Old Egyp- tian and Habessynian writing, as well as in several modern Indian writing systems in which no (Sanskritic) virama is used, and the nearest analogy occurs in the old Persian. We are thus at a loss to know, except by inference from the pointed writing , or from grammatical laws , or from foreign transcrip- tion, whether we have to read as ', or as z, or as the semi- vowel i (y), and i as u , or as au , or as the semivowel u (10). In fact was pronounced as * in nuhSx 'Alisah ('Elisdh) 'Ekioa, p,>s Stun (Sit/yon) Si&\ as at in ]D"p Qainan (Qendn) KaLvar, a-nso Masrcdm (Misrayim) MeoQa'i'n; as / (y) in p-^ Iqrdgn, (Yarden) '/opJayoc, };no Mgdian (Midi/an) Madiu^i; and i as u in -fitzjx 'Asur ('Assiir) ^AaooiQ', iv> Lud (Liid) ^lovd; as au in iit:;y ,'Asau (lEldu) 'Haav, ^y-\ Ra>au (Relit) 'Pcr/aii; as u (w) in rn^-'J Ninatiih (NinewehJ Nivevi, ]v latian (Ydwdn) 3 Ia)vdv, iV") Liui (Lewi) HEBREW. 179 It is evident, therefore, that to make from any unpointed Se- mitic writing a proper and readable transcription, a previous knowledge of the language is indispensable. Modern researches have taught, that, among the pointed vowels, Jong (a) takes its rise mostly in a, short (o) in w, long (a) always in a, short (e) mostly in i, (e) in i or ai, (o) in u or aw, or in a degenerated a; ^-7- and i answer to the old heavy vowels i and u, > and i to old ai and au\ (t) and (w), as well as sewti mobile take their rise often in old a. These general facts afford a general rule for transcription, without however superseding the special consideration of each particular case. Our specimen will best show, what is here meant. But it is important, that the reader should always be able to distinguish immediately the vowels added by the learned transcriber from those written in the original text. The former are to be indicated for this purpose by the little circle placed underneath. In transcribing the pointed writing, it is not less de- sirable to facilitate to the reader as much as possible the dis- tinction between the old and the later writing. Apparently this might be attained by treating, in harmony with the re- ceived view, all old characters, including and ", , as pure consonants, and the new points as the only vowels, e. g. ^3 kiy, yj5"i roqiy? , I'l^n hd-owr, rnt3 towv , pin uwveyn. The reader, then, would in the transcription have simply to re- gard all consonants as representing the old writing, all vowels as representing the points. This mode of transcription, however, would not only offer to the eye a form of writing incapable of being read (nobody being able to pronounce combinations such as -iyl, -owv) and therefore offensive in itself, but it would not even correspond to the intentions of the authors of the pointed system , who, according to our view, did not mean as little indeed as the old writers themselves to consider the and i as consonants, which they never could have been. SEMITIC LANGUAGES. The notation - 1 , *,, * and '*, , differed from long , --, and , no longer in sound but only in origin. In tran- scription, however, the distinction is necessary; we, therefore, represent the combination of the ancient character and the points by *, u, <', o, the simple vowels if long by 7, u, e, , and i , and % and i). There is no distinction in the Hebrew writing between the sewd mobile , which we write e, and the sewd quiescens, which we omit altogether, because it was not pronounced at all. In our specimen we have, for the sake of comparison with the old writing, retained the sign of x ' in the beginning of words; being, however, there necessarily understood, it may be omitted as in the Arabic writing. The paQa% furtivum, as indicating only the natural phonic transition from the vowels i, ) is undoubted; we there- fore write t and s; to regard the former as a medial on ac- count of the Arabic _b = d (see below) there is no reason. The ancient sound xr? s was later split up into a deeper sound a) and a clearer one tr, which latter approached very near to o s, and therefore must have been spoken much like the Polish (see above p. 161). The signs ^, /, ^, d, require no farther explanation. The accents of words, where it ap- pears useful to mark them , may easily be added. 182 SEMITIC LANGUAGES. SYRIAN. a e - - u el 6 u ^ 1 - 01 _ > > - h - - *~ - f - x - - *. _ - bx 9>y - X - y - \^ t - * - \^ >m& 1 5 \ W d,d n s z r 7 > - - P,f ^ m - w Specimen. o ,F o .0 7 o .0 o y .0 = * . _, . . o 2. .\\ A-.0 . t>., V * >- > > o .01000 .o v .0 > . < .o looio . . < jotnJ . o o. . o > V .o or o . | TL_^ y-oo . . JBQ*J ^.o V . o 7,t> 7 2. |C(310 U^ci JC010 . Gen. 1, 1 -6. 1. BerlsW bero> } aldhd } yoQ semayo> iveyo$ >arlo\ 2. War Id' hewo& tuh ivebuh ive^esuj^d' >al >apai tehumd*, weru%eh ddloho* mera^efd' lal } apai mayo'. 3. Wetnar } aldhd': nehwe> nuliro* icahewd' nuhrd>. 4. Waj(ezo> >aloh6> lenuhro* desaplr waferas >dldho> beQ nuhro } h^esuj^o: 5. Waqerd> >aldho> lenuhro> >im6md> ^d' qero> lilyo> wahewo* rams6> icaheivd' ,mfro> yaurnd' jad. SYRIAN. 183 Remarks. Besides the old points of the Syrian writing, we find in the manuscripts, since the 7 th century, also the Greek vowels added to the Syrian letters in small figures, as we have exhibited them above. They represent the five vowels a, -* (as vowel-sign) or ] quiescens, short, when placed without either, and w, always united with o (as vowel-sign), is sometimes long and sometimes short. The short * e re- presents in the language at the same time the short i; we follow, however, in this respect the indigenous writing, transcribing it by e. The want of an indication where an audible sewd is to be pronounced and where the consonant alone without any vowel is spoken, exists also in the Syrian writing. Tradition alone supplies this imperfection. In the modern pronunciation the feeble consonantal sound of j 'is often dropped entirely, although the letter is still written, and the appertaining vowel is then immediately connected with the consonant preceding the | : e. g. daloho* , is pronounced instead of de>alohd'. In such cases we propose to keep, for the sake of etymology, the sign ', but to place it over the vowel connected with it, writing e. g. daloko\ The same letters represent in Syrian writing the ex- plosive and their corresponding fricative letters. Sometimes the former are indicated by a dot over them , the latter by a dot beneath ; but usually those dots are omitted. In every case the transcription has to follow the pronunciation. 184 SEMITIC LANGUAGES. ARABIC. ! -'- - t * z* <3~ ^ ^ t t - - " "~ " I ~~ 15 O O r O LT : O O J V s> V.- ^ _ -5 ad ,' ' _ K h - - - '* A 1 ^ - x r - - w w - ^Q - 8 - J ai ?< - d - ? - $ t d n 5 iS- \J (j r - b m f ? Specimen. 1. -> - o -. o-o - - ","' * J QJ^J QjAJU, 3. ,] 5. 5UJ s ARABIC. 185 G < ^ ~ ?. -x*.*.^ r V^ 1 - - y o y ~ *, f - I *j .eol ^-Jb 5 x lib I ul , j,o, ,, i ,^ , , SKJ , , w Qj^AJf., * U! Suratu 'I baqarati. Bismi 'ttdhi 'l-rafimdni >l-ralilmi. 1. ^/dlika 'I kitdbu. Id raiba filii liuddn lil muttaqina, 2. alla- dtna yu'mimma bi 'I yaibi, wa yuqimuna 'l-saldta, wa mimmd razaqndhum yunfiquna, 3. wa 'lladma yu>minuna bimd unzila ilaika wamd unzila min qablika, wa bi 'I d%irati hum yuqinuna. 4. Uld'ika 'ala hudan min rabbihim, wa uld'ika humu 'I mufliKuna. 5. Inna 'lladlna kafaru sawd'un lalaihim a>andartahum am lam tundirhum; Id yunninuna. 6. J%atama 'lldhu 'aid qulubihim, wa 'aid samtihini) wa laid absdrihim yisdwatun, wa lahum 'addbun ladlmun. 7. Wa mina 'l-ndsi man yaqulu : dmannd bi 'lldhi wa bi '/ yaumi 'I d^irij wa md hum bimu^minma, 8. yu%ddi>una 'lldha wa 'lladina dmanu, wa md ya%da>una illd anfusahum, ira md yas'uruna, Remarks. Our only object here is the written language and its pro- nunciation, as it has been fixed very accurately for the Koran and faithfully handed down. This pronunciation is still fol- lowed by the Readers of the Koran, and is in use in a num- ber of Bedouin tribes. The manifold deviations of the mo- dern dialects have never been admitted in writing; and the pronunciation of the Koran is still everywhere understood and regarded as the best. When, however, in particular cases it is desirable to render a different dialectic pronuncia- N 18(3 SEMITIC LANGUAGES. tion by Roman characters, then it is necessary to determine first the individual pronunciation as is done by Eli Smith in his appendix to Robinson's Palestine and to transcribe it accordingly. Arabic writing, like that pronunciation which the Or- thoepists teach, distinguishes only three vowels, a, t, w, which may be long or short and two diphthongs, ai and ait. The latter have sometimes lost their second element in pronunciation, though in writing it is retained; f. i. .JLc, *j^ >alai, salauta, are pronounced ,'ala, salata. It is the same as with the i subscriptum in the Greek, which is still written but not pronounced. This analogy will justify our writing in such cases even in Arabic for etymological reasons an i or u sub- scriptum: tola , salata. In the Article al, the vowel, when fol- lowing upon a vowel, is passed over, which, like other eli- sions, we note in the European manner by an apostrophe. The sign Madda is either a mark of the length of the vowel, and then to be rendered accordingly ; or it represents at the same time * Hamza, in which case this will have to be added: ^.lli sana>dnu. The perpendicular FatKa, also, is mostly only the sign of a long vowel. Hamza (see above p. 68) must necessarily be transcribed only in the middle of words; in the beginning its omission produces no ambiguity; we write therefore yu-minuna, but alladlna instead of } alladma. Our sign of / for p shows its phonetic relation to the weaker * >, and has moreover the practical advantage of being convenient before a capital letter in the beginning of names: lAli, lAkka. Of ^ = K and o = q we have spoken above (p. 69), as also of the linguals or gut- turo-dentals -b, -b, LN 3 ) O^j d, $, , z. With respect to the accu- rate pronunciation of the latter sounds, especially the medial ^3 d, which hitherto by European scholars has been taken for a tenuis and therefore connected with the basis t (th : t, 't, t), ARABIC. 187 we refer our readers to a special treatise by the author, in which these questions are discussed at length. l It may be doubted whether - is to be rendered by g or /, the former representing the older and purer, the latter the more general modern pronunciation; we prefer the latter, for which we find an additional reason in the fact, that the foreign languages which have adopted this letter with the whole Arabic alphabet or in Arabic words, almost universally pronounce it/. The Tesdid - is a sign of reduplication of the consonant over which it is placed; in transcribing, the consonant itself is to be re- peated instead. The J asm * is of importance only in syllabic writing, showing the absence of a vowel; in transcription it requires no distinctive mark. In all cases of assimilation of consonants, the etymological point of view must prevail in transcription as it does in the native writing, the assimilation being generally self evident even for us. We write therefore radadtu, a^adti^ not radattu, a^atti. The article al ought to be always kept separate , according to the custom in all European languages; only, before the solar letters its assimilation, pro- ducing a closer connexion in sound with the following noun, may be marked by a line of connection. We therefore write al kitdbu, but al-raKlm, al-ndsu^ [to show the pronunciation (irraKtm, annasu. The almost general extinction of the sound of h at the end of words is no reason why it should not be noted there , as in the midst of words, by 7t. In the same manner, at the end is to be transcribed h, when pronoun- ced so; but when spoken as , is to be written t. There is no difficulty in making this distinction, as the different pronunciation of is determined by a definite rule. 1 On the sounds of the Arabic language and their transcription. Publi- cations of the Berlin Academy 1861. N2 188 SEMITIC LANGUAGES. GE',EZ (ETHIOPIC). O ft - - eh U >a } a - Ka ha flo T* A, pa ba p'a ma fa wa Classes a e 1234567. a In Ml- Ml, *n n, Yl *P e 6 ka ku kl kd ke ke ko e I vu ai ait *M1/. : ifbC : 'J'J : 6A1- : Specimeu. : A-AC : o :: Gen. 1, 1 5. 1. Bak*adanil gabera >egzi>abeKer samdya wa medra. 2. Wa medrsa ntdstarn wankonat deliita wat'elmat mal'elta k'aldi wa manfasa >egzl>abelier yet'elel malldta mat. >. Wa yebe >ea;i } abeKer layekun berhdn wa kona berhdn. 4. Wa T&eyo >egzvabeKer laberhan kama sandt wa falat'a >egzl>abeKcr nui-l^iln lii-rltdn ir >n<~rkala pelmat. 5. Wa samayo >egzvabelier laberKdn 'data wa la f' el mat lelita wa kona lellta wa f'abh'a wa kona maldlta >ah'a - ch "5l U > [f t n TT i - UJ i 5 ft H , , K K h k g F - X k g $ - X t } n s z p b P' m f - Specimen. :: IMTdT : (D yt r I Isenberg, Amharic Gramm. p. 14. BaKawarydt zaman babeta Krestlydn endeh ydlac andenat na- baracebdt huldcau ande segd andlt nqfsem 'eskihonu deras Krestlydndtem hulu ba Krestos kato altalayum huldcau ya 'Adam lejoc enda nabaru basegd, huldcaum lardsdcau ydla Krestos yaVafu ^dt'e>dn enda nabaru >endehum bahdimdnot huldcau bdnde Krestos ddnu, 192 SEMITIC LANGUAGES. Remarks. The Amharic alphabet differs from the Ge>ez only by the accession of several new sounds, which by the inhabitants are called jjslamitic", being indeed for the most part Arabic sounds. To these belong especially the palatal letters c, j, ?i, a, , c\ all derived as to form, from the corresponding dentals. The sibilant m, which originally, as we have observed in reference to the Ge'ez, seems to have been sounded as , has been divided, as in Hebrew, into two sounds; the letter w being confined to the sharper sound s, which must have approached the Po- lish #, and H being added to represent the deeper sound s. The letter rf, corresponding in Ge'ez with the Arabic -^ K, had lost in Amharic its strong breath, being almost weakened to the sound of the simple U, h. In a subsequent time, there- fore, when Arabic words, containing the letter ~, A', were in- troduced into Amharic, a new Amharic letter was wanted to express this strong spirans, and ' I h was added for this pur- pose. We distinguish in consequence this new letter from the older one only by a diacritical point, placed under it, K. The vocalisation is the same as in the Ge'ez. OLD EGYPTIAN. 193 OLD EGYPTIAN. 'A jra &7 }>05S=3 ^li .C=^>5 \->& I/ 1 AAAAAA 5 ^J lo i*- u < 57 / J K ^ RL c 194 HAM1TIC LANGUAGES. J: Specimen. \\ m 1 x\ o ^^ n T S /W\A\\ /I o . 1^. tg). Takrut (nS^n , TaxtAw^g). Kanbut, Kambatt (Kaf-ifiuarjc). ffiiarsa (cuneif. %sayarsa, ). Klaupatra. Pian% (Oivayj^. ahu (Cefil). Jam (SOflJ). a/ (5.q, AA^). wia (AIHJ, 1*6). j5ar (Sya). rceAm' (negros and) nehesitu (negres- ses ). iuma (6IOW, IOW). kahu (K3g). te ((^OflT), n^fe (AHOK). %r/i (^OOpft). ska (CKU). * (MATOJ). A|6w (&lfkm, ^ASlOTJ). %i (8 AT ). (COTTIT). Remarks. The hieroglyphic writing was at all times essentially an ideo- graphic writing, in which every sign expresses a whole idea and its corresponding words. It is true that we find from OLD EGYPTIAN. 195 the oldest times a certain number of pure phonetic signs inter- mingled with them, but of secondary nature. Those are not intended to supplant the ideographic signs, but to suggest the proper words for them, to supply them with grammatical forms, and to write foreign names. Besides those two classes of signs, there are others of an intermediate nature. It is evi- dent that our transcription could not reproduce in any way this complicated system of writing ; we have only to deal with its phonetic part and to determine the different sounds of the ancient language. This task has been accomplished once al- ready by the Egyptians themselves, viz. by the Christian Egyp- tians, when they changed their indigenous writing for the Greek alphabet, adding to it six new characters for the same number of sounds peculiar to the Egyptian and unknown to the Greek language. The comparison of the old Egyptian with the Demotic and Coptic writing, and the examination of proper names transcribed anciently from hieroglyphics into foreign languages and vice versa, are the principal means to determine the old Egyptian pronunciation. They are per- haps not quite sufficient to remove all the doubts which are still entertained amongst Egyptian scholars, yet I may refer to what I have said on this point in another place. 1 The pronunciation of the Egyptian sounds, as given above in our transcription , is what the later Egyptians themselves gave to the hieroglyphic signs. They may have erred in some points, but those points are doubtful also for us, and it would therefore be advisable not to decide any thing in advance of this later pronunciation , before those doubts have been tho- roughly removed. We add here only a few remarks on spe- cial points. The short vowels rarely were written, but were regarded as conveyed in the respective consonants. The three vowel 1 Konigsbuch p. 169 sqq. 196 JIAMITIC LANGUAGES. signs \\ ; g^,, - o, which are imperfectly distinguished in the corresponding Demotic, are not distinguished at all in the Coptic words, and in the transcription of Greek and Roman names. All three are rendered in the Coptic and in these transcriptions by the vowel a, with which e and o are regarded as identical. In the hieroglyphic writing, however, they very rarely interchange with one another. There q is mostly found in the beginnig of words, j|^ mostly in the middle or on the end of words. The sign ^ a seems to have expressed originally long a, which not seldom passed into o, as from the Demotic sign 2*T = o, representing the hieroglyphic ^*^ n , as well as from the Coptic , might be inferred. To the same sign we find in Hebrew y often corresponding . and this letter might then also indicate rather the lengthening of the pre- ceding vowel, not the consonantal sound /, which seems to have been always unknown, as well as the sound of * ', in the Egyptian language. The sign \\ appears first in the end of the Old reign, and in the perpendicular form \\. This may have been an abbreviation of 1|1), but seems to be used rather for the short i, in contradistinction to (jl), which represents long i or e. Between jj> and the later introduced sign Q, both standing for u, there is no difference to be observed. More seldom is f\ , which perhaps was meant for long u or o. Almost all the consonants seem to contain originally a cer- tain implied vowel, or at least to unite more easily with cer- tain vowels then with others. We observe principally two classes of consonants in this respect, the one of which prefers to be followed by the vowel a, the other by u or i. In this, too, we find the reason, why in later times the consonantal value of certain signs has changed through the influence of the following vowel. Especially the closed vowels u and i caused not seldom, as we know also from other languages, an assibilation or softening of the preceding consonant. Accordingly, we find A mostly united with a, ^z^s more OLD EGYPTIAN. 197 often with u or i. To both corresponds k in Coptic and in Greek names, and r> in Hebrew names. It is true that we find not seldom Hebrew p q for hieroglyphic A\ but this is as little able to prove, that the Egyptians had the merely Semitic sound of p, as the regular Arabic writing *J for t re- place one another often, though not in all cases. The fol- lowing characters <^ == ^ 7 ~" \ , J,, which occasionally change with one another and sometimes also with the preceding cha- racters, show a tendency to duplication, tt. Afterwards, they are inclined to pass into the sound of the Coptic Z. = c, which Coptic sign indeed seems to come from the demotic form of hiero- glyphic J^ The original hieroglyphic pronunciation t was howe- ver known even as late as in Roman times. The same character is found sometimes as substitute for the Hebrew x, as in I o=L) ^ ar ?5 which has been compared with the Hebrew s or -\a .SW, Tyrus; but in these cases we must remember 198 HAMITIC LANGUAGES. that by the side of the Hebrew x there is commonly found, as an older sound, the Aramaic -o, as in >nyo, Turd>, from which also the Greek form TVQLOI; is taken. The choise of the hieroglyphic ^ was due perhaps rather to the decided emphatic pronunciation of "O in such cases then to the assibilitated pro- nunciation of x. The medial sound of d in foreign names was sometimes indicated in hieroglyphics by ^, Q| and in De- motic by Zl i. e. nt; this proves sufficiently that the Egyp- tian language did not use the simple sound of d. fij^ p con- nects itself readily with a, D with u or i. The identity of the consonantal value in both results from the fact, that ori- ginally D was added to the less frequent ^ in order to in- dicate its phonetic value. There is no doubt about the sound of 6, which before u and i is mostly written J, before a, often ^J. The sounds of m and n have never changed by the influence of the following vowel; o = nu dropps its inherent u only in later times. The characters fD and j[ , although they rarely interchange, seem to be distinguished chiefly by the subjoined or inherent vowel, the former preferring a, the latter u or i. The closed pronunciation of u and i caused naturally a strengthening of the breath in \. The same case happens with \ and , the former of which was originally explained by the latter (as ^ by D, 55* an ^ ^ by H or P, 5j* by 6 or J), but differed afterwards from it by inclining rather to a, wilst inclined to u or i. It is for this reason that in later times passed more frequently from the pronuncia- tion of % into that of is. Amongst the characters representing s, 4 unites constantly with u. Bill and C3C3 represent both s, but the former prefers a, the latter u or i. There was no distinction between the two sounds r and I in the old sacred language ; it belongs rather to the later popular dialects. The lion JBa> as r, and more frequently as I occurs principally in foreign names since the time of the later dynasties. If, now, any one wishes to denote in the transcription this OLD EGYPTIAN. 199 system of vocalisation , which might be compared with similar facts in the Old Persian and Old Slovenian languages and which was more in use at certain times then at others, he ought to chose one and the same diacritical sign, as, for in- stance, a point underneath, so as to distinguish from the others those consonantal characters which unite preferably with u or i, and at the same time to signalize their frequent change in later pronunciation through the influence of these vowels. The sound of Ij might also be provisionally denoted by a with a point above. But, if we consider, that this distinction of the two classes of signs did not indicate originally a phonetic difference of consonantal value, that there was never a syste- matic uniformity in regard to their use, and that in any case several points of this question remain still unexplained, we cannot attach a great value to the introduction of this diacritical point in our transcription. Scientific exactness, however, demands, that all the vowels, added by the tran- scriber conjecturally, are to be signalized as such to the reader. We write therefore, as in other languages, for supposed vo- wels 0, |, u and if the vowel is quite uncertain, e. 200 HAMITIC LANGUAGES. COPTIC. 8 A K (U) - * (X) e H o SB 6 - - cy j (T) OT T & n c p * w M 61 Ol AT t'T GOT ~ n & M ? J (*) (E) (t) fc) (t) e A a & # - X M e oo c - s i (u) u t d N s r ^ A c _ _ ai ei 01 an eu ou p b m / ph ks ps C ti Specimen. IHC &e eT ATMAcq ^en Sno^eeu fiTe f ^en ru e^ooT HTG Apyj^HC noTpo gunne JC gs.n WAIJOC ATI cSo^ c& neieST e J^HM eT^fw mioc. 2. ZC ^ OW5H cH T ATM^.C ROTO flTe fll e neqcioT c<^ neieST OTOg, ^.rii COPTIC. 201 lesus de et aumasf %en Bethleem ente ti ludea %en ni ehou ente Erodes puro lieppe is han magos aid ebol peiebt e Jerusalem euco emmos. 2. Ce of thon phe et aumasf puro ente ni ludai annau gar e pefsiu sa peiebt uok ani ce entenuost emmof. Remarks. The vowel H corresponds, like the hieroglyphic [j(j, to our long -i or e, and CW to u or o. The compound OT for the simple vowel u is taken from the Greek ; and when this vowel forms the second part of a diphthong, it is, as in the Greek, rendered only by T; we have therefore to pronounce &T and 6T as au and eu; but the diphthong OM, which in the common Greek is not in use, was therefore written GOT, in order to distinguish it from the simple OT = u. There are on the other hand three diphthongic combinations with I, viz. $j GIj Ol = ', ei, oi. It is probable that we have also to re- gard the combinations Hf , WM and HT (HOT), OOT (WOT) as diphthongs: a, o*, en, 6u; and that even di and aw, as well as the single a, existed in the language, without being dis- tinguished in writing , as and a were not distinguished in the Greek alphabet. The single T occurs in a few Egyptian words, changing with J, e or H ; it may then be expressed by u. The marks _!_, _n_, _1_ over consonants indicate usually the indistinct vowel, which we write e. The peculiar Egyptian sounds 6 and X show even by their form that the former springs from &, the latter from t (see above). It seems therefore that their pronunciation approached to the Polish sounds, written in our transcription by c and c. They were most frequently confounded with one another and passed afterwards into the softer sounds j and j , as K , T , II were pronounced in later times #, J, b. The letters I* and & occur only exceptionally in Egyptian words for K and T ; but they were usually pre- served where they occur in Greek words. The Greek letters 202 HAMITIC LANGUAGES. DQ, O, <$> had by no means in the Coptic alphabet the pronunciation of onr ^, 0, /, but represented the aspirated sounds kh, th, ph; they were almost peculiar to the Mem- phitic dialect and are often resolved into K , T& , JT& ; while the Theban dialect usually keeps the tenues K, T ? IT instead. The letter ) , our % , belongs only to the Memphitic dialect and is replaced in the Theban dialect by & li. The letters ^ and <" occur regularly only in Greek words; the two for- mer are sometimes met with in Egyptian words instead of the ordinary combinations KC and ITC; and <^ seems to have been pronounced like C. The sign T represents the syllable TJ, which latter is often written instead. BEJA (BISARI, ETHIOPIAN). a e o $ u a m y h k g - k g n - - 3 - ' s y t d n s r I - b m f w Specimen. Bdbu iydne hoi, ant ogauib gtbhe. Am asogimek liok^ baruk inkertinia hcb. Batuk otu ddi'tui; nauatrit kitkai; nauatrit tike- tiek, areyi hoki. Anikakan, baruh dine. Ura otak tu sdgal w- hat'dit tdkat ehe, shule jilldida. BEJA. 203 Remarks. The name of the Beja is well known to the Arabic writers of the middle age, and designates still the different tribes of the Bishari , Hadenduwa and other descendents of the Blem- myes of Roman times and of those Ethiopians, whose chief town Herodotus calls Meroe. They dwell in the country between Egypt and Habesh , east of the Nile. The dis- tinction between long and short vowels in their language is not well developed; they are all rather long; which is more perceptible, when the accent of the word falls upon them. It is even doubtful, if the combinations ai } ei, o, aw, are to be taken as diphthongs or as two syllables. We prefer there- fore to leave all the vowels without indication of length ex- cept where sometimes a decidedly short i or u appears, written by us i and u, and to use more frequently the accent, which falls for the most part on the last or on the penultimate syl- lable. It is remarkable, that we meet also in the Beja the peculiar class of deep gutturals, which we found in the Abyssi- nian language approaching to the compound sounds of kw, gw, and which we write also here k and g. On the other hand, we observe the cerebrals t and d, specially found in India, and resembling in the Beja sometimes a combination of tr and dr. There is no p , as in the Arabic , and the letter j is very rare, and seems to be taken from the Arabic, as it mostly appears in words taken originally from that language. 02 204 IIAMITIC LANGUAGES. GAL LA. Tutschek, Grammar 1845. " a d)d,d w - - h - ' e k 9 c,q - ch - e tch(tzfz) dj(dz^z) tsh - i i t dy - v - - y " - - - - h - e k 9 k' - x - e g 1 c' - s i i t' d 1 - n - - y o o - d - - z - i u u t d t' n B Z r I - b P' m f - w Specimen. Wak'ayo la/tana, goftako: ati na guba teza, ani zi dalan taa. Ho haman nati dvfe^ aka mukni adu narra k^abu^ ati liama nati fcabi, goftako^ gadiza na tai. Zi ivamadetam o/a, zi wamadetani bida', batinana batte, nan clabin, bae zin d zababi na olc'i. (Tutschek, Gramm. p. 84.) Remarks. The difference of long and short vowels is not clearly de- veloped, except in a and a; the latter of which Tutschek writes also d or a according as it is contracted from awa or TAMASEQ. 205 aya. The tone of all the vowels at the end of words is some- times almost entirely lost, in which case he writes them in a smaller form. We prefer the same indication which we have already employed in other languages for the same purpose, viz. the sign of brevity. The consonants in the third column are exactly the same as the corresponding consonants of the Abyssinian language, according to the pronunciation of a Galla man, whom I met in the Sudan; and it is, therefore, ques- tionable, whether this particular kind of tenues belongs ori- ginally to the one or to the other of these languages. The letter f of Tutschek seems to be our letter t', inclining to the Polish c. We prefer the transcription by #', corresponding to the soft sound f/', Tutschek's dy. The sound of his n comes nearest to our n. I should not wonder, if also the sound t'' existed in the Galla, although it has not yet been noticed. The letter, which Tutschek writes d' is almost the same as the d of the Beja, except that it is perhaps still softer and seems to contain more of n than of r in its pronunciation. If there exists any corresponding t, it is at least very rare; but the second /, which approaches very near to the Polish , be- longs to the same class; we write it therefore /. TAMASEQ. ' 1 "l" .'.J 1 1 - XI 1 a a - -f n,u,A 1 - CD, D a 0,O i i)JL a - - - h - it - 9 - - r I k g n x - - 9 n s z y t d - (s) z t d n s z r I - b m f - w 206 HAMITIC LANGUAGES. Sped men. As kelad .yawalen a^j^uten^ aliulay iycn dey emir en tafsit kelad ubelbel; ii_kr akal s takat. I-yela A- azlbard; yusa d yur es, inna a: ewod, kai ahulay , ma full teged takat tare)'? Inna s ahulay: elkamey ulll, tarunet s takat. (Hanoteau, Gramm. Taui. p. 135.) Remarks. The Libyc branch of the Haruitic nations is still spread over a large part of northern Africa. We know best at pre- sent the languages of the Kabyles and of the Imusay or Tuareg by the two respective grammars of Hanoteau. The latter language, called Tamaseq or TamaZeyt (with the addition of the feminine t to the gentile name) is more free from Arabic influence than the former. A peculiar alphabet of old Libyc origin is very generally in use, though without any literary application in books. These letters, which are first printed in the grammar of Hanoteau, are called Tifinay (plur. of Ta- fineq or Tajineyt). The three vowel signs, the first of which might rather correspond to the Arabic \ than to the Latin a, and the two latter of which are used also for the semivowels w and y, are of so rare and indefinite use owing probably to the influence of the Arabic writing system that our tran- scription can only render the living pronunciation, not the Tifinay writing. "We write instead of Hanoteau's #, kh, ch, j\ n, ou, according to our system (l, %> , , h, w. The letter ii, being used only before gutturals, may give up its diacritical point. The letter .s and , according to Hanoteau , perhaps also the letter ^, occur only in Arabic^words. The letter t is often replaced by or confounded with d. The description of the letter, which we write 0, is not sufficient to remove all doubts about its pronunciation. The sound of the letter t being that of the Arabic should be rendered by /, and we must protest expressly against the transcription r of Hanoteau, rh of HAUSA. 207 Dr. H. Barth, and any other which takes r as its basis. It would cause a great confusion, if the very common blunder of Euro- pean travellers, to whom the sound and etymological value of the Semitic is unknown, and who invented the new French word razzia i. e. sjic, yazzwah (impetus, incursio), found its way into linguistic science. In Tamaseq (contracted from Tamaseyf), where y and q constantly interchange, it is even easier than in other languages, to see that y has nothing whatever to do with r. e a d e e HAUSA. k g n h - m VX" *^\ 7 * / f' V \ ts(c) dzQ) - 8 Z y g t d n S Z r I o 6 ts dz - u u p b m f - w Specimen. Ydo muka taffi farauta^ mu uku. Yaro da obansa suna da bindiya, ni ina da dusi tsikin alsifu. Da muka taffi tsikin ddsi babu karre tare da mu; amma'oban yaro ya sanni enda ndma si ke. (Schon, Hausa Gramm. p. 165.) Remarks. We give the alphabet as it has been reduced to our prin- ciples already by Schon in his Grammar of the Hausa language. 1862." It seems that in the Hausa no monosyllabic diphthongs are used, all vowels being pronounced separately. The combination ts originates partly in the contraction of t and s (cf. itasi or itsi, tree). In such cases at least the tran- scription of ts, dz^ seems preferable to that of c and j. In the specimen, not all the existing distinctions of letters are exhibited. 208 HAM1TIC LANGUAGES. NAMA (NAMAQUA). a a e e o 6 (o) i I u u dei etc. au ou ai ei oi ui q - - h k g n X ' - t d n a g r - b m - w Clicks: pal. /', cer. /, dent. /, lat. // Specimen. iGuro mts. Tita ge iquta sa Zui-ngoata; ikara Zui-iigoaz ge ti ei-id u-hd tite. Taree? Ama Zui-agoabada ge nl hoa %un famei /aw, mam zl nl igom. iGam-iiei mis. iQub sa Zui-iigoab tonsaz ge iiause fgei ihuru tite. Taree? Zui-ugoabada ge nl tau zl nl mam eda iieib ions dawa td ia%are, nu, igai-dl, ihomi zl ga^a- ma; eda ueisa hoa ihdgu ma fgei, igore, gare zt gan-gan. (Wallinann, Nama Grammar p. 83.) Remarks. Wallmann in his w Formenlehre der Namaquasprache. 1857." has already introduced the Standard alphabet. He attributes to the five vowels a pure and an imperfect pronunciation, which latter he writes consequently a, e, *, o, u. The existence of one aspirate k = kh would be very strange. Tin d all in his ^Grammar and Vocabulary of the Naniaqua-Hottentot language" p. 15. compares this sound expressly with the q of our Standard Alphabet. We prefer therefore this writing, whilst we take his gh for our %. The characters /, Z, y are introduced by him for foreign names. On the click-sounds see above p. 81. We find in the Koran a dialect of the Hottentot language, according to Appleyard (The Kafir language p. 17, sqq.), the same clicks as in the Nama dialect, and besides the letters c, , and y. With regard to the gutturals, Appleyard gives three fricatives, without a sufficient description however. He says : n ch resembles the Dutch , - and -^, as well as the letters and 4* , and - , differed only in form not in sound, and were consequently also in transcription not distinguished from one another. But the circumstance that the first letters k, g, j%, |, d, occur exclu- sively before the guttural vowels a, 0, w, and the second letters lc, h s ? '!/> s "> it $ are, like the palatal or dento- palatal letters, not combined with u, nor s, y, t, d, like the gut- turals, with 1. The letters K/i, yh, %h, ch, jh, t, d, z are employed only in Chinese words. The forms of the Manju letters show, that only five vowels and thirteen consonants were originally distinguished, the others, which have only a secondary form, having as we may presume, arisen as sounds only later. The letters u, 0*) W r ^ 3 a o u i e o u ^ (0 - (/t) /i /r 9 (?) 2 21 G TATARIC LANGUAGES. Specimen. . 3 Q^-4-j A ^ i 00 .( Kaseinbeg, Turkish Gramrn., transl. by Zenker p. 17. Ei ogul spile bilmis ol Ri Kaq sobKdnahu ve ta'dla dsikdr ve nihdnde ve yerde ve kokde ve bu fehdnda ve ol jehdnda ,'aqlile idrdR olunur; amnia Kendinin ddtt serlfi tasavvuri 'aqtldan mu- nezzehdir. Amma eger dilersin-Tci allah ta'dlani bilesin, evvel Kendi Jcendini-bil, ve Kendi Halindan %aberdar ol', zird her Kim Kendini bildi Kaq sobKdnahu ve ta'dlani bildi: bu spzden maqsud ki sen bilinmiSsin ve ol biliji-dir, ya>ni sen naqissin, ol naq- gds-dir. Remarks. The notation of the vowels is so variable and imperfect, that the transcription can only be regulated by the living pronunciation. Long and short vowels are generally not dis- tinguished in genuine Turkish words. The connective ?', however, called kesri iddfe, is shorter then the common i; we write it therefore i. In foreign words the long vowels are usually pronounced as such and may be so written. The accent of the words is not very distinct, but floating as in the French language, and depending upon the whole sentence: we indi- TURKISH. 217 cate it therefore only in exceptional cases. The 5 is but a fulcrum for various vowels and must not be rendered by our consonantal sign ', the sound of the Arabic hamza * not being used in the Turkish language. The letter h at the end of o o a syllable is pronounced only after a long vowel (t, 3? L5)> after a consonantal letter it indicates only the presence of one of the short vowels e or , which we write then in our tran- scription instead of h. According to the vowel -harmony the four guttural-vowels a, o, u , i combine together , and the four palatal vowels c, o, u, i likewise. As in other cognate. lan- guages, the guttural consonants >', q, K, %, /, t, d, s, z, com- bine with the guttural vowels, although the latter four (t, 5 *> !& > (J& ?> ^ 3 ) ^ ft are no * f un d m "" ginal Turkish words, but only in words received from the Arabs or Persians. The specimen is taken from the Turkish-Tataric Grammar of Mirza A. Kasem-beg, translated by Zenker (1848), p. 17, and shows at the same time several essential deviations in our representation of the Turkish sounds from that of the author. We had the great advantage to consult personally one of the most competent scholars in this matter, the consul Dr. Rosen, who is both practically and scientifically perfectly acquainted with the pronunciation of the Turkish language, as well in Constantinople as in the chief provinces of the Turkish empire. P 218 TATARIC LANGUAGES. TURKMENIAN. q h k g n xO) y * 3 - s y t d n s z r p b m f w Cf. Ilminsky, in the Bulletin de 1'Acad. Imper. des sciences de St. Petersbourg , t. I, 1860. p. 563 sqq. KAZAK (WESTERN KIRGHIZ, small horde). Guttural (hard) vowels: a d o u i Palatal (soft) vowels: e e f i g if k g n t d n p b m I I Specimen. Asin dstn dstna bireket bersen bdsina, bgdenedei zoryaldp kiryd'uldai kuryaldp Kidir kehin kdsind nose turydn boz uige Kidir atd daresin bet bisese bul uidun unike kursdk koterib u_yd koze zaresin^ etc. (Speech of thanks after a feast.) Remarks. The Kiryiz (Kirghiz] are divided into the eastern tribes (the black or mountain Kirghiz or Burut), who alone call themselves Kiryiz, and the western tribes (Kirghiz -Kaissaks) who call themselves Kazak, and are subdivided into four hordes, the great, the middle or Sibirian, the small, and the inner or Bukeyew horde. The alphabet, as well as the specimen, belong to the small horde and have been communicated to the author by Dr. L e r c h. He has observed that some Kazak individuals pronounce the letter z as /, and others sometimes % instead of the common pronunciation k. Ilminsky writes q instead of k before the palatal vowels. SAMOYETIC DIALECTS. SAMOYEDIC DIALECTS. 219 Castren. Castren. a o u y - ' - h - a o u i e u kg y x - c 5 - s z y c 3 - - - t d n s z r b I p b m f - w Palatalised consonants: tdcnszrl Standard Alphabet. Guttural (hard) vowels: a o u i - > - h - Palatal (soft) vowels': e o u i e u kg n x - J - s z y t d - - - t d n s z r r I p b m / - w Palatalised consonants: t' d" i n s z' r I' Remarks. The sounds represented above belong to five different dia- lects, of which none possesses all of them. The Yurak wants the letters: e, %, ?, s, z, c, jf, d,f; the Taugi: g, e, o, u, %, A, r, r, /, , c", _/, 2, z, t, t' ', d, p, w; the Yenissei: i, e, o, u, X, r, s, z, c, j, z, z, t, t', d, p } w; the Ostyak: ', r, r, z, *', z, t'; the Kamassin: i, r, r, c, /, s, 2, t', d^ /. About the vowel u see above p. 210, and about i p. 54. The letter 5 is principally heard at the end of words, when another consonant is drop- ped. The peculiar letter, which Castren writes f, but in which, as he observes, the sound of r predominates, seems to be a cerebral r. (Castren, Gramm. der Samojed. Sprache, herausg. v. Schiefner. 1854.) P2 220 TATARIC LANGUAGES. MAD'ARIC (HUNGARIAN). a d o 6 u u k g e e d 0(0) u w() i i cs(ch,ts) ds ty cz(c,tz) (ds) t d p b a a oo u u e e o o u u i f ny n m h - S Z8(Z) f v r I Standard. k g - h - a - s z t'(ty) cf(dy) n(ny) - - i (fl - - - I d n 8 Z p b m f Specimen. 1. Es Ion az napokban, Augusztus csdszdrtol parancsolat adatek ki^ kogy mind az egesz fold beirattatnek. 2. (E beiratds lett eoszor, mikor Sziridban Czirenius tiszttarto volna.). 3. Mennek vala azert mindenek^ hogy beirattatndnak , kiki az o vdrosdba. 4. Felmene pedig Jozsef is Galiledbol, Ndzdrctnek vdrosdbol Judea tartomdnydba , a David vdrosdba, mcly Bethlchcmnck nevcztctik, mivelhogy Ddvidnak hdzdbol es hdznepe kozul valo vala; 5. Hogy beirattatnek Mdridval, ki neki jegyeztetett vala felesegul , es vala vdrandos. Ev. Luc. 2, 1 5. 1 . Es Ion az napokban, Augustus casdrtol parancolat adatek ki, hod' mind az eges fold beirattatnek. 2. (E beiratds lett elosor, mikor Siridban Tirenius tisttarto volna.). 3. Mennek vala azert mindenek, hod' beirattatndnak , kiki az o varosdba. 4. Felmene pedig Yozef is Galiledbol, Ndzdretnek vdrosdbol, Yudea tartomd- MORDVINIAN. 221 ndba, a David vdrosdba, mel' Bethlehemnek neveztetik, mivelhod' Ddvidnak hdzdbol cs hdznepe kozul valo vala; 5. Hod' beirattatnek Mdridval, ki neki yed'eztetett vala felesegul, es vala vdrandos. Remarks. Besides the fully assibilated palatals c and J, there exists in the Mad'aric language another class of slightly assibilated palatals, corresponding to the Polish and Serbian e, j, and still more to the Cheskian t, d" , n. They are now incon- sistently written ty, gy , ny, ly , the two former being uttered at exactly the same point of the palate. Undoubtedly they ought to be written either K(ky) and g(gy}, or t'(ty} and d'(dy). As, in fact, they approach more to the dentals than to the gut- turals, and are pronounced even nearer to the teeth, than c and J, and as moreover ty and gy (dy) not seldom are derived from the dentals t and d (Dorotya = Dorothea, gyemant = dia- mant) and instead of gy in former times di or dj, and sometimes even dy, were also written, we naturally prefer to write t' and d\ as well as n and I'. The explosives of the first column are the real dry tenues (see p. 134). MORDVINIAN (MokXa dialect). k g n h - t J - s z y t t d n s z r I p b m f v - Hard: a o u i Soft: e e i Palatalised consonants: /', n\ s' y d', etc. Specimen. At a i traks. Al'at asil alasats, tak son traksints tanks kambras sots, San'ts sgn asize ars'a, sto kambrasis traksti af ladai^ ozas' trakst 222 TATARIC LANGUAGES. lanks, senksa sto ickezi yalya molems ez yors'a. Son ozas\ kar- mas trakst aicCaminza; traksis' antsak, ozadit ala askilai. Al'as trakst paritsisi; traksis' kolai savir moli. Al'at ketsa mandil, trakst kosarize, sondiinza arai, son mant-ezda ardiz tui. Trafots' kolai savir moli^ livskidi i lek#i; a ardimats as son, kut' i savik. Traksis mele alat alu picas'; af madrena: traksis' asiz s'as saca arnemd. A s'avik erevi sodams: kona savir yakamd sacs^ smdi af lindema. (The peasant and the cow, a Mordvinian fable. Gramm. of Ahlquist, p. 120.) Remarks. The Mordvinians live with few exceptions on the upper and middle banks of the Sura, a tributary of the Wolga. Their language is divided into two dialects, the Moksa and the Erse, the former of which is treated in the lately published Moksha- Mordvinian Grammar of Dr. Ahlquist 1 ), whose personal experience of the spoken language we had the advantage to consult. His Grammar is printed in the Standard alphabet with a few unimportant deviations. We should prefer to write e instead of o, although the sound is the deep English sound (a), of which we have spoken above p. 50 sqq. His hard i ( bl ) is our '. As to the peculiar t, which is pronounced with a more lengthened pressure of the tongue against the upper teeth, the description seems to indicate the same emphasis, which we have met with already in several other Tataric languages. For this reason we prefer to write t. The Mordvinian lan- guage participates in the vowel -harmony which is found in the cognate languages, and its consonants are subjected to the influence of the following palatal vowels by assuming a shade of y, expressed by the palatal line: r I' etc. 1 Forschungen auf dem Gebiete der Ural-Altaischen Sprachen. I.Theil 1861. LIVONIAN. 223 LIVONIAN. k g t d p b m (h) s s Specimen of the Kolken dialect. Un se puoga kitis temmin: o iza, ma um patti tiend vasti tovast un vasti sinda, un eb uo emln vert, ku sa mmda ents puogaks nutdd. Bet iza kltis ents puosidin: tuogid nent ama yuvdd ornd un edigid tenda , un dndagid temmin suormiks keddi un kengad yalgi. Un tuogid ud liebiz vdskiz un tapdgid sie, las meg siomi un lustlgil velmi. Siest ku min puoga vel' yera kuolin, un ni ta um taggis yels sond^ ta vel' kaddin^ un um lievtid. Un ne urgist lustlgil velda. Ev. Luc. 15, 21 24. (Sjogren-Wiedemann, Liv. Gramm. vol.1, p. 354.) Remarks. The Livonian language is at present spoken in Livonia by only eight persons in the village of Sails, in Courland by several thousands. Of the two Courland dialects of Kolken and Pisen, the former has one vowel more than the latter, viz. e. Since it has besides the two palatal vowels o and w, also the two guttural vowels e and /, the vowel-system is very complete. The vowel -harmony however has left only a few traces. The sound of h is almost entirely dropped, and / occurs only in foreign words. The standard work upon the Livo- nian language is that of Sjogren, edited by Wiedemann, St. Petersbourg. 1861. in 2 vole. 4, and the alphabet used 224 TATARIC LANGUAGES. in this work is our Standard alphabet. Only, j is employed instead of our y, and a instead of e. There is a difference made between the a of the Pisen dialect, and o of the Kolken dialect; the first is an o still more open than o. We should prefer to write both o, as they do not occur in the same dialect. The vowels o and o are essentially the same as our e and i. Our proposition, however, to write them so was at the time of publication not yet known to the learned editor of Sjogren's work. TAMIL (TAMULIAN). 6J PT- Sanskrit sounds <5t> Go rr^ - /i CT* FT* (63 LLJ / L. L_ 63ST en m m (53T IP - 0, &, 5 n GO * u u LQ QJ - a a e e 06 i I u u k y,g w - h * *3 w y J t d w i s r r n r - t d,d n r / s p b m V - TAMIL (TAMDLIAN). 225 Specimen. Ycnnattindlyenil Parabaran tammudiya ore peddna kumdrani oizucdzikkadavan yavano avan kaddvpoydmal nittiya zivani adi- yumpadikku averiye koduttu icvalavcii ulayattdridattil anbdi imnddr. Ev. Job. 3; 16. Kumdran aoani noki, tayappane tevanukum umakum virotamdi pavam ziten itumutal umudiya kumddan yandu arikapaduvaterku nan pdtiran allavendu sondn, Ev. Luc. 15, 21. Remarks. We follow principally C aid well in his "Comparative Gram- mar of the Dravidian languages." London 1856. The vowel i is mostly weakened from final a and a, but "every trace of the sound of a has disappeared", says Caldwell. It is mostly long, but sometimes also short. It is evident, therefore, that we have to do with a simple vowel, not with a diphthong, as it is represented by Caldwell and others (ei or ai), and that the Tamils were right in giving to it a simple sign. It "accords in sound very nearly with the sound of e or ey in Turkey 1 '', according to Caldwell. "We have no doubt, that it is the same as that Tulu vowel, which has been compared by a good observer to "a short and indistinct w." These different descriptions lead us to believe, that it is the vowel peculiar to most of the Tataric and several Slavonic languages, which we write i (see above). The diphthong au occurs only in Sanskrit words. The slight change of sound which all the vowels, except w, undergo after the cerebral consonants, in- cluding partly also the common r and , is, in connection with certain traces of the vowel-harmony (Caldwell p. 101. 136.), most interesting for the linguist, but cannot be represented in transcription. With regard to the consonants, the letter , d is erroneously taken for a semivowel by Caldwell (p. 108) who writes it R. The Tamil Grammarians themselves divide 226 TATARIC LANGUAGES. their consonants (C. p. 102) into six surds or explosives (val- linain): , c, , , p, t(n) , six nasals (mellinam): n, n, n, n, m, w, and six semivowels: y, r, I, v, r, I. The pronuncia- tion of the letter (R) is generally described as a peculiar com- bination of t and r, or, if sonant, of d and r, which again shows its explosive nature, and prevents any transcription with the basis of r or R, instead of t (or d). The Tamulians would certainly have arranged their varga's, as we have done it, according to the Sanskrit principle, if they had not followed too closely the Sanskrit alphabet, in rejecting at the end the four letters r, /, , n, which the Tamulians have added to the Sanskrit alphabet. For the old Vedic 35 I had disap- peared in Sanskrit, and T r corresponds in the Tamil, as in the Hindi, Hindustani, Sindhi etc. to the dental r. The Dra- vidian and Hindi r was derived from d and was probably slightly different from the Sanskrit r; we should even prefer to write the Dravidian sound /, if two diacritical signs were not too heavy, and if the transcription r were not already too generally received. It is indeed our opinion (see above p. 99. upon the Hindi letters ^ r and ? rh) that the Tamu- lian letters t, d have a similar relation to the cerebrals t, d, as the palatals c, /, have to the gutturals k, g; and we take , c?, r(r) as peculiar slight assibilations or vibrations of the cerebrals , d, r, approaching to the combinations $, dz, rz; for there is physiologically a very slight difference between r and s, r and s. At all events we must choose single cha- racters for the single Dravidian letters. This assibilation of li 4* T pushes the Jtip of the tongue a little forward and nearer to the dental point. Hence the letter <53T, our w, which occurs only before d and at the end of words, and which originally belongs certainly more to the cerebrals than to the dentals, as even its figure shows. By a Tamulian euphonic law, the surd letters k, c, t, t, t,p, are pronounced as so- nants, wherever they occur singly in the middle of a word, TAMIL (TAMULIAN). 227 and three of them , k, c, t, lose in this position even their explosive nature and become sonant fricatives. No sonant letter, on the contrary, begins a word. This law explains the fact, that in Tamil and partly in Malay alam the same characters serve to express the surds and the sonants. Our transcription must of course follow in this respect the pro- nunciation. If in the middle of a word the surd letter is to be pronounced, its character is repeated. The sonant <5 (k) is pronounced y; the sound of g is sometimes retained in Sanskrit words. The sonant & (c) is pronounced z, "as a very soft sh"; the sound of j is sometimes heard "in vulgar Tamil", and "in the use of those Sanskrit derivatives in which the letter ^ / is found in Sanskrit." The sonant letter &) (t) is pronounced "with the sound of the soft English iK' '; the sound of d occurs only, "when it is combined with a nasal, as in andam" The Tamil 6LI has not the English sound w\ it is generally rendered by v, and we keep this transcription, although the description of this sound might raise the doubt, whether it were not rather pronounced like the w of middle Germany (see above p. 75). The Tamil is destitute of sibi- lants and aspirates, as well as of the simple spirans h. The letters s, s, s, h, if occurring in Sanskrit words, are repre- sented by the corresponding Grantham characters; s in Sanskrit derivatives of earlier date is replaced by the Tamil c or , the Sanskrit s by the Tamil t or d, sometimes by r, or even by t or d; the Sanskrit s sometimes by t, cor z, and sometimes it is omitted altogether. The Sanskrit h is omitted in the Tamil. The connection of consonants and vowels is analogous to that in the Sanskrit, the above given vowel characters being used only in the beginning of words. 228 TATARIC LANGUAGES. MALAYALAM. o$io 6) <&> 00 .OJ S? S CUO 00 O rm a oj eru 3 00 oo Ud OTO CQJ ro cu LQ OJ a a e e oo i I u u r f I I ai au A # n h s - kh gh c " 3 n s y ch jh t d n s i ih f)h t d _ m r _ _ r r t d n 6' r I th dh p b m - IV ph bh Specimen. Entakondennal Deiwam tande ega/jatandiya putane, awenil unsivasikkunawen orutenum nasiccapokdte, nittya/jiwan untaken- tunnatina, taruwdn takkawannam eteyum lokatte snehiccu. Ev. Job. 3, 16. Appol makan atvanota, appane, nan swerggattina nereyum, ninde munbdkeyum pdpam ceytirikannu inimel ninde makan enna collappctuwdn yogyanalla enna paranu. Ev. Luc. 15, 21. Remarks. In Malay dlam e and , o and 6 are represented only by one character; in our transcription, however, they ought to be marked according to their quantity. We have excluded TDLU. 229 the compound letter H, which is generally exhibited in Gram- mars. There is a peculiar nasal in Malaydlam, Telugu and Kanarese, which is pronounced m at the end of a word; but it may also euphonically be substituted for any other nasal and will then be pronounced accordingly. In our system it need not be marked. The letter Qj is pronounced w, not v, as in the other dialects. Cf. the Grammar of the Malayalim language by the Rev. Joseph Peet (Ch. Miss. Soc.). 2 d ed. Cottayam. 1860. TULU. k g n h s kh gh ,/ h t y ch $h t d n s T I th dh t d n s r I th dh p b m - V ph bh a d e e 06 i I i u u T f ai au Specimen. Yesu, Yehudada Bethlehemudu arasdi Ilwoda dinoledi piitti bokka inda yotiseri mudayididi Yerusalemagu battidi Yehudye- reyi arasu ddi puttindye volu ulle dayeg andunda yenkulu dya bollini mudayidi tiidu dyagi drddhane malpere battd andidi panderi. Ev. Matth. 2, i. 2. Remarks. The Tulu is ordinarily written in the Malay dlam character. The vowel, which we write i, has been compared to a short and indistinct u (see above p. 225). 230 TATARIC LANGUAGES. KARNATAKA (KANARESE). 3D S? So w 6 a -So a go a a e e oo i I u u if a au k g n h s - kh gh c 3 n 8 y ch jh t d n S i th dh t d n s r I th dh p b m - V ph bh to Specimen. Ydtakkendere dtanalli visvdsa riduvavarellaru nasavdgo.de nitya jivavannu honduva nimitta Devaru tanna vobbane mayanannn kotta lidge jagatanna astu prlti padisidanu. Ev. Job. 3, 16. Adare maganu avanige tandeye paralokakke virodhavdyiyii ninna mundeyu papa madidd/iene ndnu innu ninna magarendu kareyalpada yogyanalla annalu. Ev. Luc. 15, 21. Remarks. In Kanarese the letter t is confined to the poets. TELUGU (TELINGU). 231 3D TELUGU (TELINGU). C3 S 2o So o d & w etc. & 80 32 C? 3 3 e> * # n h g - kh gh cf id \sylt I yJv n s y ch jfh t d n s i th dh t d n s r I tli dh p b m 5 V ph bh (0 a d e e oo i I u u r r I I d d etc. ai au Specimen. Yendu vallanante ayanayandu msvdsamunsevddevvado vadu nasatnu pondaka nitya jivamunu ponde nimittamu Devudu tana yoka kunidruniccinattuga dagattunu prttipallacenu. Ev. Job. 3, 16. Appudu kumdrudu dyana to tandri paralokamunakunnu nl- kunnu virodhamugd pdpamu cesi yunndnu yikamldata nl ku- mdrudani piluvabada nenu yogyudanu gdnanenu. Ev. Luc. 15, 21. Remarks. The two first letters of the second class have two sounds ; they are pronounced c r and/ in all Sanskrit derivatives, and in Telugu words before , , e, 0, ai; before the other vowels they are pronounced t and d, as in Marathi (see above p. 109). "The letter t is found in Telugu (as in Kanarese) poetry, but in the modern dialect of the Telugu it has fallen into disuse." 232 MONOSYLLABIC LANGUAGES. KWAN-HWA (MANDARINIC) dialect of Nan- kin. Vowels. a Q L/,w,yw kky,w,yw tSu> tshw tSy,w,t/w tSfly,u>,yw ty,w . tllif,u> PH phy e i i v r att(ao) ai ci eu Vowels with end-consonants. an eh in un an en in Consonants preceding a vowel. n Tones. phin,, the floating; saw', the ascending; /-//7/? v , the descending; cV N , the returning. High phin pa, Low phin pa, San pa 1 Khyu pa^ i Specimen. A g ro A 3<"> Ev. Matth. 2, 1. 2. KWAN-WITA (MANDARINIC). 233 1. Hi,-lyu\ wan t &,, Ye^su, k? sen, yu, Yeu L thafi Pe\-l- heh L , yen! po\ s su\ zin^ tsf tun, fan tsfi Ye^-lu^-sa^-lin. 2. Ywe\: sen, _, wei Yeu^-thafi zin^ wan L tse*, nan, tsafi? No 1 tsa tun, faii L kyan" khi, sin n ki lai L pa tsi,. Remarks. The Ktvan-hwa or Mandarin dialect is spoken by the people of the middle provinces of China and likewise by the higher officers and cultivated classes throughout the whole country. In this dialect, which is better known in Europe than any other, the monosyllabism is developed to the highest degree, every syllable being a whole word ending with a vowel or one of the two nasals n and n. In former times the Kwan- hwa distinguished surd , sonant and aspirate consonants , as we have shown elsewhere l ; at present the sonants have disap- peared. The letters f and w are always followed by a vowel; all the others may have inserted between them and the fol- lowing vowel one of the semivowels y or w, or both of them, as our alphabet shows. European scholars use mostly instead of these semivowels the full vowels i and u (or dialectically e and o). We do not repeat here the reasons, why this custom is scientifically and practically inconvenient. We have spoken in the same place upon the letter z , as to which we are not sure, whether its actual pronunciation is not rather /, as its place in the sound-system as well as the description of the sound by some scholars, seem to suggest. The vowel u is often dialectically pronounced u, which may be written wher- ever it seems suitable. The sound of r occurs only in one word, formed by this single letter, but with different tones. It is commonly written by the grammarians eul, or ulh, or 1 Ueber Chines, und Tibet. Lautverhaltnisse. Schriften der Berlin. Akad. 1861. 234 MONOSYLLABIC LANGUAGES. vrh etc., but it is nothing else than a vocalised, probably ce- rebral, / (or /) which we write consequently r. The vowel, which we had formerly proposed to write r, according to the pronunciation of Mr. Guzlaff, whom we consulted about it, seems to be derived in the Chinese system of sounds from the vowel u\ but it is pronounced entirely like the Tatarian and Slavonic "hard" i. The Russian missionaries represent it therefore by their i>r, and we have to render it consequently by /. It occurs only in the words si and tsi\ The tones which, in Chinese, are an essential element of speech for the distinction of words, were hitherto represented by European accents of quite a different meaning, or not expressed at all, and some scholars used the same accent for one tone and others for another; for ex. Morrison and Remusat represent the saw-tone by pa, the khyu-tone by pa', the high as well as the low phin-tone by pa; Marshman and Medhurst, the saw- tone by pa, the khyu-tone by pa, the two p/mi-tones by pa; Medhurst in the Fu-kyen dialect, the high saw-tone by pa, the low San- tone by pa, the two phin- tones by pa and pa, others the saw -tone by pa, the khyu-tone by pa'. A new system was under these circumstances indispensable. The system, which we formerly proposed and have repeated above, follows as closely as possible the indigenous writing, com- pleted by the missionaries of the southern provinces. The Man- darin dialect has only five tones, the phin-tone alone being divided into a higher and a lower; we omit therefore the little horizontal line which distinguishes the lower i. e. deeper pronunciation of the other tones. HOK-LO. HAKKA. 235 H K - L 0. p i i u d e -I o u t ai au oi en k 9 kh n ' h - Tones. tf dz tsh - - y high phin pa, ts dz tsh low n P a i t d- th n s I high son pa 1 P b ph m - low w high r> khi pa T low y> pa>- high nyip pa, low Pi Remarks. The Hok-lo dialect is spoken in the north-eastern part of the province of Canton , in the department of Tsau-tsyu. The alphabet, as stated above, has been furnished to the author by the Rev. Lechler who lived several years in this country. The nasalisation of the vowels is less open and more squeezed than in the french vowels. The Fu-kyen, to which the Hok-lo dialect belongs, distinguishes all eight tones, but the low son -tone is pronounced with a peculiar modification, which might be expressed by pa'. Vowels. a e o i i u in an ai oi m cu HAKKA. Consonants. Tones. ft* n h high phin tsh - s y low phin tsh . - son th n s 1 khi ph m f w high nyip low nyip Q2 pa, pa, pa 1 23G MONOSYLLABIC LANGUAGES. Specimen. 1. Hi,-Jut\ icon, ku? si,, Ya,-si t kP yen L tshut\_ v , tshafr Yu^-thafi kok^ pa\_-lfi-hcn, yip\, >/u, ki' tsak\ yu, tshoi len L ka? nyin, tehoi, tun, phcn 1 theu L Ioi 2 tau" Yaji-lu^-sa^-lan, kin, san . 2. Kan, yoii" w<: yu, tsak^ nyin L tshut^ se" loi A tso^ Y\t L - {hai> nyin L ka? won 1 ; ki^ tehoi, lai^ tsa{\ tlian^ li,f Nai^ tsJioi, tun, pen, khoif tan 1 kya, sin, syuk^; so 1 yi, thit\ sf lcri L pa? fo^ ki^ Ev. Matth. 2, 1. 2. Remarks. The Hakka dialect, as spoken in Hoh-koh, has been already reduced by the Basle-missionary Rev. Lechler to the Standard Alphabet, in his translation of the Gospel of St. Matthew (Berlin, I860) from which we have taken the above specimen. It has one tone more than the Kwan-hiva, vix. a high and a low nyip - (zi-) tone. It has moreover a vocalised w, which is to be written w; but it has, on the other hand, not the vowel ?, nor the consonant ~, instead of which it uses ny. The Hakka dialect, as well as most of the southern Chinese dialects, admits besides n and m other consonants at the end of words. In fact, all the words having the nyip or "returning" tone, end in one of the three consonants k, t, or p. In compound words, however, and in other cases, these final consonants, when preceding another consonant, are not pronounced. If, therefore, we find in the translation of Mr. Lechler the names of Bethlehem, Jerusalem, Abraham rendered by Pak-li-hen, Ya-lu-sat-lan , A-pak-la-hon , we have to read: Palihen, Yalusalan , Apalahon. In such cases, we prefer not only to put little lines between the single syllables, but to put the apostrophe instead of the elided consonants: Pa\-l^-/ien 11 Ya L -lu"-8a\-lah n A,-pa\ -la/'hon 1 , or to omit even the apostrophe, the elision being sufficiently indicated by the tone. THAT. THAI (Siamese dialect). 237 cV -]- e - o ai #z aw am if) si 2. th, n. th, 18. th, 19. n, 20. d, 21. t, 22. th, 23. th, 24. th , 25. n, 26. b, 27. p, 28. ph, 29. f, 30. ph, 31. /, 32. ph, 33. m, 34. y (j), 35. T, 36. I, 37. W (v) , 38. S, 39. 8, 40. , 41. A, 42. I, 43. a (0), 44. A. - They state expressly, that there is no distinction in the present pronunciation between the different letters transcribed alike, except that certain letters viz. our letters M, ^, ch, th, th, ph, f, s, s, s, K, are always followed by a vowel with the high ascending tone. This great number of identical sounds, amongst which we find 5 kh, 4 ch, 6 th, 3 ph would be quite unintelligible and embarrassing for the linguist, if we did not distinguish them in transcription as well as they are distinguished in Siamese writing, and this is only possible, if we are able to distinguish them at least etymologically; for it would be absurd to sup- pose that this identical pronunciation existed from the be- ginning. Now, the alphabet will be understood at once, if we reestablish the ancient order as we have done it above. Taking those letters, which differ only by a slight break in one part of the character, as later variations derived from THAI. 239 one and the same sound, we find but five original divisions of every one of the five explosive classes, in perfect harmony with the system of the five Devanagari - classes and with the Siamese Pali -alphabet as communicated by Low. The same Indian arrangement is evidently followed in the subsequent letters corresponding with the Sanskrit y, r, Z, w, *, , s, A, /, to which are still added two other letters fl, the fulcrum of initial vowels ( 3 ), and Ef a second k. The fifth column con- tains the nasal letters and we learn from it, that the palatal letter flj, which now sounds only like ?/, was originally the palatal nasal n. The fourth column corresponds with the aspirated medials of the Sanskrit. It seems that even at present this original value has not quite disappeared and that the guttural *JJ at least is still heard as gh. The strong aspiration may have misled many a foreigner respective the real pronuncia- tion of the first part of these compounds. The third column ought to represent the simple medials ^*r - J A3 OO Tones. The floating or natural tone, not written. The acute (an 1 mylt) _?_ The grave (sye 1 pauk) i ad k g n ) is kh gh e 6 c j n i 1 - ch jh i 1 i u u t d n j r th dh d u t d n 1 s th dh ii au p b m u i - ph bh Tones: a a' a x MRANMA (BURMESE). ' 243 Remarks. The Burmese writing is syllabical as the Sanskrit; the in- herent a is not written ; the mark c placed over a consonant indicates the want of any inherent vowel. The vowel -signs, as given above, are all initial. The vowel called triphthong by Latter, which is composed by the signs of u and 4, and pronounced sometimes "intermediate between o and u'\ and sometimes as "a short oi or long z", seems to be our -4, I of the Tatarian and Slavonic languages, which we have found also in the Chinese language. The nasalisation ("anusvdra") affects only the vowels a and u. The pronun- ciation of the first diphthong is, according to Th. Latter, ei, not ai. The only Burmese sibilans corresponds with the Sanskrit ^T -s and seems to be pronounced very near like the Polish s. We may therefore write it s, if with any diacritical sign at all. The present pronunciation of the Burmese letters deviates in many cases from the received orthography. It seems there- fore advisable for missionary purposes to follow the actual pronunciation, whilst in linguistic researches it will be ne- cessary either to transcribe exclusively the old orthography, or to mention it at least in every single case of deviation. Cf. the Grammar of the Burman language , by F. Carey, Se- rampore, 1814. and the Grammar of the Language of Burmah, by Thomas Latter, 1845. All pure Burmese words are monosyllabic. Of the peculiar Chinese tones, only three are in use among the Burmans, of which the floating or natural needs not to be indicated. 244 ISOLATED LANGUAGES. Y U K A G I R I C. k 9 X - \f V I / V V C J n s z y t d n - r p b m w Specimen. Omoca ddil terikadeni lengi. Ke lomdok yonjuririma? Pon- bure oille. Kaweik adakun inlicaon kecim. Ke lomdok yowju- ririma? Yonjode oille. Kaweik kecim adakun monogoj[. Ke lomdok yonjuririma? Serezen moza, oille serezen pugalwe. Ka- weik adakun comogina kecim. Remarks. The alphabet and the specimen are taken from A. Schiefner, Ueber die Sprache der Jukagiren, in the Melanges Asiatiques t. III. 1859. p. 595 sqq., who there uses already the Standard Alphabet. CHUKCHIC (CAUCAU). kg n % h t d n p b m Palatalised consonants : g X t' r I' y r I 'W JAPANESE. 245 Specimen. y[wa%o , ^a^avnetwa , j[ uci. Minkri ^itirkin ? Torekaite geal- lim. Tawa% warkin ? Ketai, mintawa^om. Inan yarsnok hiwut torekaite minrestim. Men^ostin? Doten %ac. Renko Inege %ilkit. X wa X wki! min%ametwamik kinemal. Remarks. The alphabet and the specimen are taken from the treatise of L. Radloff: "Ueber die Sprache der Tschuktschen\ in the Memoires de 1'Acad. Imper. des Sciences de St. Petersbourg, VII e serie, torne III, n 10. 1861. The author of this treatise has already employed the Standard Alphabet with a few ex- ceptions. With regard to his a as corresponding to the Rus- sian fl 'we are not sure , whether this vowel is really different from our e. a e o JAPANESE. k g t,ts,t8 (p) (n) n m ./>, \ indicates the lengthening of the preceding vowel. ? (0 indicates often the shortening of the preceding vowel, or the doubling of the following consonant. iJAi, final nasal. 246 ISOLATED LANGUAGES. f 7 /a ga a >r / ke f ki i 3 Ao a* ^o ^T o \> ta da 7 r te dzi^dzi y M Jo \*O /<-jy ^J^ 60 tf' s* pa ^ pe U. pi j^\ J30 7 ? )w f* ba * be U." U ^$ ^ 7" 67* > fajcajw. -X A-. it A-- ^ /o,". . "7 /" r na ff - ni ^ wo ^ mi ? ma y me I mi t mo l\ mu f sa -t Sf, v- j*,# V so X f za 1? ze ^> ^ xltv / Z J* A ^p ya - - - - B yo a. yw 7 ra V re 1) r a ro /u- rn VI wa 2. we f wi 3 wo - Specimen. Japanese proyerb. Kagami fa sugata no yosi-asi fo miru mo } kokoro no ktyokn tstyoku wo tadasi aratamenu ga tame nan. Japanese proverb. Kon-nitst wa. Nani-wo o me-ni kake-masiyo kaf Mo sii- kosi oki-no-wo o mise. Ki-ro wa nozomi-masenu. Hei, zui-bun deki-maw. Sore wa Nippon-no fi-doi'i-de ari-masfika? Si gu- watsu ziyu yokka-de am-masu. Fito fako ffyak kin iri-ni nasare. San futo-de itei ydto-ni nari-masu (three feet make one yard). Is-styaku-no to-wo itei ziyo to -l-masu. J. Hoffmann, Shopping-dialogues. JAPANESE. 247 Remarks'. Of the two Japanese formes of writing, the cursive Firo-kana and the more square and distinct Kata-kana, we have exhi- bited above the latter, which has been reduced already to our Standard Alphabet by J. Hoffmann (the learned editor of Donker Curtius's Japanese Grammar, Leyden, 1857.) in his ^Shopping-dialogues in Dutch, English and Japanese, Leyden. 1861." The Japanese used first for their language the Chinese writing. In the 9 th century they derived from it a Japanese alphabet of their own and limited it to 47 syllabic signs. These syllables were put in such order as to form a little poem, the beginning of which I-ro-fa became the name of the alphabet. We learn from it, that the Japanese language of this time distinguished only ten consonants, which were com- posed with the five vowels a e i o u; three combinations were left out as not existing in the language, viz. yi, wu and ye (or we). The old consonantic system was therefore this very small one: * ' - y t n s r p m - w Afterwards the sonants ^, h e o ts dz tsh n s z i u ts dz tsh - - - t d th n s z p b ph m - - a e m n y = q ' ' q I'll Specimen. QT Foucaux, Gramm. sur la langue Tibet, p. 195. Dei tshe de'i dus na yul Bdranase >-dir dran sron l-na br-'gya zig y-na-s te; dran sron de dag gis s-ton pa Udpdla zes b-yaba dampd-i tshos s-lob tsin bs-gom pa la d-ga-> bd-s kun tu zin; su la dampd-i ttihos yod pa de b^dag la s-mrana. R 250 ISOLATED LANGUAGES. Remarks. The Tibetan writing is syllabic like the Sanskrit, from which its characters are derived. The orthography of the Tibetan literature was fixed at a time, when the language was still in a very different state. Many letters which then were pro- nounced, are at present silent. We have the choice, either to give up entirely the old historical orthography, following only the actual pronunciation, or to seek for a compromise between both. We have made a proposition to that effect in the above (p. 233) mentioned treatise, and our specimen will best show what we mean. As almost all the silent letters pre- cede or follow the root, which alone is pronounced, they might all be transcribed, but separated from the letters pronounced in a conventional way, and the altered pronunciation of a few other letters might be placed above those of the old or- thography. The brackets as in bs)gom = yom; d)ga(' = ga; na(s = ne, which we formerly proposed, seem to be less convenient, than a separating line, although also this line has in our European writing an other meaning. The change of pronunciation occurs principally in the letters a, r/, >, m, y, which become sometimes e, rf, c?, n, ts , and might then be written , y, r, m, y. Other minor changes, as the softer pronunciation of b before vowels, or tlie sharper of d before r, might be omitted. The letter ^ was originally a weak nasal, but is now, if heard at all, weakened into the sound of our >. The vowels, except a (which is not written), are expressed in the middle of words by the signs known from the Sanskrit, and in the beginning of words they are added to & as their fulcrum; the same letter without any vowel-sign designates, as in Sanskrit, a. The quantity of the vowels is not distinguished in Tibetan, except in Sanskrit words, where the long vowels are commonly expressed by adding underneath the letter H. GEORGIAN. 251 GEORGIAN. e o i u i (u) 8 - - b i 6 6 d - 1> ^ $ y, B - a 3 i oj 9 - 3 q _ _ _ x h ^ 9 k - x r C' 3 $ _ s z t' d t - - 1 * * t' d t n 8 C r I p> b p m - - w JEkc Specimen. b-ltf> OwjOT^ODO 0DO'|^noo job ^bocn-Jd-n^ibo ^'jjoi-' obl)5> oX-jiocnblfb : Brosset, Elem. de la 1. Georg. 1837. p. 268. T>midata da udlewelta moPameta Dawit da K'ost>ant>inesi. Ese udlewelni mot'ameni iqwnes natesawit Kartwelni, sazywar- tagan Ap^azetisata , sana^ebta-gan Aryuetisa; aznaurni iqwnes Pomit, da natesawni ertman-ertisani , udlewelni da dlierni brdo- lasa, mj^edarni suenierni da ganikmulni tqobasa mPertasa. R2 252 ISOLATED LANGUAGES. Remarks. The vowel c is described as a very short e and occurs especially after u: ite. In the same way i mostly forms the second part of a diphthong ai, ot, but sometimes it occurs also alone at the end of words, as in saidumbt; it is rendered in Russian by H or L. As all the characters and their order are derived, like the Armenian, from the Greek, the sign of the vowel u is originally a composition of o and i> (Greek ov}. The Greek vowel v alone does not belong to the original Georgian alphabet, but is added afterwards after u (without any numerical value). It is found, however, in a few words, 8 r ,d (=$widi), seven, and may then be rendered by u: sud. About the dry tenues, which we write ', t> etc., and which are common to the Georgian, La$ian y Minyrelian, Suanian, Ab^asian, and other Caucasian languages, see above in the Ossetian alphabet, p. ]39. We are not quite sure of the pro- nunciation of that letter which we have rendered by j^. Euro- pean Grammarians usually write kh instead of our % and kkh instead of our %. It seems that the latter is the corresponding fricative to s z y if d t> - - t d P n s z r I I p b P' m ~ - w Remarks. The letters, which we write with > are apparently the same dry tenues, on which we have just spoken p. 247 (cf. p. 134). On the aspirated 1 see above p. 172. 254 ISOLATED LANGUAGES. ALBANIAN (Toskan dialect.). f - - x - a * r - X Y *J YJ % a o I V OV T d V a V d 7i b ft (f /S e - - - h - - a k g - X Y - e o * 9 n s z y l ' i u u t d n s z Q d r 1 p b m f v - Specimen. Kjf vjf j.tb(>i rd vjg ftfi'd f.tb()i()oi> , e t xj Otrf xj do vya rjc vm i xiy xjf xje ax6[ia no. ^.jfQf. J'ltQ ^fff oa djf t u btti't TE dvfiaiLt c eciy , %ji xta 9 t nny vdt dtc vt. I vyeci djdtyf xje OTLOU vdt dec, vovx ovubvt, no att f J[odt vd? art IE dent, f aije f yjivt TOO Toobf're ( f. vdt GTUV T fi(>, f dat> vdg yQa TE TVQ 7iQ re (>ITOVQ. Popular tale. J. G. von Hahn, Albanes. Stud. II, p. 167. Ke ne ml/ret nde ne vend e mbrcteron^ e i He Gene Ke do te vriteiy nga ne nip i tiy Ke Ke akoma pa Ure. Per kete pune sa dyem beine te duvdize V etiy , Ke kis , i stiy nde det e i mbut. I treti dyale Ke stiu nde det, nuk' umbut, po taldzi e hodi nd* dne te detit, e atye e gene tea teobene e e muare nde atan te ture, e e ddne nde qra te ture per te riture, * P fl / Q " * Q POLYNESIAN or MALAYAN LANGUAGES. MALAYAN. 255 Remarks. The Albanian language is divided into two dialects, the Toskan and the Get/an. The Toskan write with Greek letters, the Gegan with Roman. We follow the Toskan grammar of J. G. von Hah n in his learned Albanesische Studien, Wien, 1853. 2 d P. In the Gegan. dialect occurs the freuch nasa- lisation of vowels, expressed by the addition of v , which we render by the sign over the respective vowels. MALAYAN. In foreign words. * - & o J> - O U" c s _bJs o a a a e e OQ i i u u 3 - U - H | ^ n - - X / c j n - y t d s z i d - - - - t (2 n s r I G d s z p 6 m - w f 256 POLYNESIAN or MALAYAN LANGUAGES. Specimen. Makka tnqh'aluwarkan dlya dqrripadda karajdanha padda bdrah (dydjm yah dikahqndaklna , dan dimuliydkahha a/can bdrah siydpa yah dikahqndaklna dqhan tdhan qodqratha. Bdran si- ydpa mqmmunuh orah dqhan tiydda sqbqnar Kaqha niscdya di- siksa allah dqhan dpi ndraka yah a mat hdhat. Schleiermacher, de 1'influence de l.ecrit. sur le langu. p. 602. 604. Remarks. The Malays, like the Arabs, distinguish in writing only three vowels, short and long, a a, i I, u u, using the same signs for i and e, I and , u and o, u and 6; i^_ is always a, but is sometimes a, and sometimes the indistinct vowel, which might be transcribed e, or as we prefer it in this case q. The pronunciation of ^ and g is described as between our c j and t tf', it seems therefore that our nearest expression for them is c and j. We write the merely Arabic sounds as we write them in Arabic. We have followed principally the exposition of the Malayan grammar by Schleiermacher in his book: de Vinjiuence de Fecriture sur la langue, 1835. p. 409 sqq. There, however, the two cerebral letters t and d are not mentioned. They are at present, as it seems, fallen into disuse; but the new invented sign 3 of the Malayan alphabet proves, that this letter d was used, at least in former times. The corresponding t was expressed by the Arabic -b, which after- wards was commonly confined to the words of Arabic origin, and replaced in genuine Malayan words by ^ t. BATAK. 257 BATAK. 7-7 o o OC e a e o ca ja pa ba na na ma ha,, sa ya va la wa Remarks. There are three dialects of the Batak, which is spoken in the north of Sumatra, viz. the Toba, the Mandailin or Ankola, and the Dairi. The characters given above belong to the Toba dialect, except -^ ca, which occurs only in the southern branch of the Mandailin dialect. In the same dialect ^^ is written instead of ^^ >a 7~~? or ""^ instead of "^ ha or for sa, . m a ^ *^ for ~o na, -,O OJl oan jh OS ' ' O I v Specimen. Negari Bali vonten tiyan; budinnipun lahkun dennin rosa. Padamellannipun sabin. Namannipun jaka Pirannon. Miren j^abarj yen negari Messir havis teda. Jaka Pirannon lajen kesas daten negari Messir; bekta dagannan pantun liutavi huvos. Saren dumugi negari Messir , kapangis tiyan dusun hin Karas. JavaanscheSpraakkunst door Cornets de Groot uitg. door T. Roorda. 1843. p. 68 Remarks. The system of vocalisation is essentially the same as in the Decanagari. The vowel e does not occur in the beginning of words. The vowel a which after any consonant is not written at all, has two different sounds, according to certain rules; the one is our pure a, the other a somewhat closer and there- fore more indistinct a approaching to our o, as Qx approaches to an indistinct e. We write therefore those two sounds a 260 POLYNESIAN or MALAYAN LANGUAGES. and e. The sign is described as a final n .', it seems to imitate in its figure the Sanskrit anusvara, but without its peculiar nature; we prefer therefore, not to distinguish it in our transcription from the full consonantal h. The final letter 7 corresponds with the Sanskritic visarya s. About the two cerebral letters n and s (s), we think that T. Roorda is right in what he observes in his edition of Cornets de Groot's Javaansche Spraakkunst, Amsterdam 1843. p. 8. As the no- tation of the final w, the different notations of r also seem to be imitated from the Devandyari, and the two signs ^ and ^l called Pa-cere and Na-lelet seem to represent the Sanskritic r and /. The full vowel characters are sometimes used as o o initials in original Sanskrit words. Instead of them the letter h is generally used with its respective vowel sign. It seems preferable to transcribe this 7t, although it is at present scar- cely audible. 1) A Y A K (Borneo). a k y n a to. ) d' - e o t d n i u p b m r w Specimen. Aton olo id'a tempon anake hatita dua biti. Dan id'abiisu intu o L o o o awen ta hamau dengan bapa: Apaii, tena akanku bay in ranw ida baria ayunku. Dan ia membayi akan awen ta pcnataue. Maka d'aton ara andau limba ta, anak ida busu menampunan kara ramo, dan hayoet akan leiou awaii ked'au, heta ia menanan ramoe awi kapapan yawie. Luc. 15, 11 13. H. C. von der Gabelentz: Grammatik der Dajak-Sprache. Leipzig. 1852. p. 45. MAKASSAR. 261 MAKASSAR. A XX x\ ka ya t'a da ta da pa ba na na na ma <** ha sa ya ra la wa Specimen. lya-minne ankdna-kdnai pau-pamvanna D'ayalankdra. Ala siydpa- siydpad'a harden lompo, a>makeya makota, a'mindwan iraicdnan parentdna D'ayalankdra. Na and'o D'ayalankdra, tau Idnibusu* pamdi*, na ddele> bitaranna ri sikamma bone-buttdna, siyagdan ri-patarintina, riyanron-kemokanna, ri-pasaribattananna. Pdssala'. Na ni-surokdna-mo pau-pauwanna. Niya> se>re ka- rden ri parasdnana, niy arena Tina-Sumpe. Naiya arenna ka- r deha, nikdna Rdaa-Adan; ma>lompo kakaraenanna , ma>ld>ban irateyaiidnin, siyaydn ri bawa anin; na kaleleyan-mo biritdna ta>se>re-ta>se're parasdnan, lambusuma siyagdan labona ri-sikamma pdkereka, siyagdan ri-tau-kasiasiya. Matthes, Makassaarsche Spraakkunst. 1858. p. 14. Remarks. The Makassar language is spoken in the southern part of the island Celebes and partly in the small neighbouring island 262 POLYNESIAN or MALAYAN LANGUAGES. Saleyer. The vowel a is inherent in every simple character; the other vowels are expressed by points added to their re- spective consonantal characters. There are no peculiar initial vowel signs , but the character ^ serves as a fulcrum for every initial vowel. We need not transcribe it. Final nasals are regarded rather as modifications of the preceding vowel and are for this reason usually not written at all, or repre- sented sometimes by a common sign i> placed above the syl- labic sign, and pronounced at the end of a word always w, and before other consonants n, n, n or w, according to the class of the following letter, imitating in this respect the use of the Sanskritic annsvara-point. Decidedly long vowels occur very seldom in Makassar and almost alone in foreign words. But every vowel, short or long, may be pronounced with a sudden closure of the throat after it, which corresponds very nearly with the Chinese z\ tone (see above p. 232), and which we represent best, as Mr. Laud proposes, by adding our hamza > after the vowel, as we have done it in those Semitic languages , where N or *= (hamza) closes a syllable. It seems, that in Makassar this final ', which is not written, replaces always, as in Chinese, the Samoyetic (p. 219), the eastern Po- lynesian (p. 259) and other languages, a dropped consonant, especially , which reappears, when a vowel is added, for ex. balla> and a becomes ballaka, whilst balan and a remains balana. The letter h is not found in the older Makassar writing; it has been introduced only in later times, and principally used in Malayan or Arabic words, hardly in pure Makassar words. The word-accents may be added on a much larger scale, as it is done in the standard works on the Makassar language by Dr. B. F. Matthes. DUGIS. 263 B U G I S. XX XN ^ ^ k 9 n i d' n t d n P b m y r I nk nt nr mp Remarks. The same sign, called ant'a, which in the Makassar indicates a final nasal, is in the Bugis a vowel sign designating an in- distinct sound approaching to a , and therefore transcribed by us a. Four signs are added to the Makassar alphabet to ex- press the combinations of nka^ nta, nra, mpa. There is no nta, which seems to be replaced by nra. These combinations occur also in the beginning of words and remind us of the same fact in many African languages, ng^ nd', nd, mb have no peculiar signs and when they occur in the middle of words, the nasal, as in the Makassar, is not indicated at all. We owe these remarks to the personal information of Dr. Matthes, who is scientifically and practically fully acquainted with the dif- ferent languages of Celebes, where he has lived many years, as an agent of the Netherlandish Bible Society at Celebes. 264 POLYNESIAN or MALAYAN LANGUAGES. EASTERN POLYNESIAN LANGUAGES. These languages belong to the poorest with respect to the number of sounds they use. None of those which are hitherto known, have the letter s, nor y , nor j^, nor any aspirate, nor even any media, with some rare exceptions. They have the three tenues &, , p, pronounced rather softly and even of these the Tahiti wants the k, the Sandwich the t. But we think that we have to regard not only the hamza > as a softening of &, but also v as the soft correspondent of jt>, and r or I of t; for we find in the language of New Zealand r and d changing with one another; the Sandwich has /, but no r, the Raro-Tonga has b, but no u, and other languages have w (perhaps w>?) instead of v. Most of them have the three nasals n, w, m, and besides h. Some distinguish f and 7t, which, however, change most frequently with one another; the Raro-Tonga and the Gambier have neither. We possess an instructive comparison of several of these languages by B. Gaussin (Du dialects de Tahiti, de celui des tics Marquises et en general de la langue Polynesienne, Paris. 1853). According to this work, we give the following alphabets in our transcription. The vowels are in all the same: a, e, i, o, u. They are very rarely decidedly long, a, , 7, e, ??. Besides the short and long, Mr. Gaussin distinguishes also two accents, which he calls "grave" and "aigue" without describing them nearer nor designating them by peculiar signs. The consonants of the different languages treated by him are as follows. The feeble guttural, which he calls "explosive pharyngienne" or "postero-gutturale" seems to be our "hamza" >. iV e iv - Z e alan d. Raro Tonga. w n n m h n n m EASTERN POLYNESIAN LANGUAGES. 265 G am bier. Tahiti. n n m n m h f Marques as^ north-western part. south-eastern part. n n m h n m h J Sand w i c h. The general system of consonants of these different languages would therefore be the following: m h f Specimen of the Tahiti language. Ua hoe e toopiti tau ta'ata i tai e hi i te i>a; >o Roo te i>oa o te tahi, >o Teahoroa te i'oa o te hoe. Ua tivu i ta raua matau n raro } i te moana; fifi atura te matau i te rouru o taua atua ra o Ruahatu; par an ihora raua: e i>a; } ua huti ihora e fa'atata aera 4 te pae va>a, hio ihora raua e ta>ata, te mavera te rouru. Gaussin, p. 255. 266 AUSTRALIAN or PAPUAN LANGUAGES. ILLITERATE LANGUAGES, AUSTRALIAN or PAPUAN LANGUAGES. SOUTH AUSTRALIAN. (Adelaide.) a a k n y e o t n r I i it P m w ai au 01 in Remark s. We follow the work of T e i c b el m a n n and S c h ii r m a n n of the Lutheran Miss. Soc. , Outlines of a Grammar, Voca- bulary and Phraseology of the aboriginal language of South Atistralia, spoken in and around Adelaide. 1846. The authors conform to the Standard Alphabet with the exception of our M, for which they write ng. A N N A T M (New Hebrides). a K g n e o t d n i u p b m f v w is dz ny Vide: H. C. von cle r Gabelentz, Die Melaneslschen Sprachen. 1860. p. 65. K s f t. y r I MARE. VITI. 267 Specimen. Is eteug natimi is ethi. is ero inhalav atamain o un. Is ika a inharei ehelc ettnan : At ekmak, alupai nyak nahaidzi inpadiain unyum imtak. Is atiakoai ira ran inpadiain o un aien. Is wat ti pan itag nodiat is eti alupat, is astsapig nain nidzi ilai asena o un a inharei , um atna o un, um apan antaka pege itag agen, is um eduaraprap inpadiain o un aien an nedo auati ahnan. MARE (Loyalty Islands), a k g n % h e o t d n s z i u p b m - v ts dz ny See: von der Gabelentz, Die Melon. Spr. p. 170. y r w kh th ph Specimen. Kei Makaze dzie onome, kei nuponi ko. Inu Yehova ono re Makaze nupo, inu toedzakore nupo wanei Aiphiti, na yara me- nene. Nupo ake iro ne makaze menu, ekewe ne inu. Exod. 20, 1. VITI (Fiji Islands). a a e e o o i I u u k ng t nd - mb n n m y r w p, /, in foreign words. S2 268 AFRICAN LANGUAGES. Specimen. E ndua na taniata sa lewe rua na luvenatahane. A t sa kaya oei tamana ko koya sa none cei rau: Tamangu, solia mai vei au nai votavota ni yau sa rota me nongu. A sa votu vei rau na nona yau. A sa tawa vunga na bofii sa mart sa songona- i-iit kpi ohu ini 1 r fu wo. KP m A to 1 idzo 1 melokd li eht eyi, eyi omo'kori aburo ko' ohu yboybo ti o 1 ni d:o' , o' si* mu oW ?v v kpo lo'h si' ilu o^kere ; ni ibe^ ni o 1 ybe 1 na> yboybo ini' re\ ni inukuna. Lake 15, 11-13. Remarks. We owe our acquaintance with the Yoruba language prin- cipally to the valuable works of the native Missionary Rev. S. Crowther (Adza^ye) (Grammar and Vocabulary 1852) and lastly of the Rev. T. J. Bo wen (Grammar and Dictionary. 1858, published by the Smithsonian Institution). Their alphabet exhibits the letters a, e } e, 0, arj(an)^ fj(n^ ny), s (s)^ z, c or fe, dz (}) instead of our e, o, u, n, c instead of our c, o ? M, h, %. If he says p. 9: "The combination wy is to be con- sidered rather as a simple sound, intermediate between w and ?y", this peculiar sound is perhaps the same, which we have written w in the preceding languages. He distinguishes (p. 1) "a third class of vowels , which cannot be said to be long, and yet are different from the short vowels. They are sounded more fully and sharply than the latter, but without the sound being prolonged, as in the case with the long vowels." lie marks those vowels by the circumflex a , <', etc. If by this description he wishes to designate the ascending tone of the cognate languages, we should write it a', c', etc, as in those languages. T E M N E. Q e o i ' u ai au oi ui k y n h y t d ft s r I s p b m f w ts gb Specimen. Wani reke katron ka nu, o wo ba tre-lomme kemmc kin, bi o Kin ka nan o sokkar, o mo tret fe a tre-gba tSanle trofat tramat ro tianle ro-kant-i, o mo konne treka ten o wo sokkar, Jid o sotto ko i? Ko bi o pon sotto ko, o botr ko ka e kenkla e yon, o ba ma-bonne. Luke 15, 4. 5. Remarks. We give the alphabet and the specimen according to Rev. C. F. Schlenker in his Temne/t 1 'rimer, Stuttgart, TEMNE. VET. 283 although his transcription seems to be in an imperfect state. He writes a, n, kh, gh, sh, zh, th^ dh. GREENLANDIC. a a e o i I u u h k g n n x r x - y t - n s z s z I P ' m f v - 290 AMERICAN LANGUAGES. Specimen. Ndlayid: mw,nf!%anu(/o } ta~a ilizimanc^mut Ps. Ill, 10. Tain((::a sxkttMutiysa, kiui^i-d! za%tup pe^Kuzzutd tazza, piumdj^tuysaK ndlaykawnut kuiiinmt. Inunil ayaytoj(umdj(p&tit. Dan. 4, 21. 22. IzumaK ajpoK , Ka- tannutaiza malugizzaganc , finnan tiy iumaj(itik , Gutip annaukku- manmatik. Acts. 7, 25. Remarks. S. Klein schmidt in his Grammatik der Gronlandischen Sprache , Berlin, 1851. supplies us with a very learned and accurate description of the Greenlandic sounds, although we must deviate in many points from his transcription of the con- sonants. Our letters: , n, j, y, &, n, j, , z, z are the same as his: A:, my, r, r, ^', ny^ y, ss, ss } s, He calls the first class gutturals, the second palatals, but be describes both classes as uttered deeper in the throat than the same classes in other languages, The second class comes very near to the common gutturals, whilst the first might be compared, as in the Getez and Amharic, to a peculiar developement of the Semitic q- class. "We distinguish those deep gutturals by adding over them, as in the just mentioned languages, the guttural point, and by doubling it over n. It is the same misconception which we have met already several times, that he takes the rough friction of ^, ^, and even n as a regular vibration, representing it by the basis r. The original explosive media is , according to this author, only preserved in the second class as y, whilst in the other classes it is either softened, between vowels, into the respective soft fricative (y , f, z, v) or, after consonants, changed even into the hard fricative (fa *, s, f). In the latter case he distinguishes the soft and the hard fricative only in the first class (r and r) and in the last (o and f), but he does not distinguish as = s from 55 = 5, nor s s from s = z. It MASSATSUSET. 291 is, however, only consistent, to show fully in our transcription the regular developement of this interesting system of conso- nants. It seems, on the other hand, that we might easier dis- pense with the compound letter dl, which this author uses for / if it follows an other consonant. As to the different accents, which Mr. Kleinschmidt employs , they seem to us not ne- cessary, if we double, after every sharp accented vowel, the following consonant. MASSATSUSET. f a k g n e o ts dz - i u t d n short and long. P b m a o Specimen. Noosun kesukwut qwuttianatatnunats koowesuonk, peyaumoouts kukketassootamoonk kiittenantamoonk ne ennats ohkeit neane ke- sukwut. Numeetsuongas asekesukokis assamainnean yeuyeu kesukok. Kah ahkwontamaiinnean nummatseseongas neane matsenehikwegeeg nutahkwontamauounonog. Matth. 6, 9 11. Remarks. The alphabet is taken from John Eliot: A Grammar of the Massachusetts Indian language, ed. by P. S. du Ponceau. Boston, 1822, and the Specimen from Josiah Cotton: Vo- cabulary of the Masts. Ind. langu. Cambridge, 1829. p. 104. There remain still several doubts about the vowels. The letters r and Z are used in cognate dialects instead of Matsaa- tsusett n. We add^- according to a remark of John Pickering, see Cotton p. 6. 292 AMERICAN LANGUAGES. a e o I R K W I S. k - t - a e o y r w Remarks. Du Ponceau, in his Memoir e sur le systeme grammatical des langues de quelques nations Indiennes de VAmerique du nord. Paris, 1838. p. 103. pretends expressly, that the Irokwois use only the poor alphabet exhibited above. According to Zeis- berger (du Ponceau, p. 259) they have also the letters ( j) hi I" y ^i A Xo) ?/i ll fJWO ho yo u Cd f/ii-u T 7m J ^r?* GT^ ^ iQ^ W ta ^ te iT fa 1 - - - - l> da $ de I f / * . A do S dla L die G dK 'y- ptt . *$* dlu P c* ma Oi me H wit Qc* wo y mit - c* wa W we W C5 wo S wu fr tr hna G waA ad s //<; Specimen. Ogidoda yaleladi hehi, galegwodiyu gesesdi dedsadoei. Dsa- gewiytthi gese wigananugoi. Ani elohi ividsigalisda hadanetesgei, nasgiya galeladi dsinigalisdiha. Nidadodagwise ogalisdayedi sgiesi gohi iga. Digesgiesigicono desgidugei, nasgiya dsidigayodsineho dsodsidugi. Ale dlesdi itdagoliyediyi widisgiyatinestanegi sgiyu- dalesgesdigwosgini uyo gesei. Dsadseligayeno dsagewiyithigesei^ dsalinigidi ale gesei, edsalegwodiyii ale gese nigohilei. Amen. The Lord's prayer. Remarks. The Chiroki or, as they pronounce themselves, Tualagi are known by the remarkable fact, that they alone of the Indian TSALAGI (CHIROKI). 295 tribes use in writing and printing vernacular characters, in- vented about 1823 by a Chiroki man called Seyivoya or with his English name George Guest. The history of this extra- ordinary invention, will be found, after an authentic relation in a Chiroki newspaper, in the Notes to the treatise of John Pickering on the Indian languages of America, translated into German by the learned Mrs. Th. Robinson (Talvj). It is in- teresting to observe, that the inventor, who could not read nor speak any other language except his own, did not proceed to the separation of vowels and consonants, but set up a syl- labarium of 85 characters, uniting 15 consonantal sounds with six different vowels and giving besides a peculiar sign to every pure vowel, lie omitted those combinations, which he did not actually meet in his language, and he added a character for the syllable nah (= w?), probably because this syllable con- stitutes the only monosyllable Chiroki word (except some in- terjections), a second one for the syllable hna, perhaps because the aspirated n seems to be the only aspirated consonant which occurs in the beginning of words, and a third for the simple vowelless s, which precedes several other consonants, for ex. st, sd, sk, sy^ syw. The sixth vowel (e) has been described as a nasalized English u of but. It seems not impossible that it may resemble the hard i of the Chinese and the Tatarian lan- guages, which would easily be decided by a Russian linguist. We should propose in this case to render it likewise by * instead of e. As to the consonants, we are of the opinion, that the Tsalayi language has no true sonant Mediae, but that our second row contains the real dry Tenues, which we ought to write kw, k, t, to, tl, and the first row the aspirates //, f, fl. We conform however to the already received orthography," whose deviation from the true pronunciation is all the less im- portant, because no third row has been developped in the Tsa- layi system. J. D. W of ford in his American Sunday School Spelling book, translated into the Cherokee language, New York, 296 AMERICAN LANGUAGES. 1824. represents all the rows of vowel combinations, also those with &, t, tl, m, as complete, and adds even several conso- nantal sounds which by Segwoya were not distinguished ; for he gives besides /, tl and dl a second /, rendered by 7, which he compares with the Welsh II, our ?, and which occurs also in the combinations tl and dl. Moreover, he uses an aspiration expressed by > before vowels and almost all the consonants (>(/, >gw, J k, >g, >y, Ids, ', cl, 'n, >s, ('1)) except only the labials m and w and distinguishes 'y from hy. In a single sheet in 4, containing the Cherokee Alphabet, we find the remark, that "in some words g, I, n, d, w and y are aspirated, as if preceded by /i." We render in our specimen this aspiration provisionally by '. With regard to the vowels, he writes v for the English u of thus, and v for the same vowel nasa- lized. He employs also occasionally two accents placed over the vowels, a, e, ?', (d), u and a, e, (i), 6, u, without, however, any explication. He also gives, besides the consonantal com- binations with s, several others in his texts, as Jcl, hy, wh> nt, nd, ntl, nn, and we are at a loss to know, how such vowelless consonants might be expressed in the Chiroki writing. It seems, that Segwoya did not provide at all for those cases, but that at present the diacritical sign . is prefixed before such consonants which have lost their vowel-sound. This is at least the case in the Cherokee Hymns , compiled from several authors, 8 th ed. Park Hill, Mission Press, 1848, as we see from a note added to p. 2 of those Hymns , as well as in The Gospel ace. to Matthew, translated into the Cherokee language. 5 th ed. Park Hill 1850, from which we have taken the Spe- cimen. Chiroki scholars will in these respects complete our 'transcription. DAKOTA. 297 DAKOTA. a e o i n d e I o u ' - - h - k g k - x y ts - ts - S Z y t d t n S Z (D p b p m - - w Specimen. Witsasta ivd tsihitkn ndpa\ ukd hakakta kt he atknku kt hetsiya: Ate, woyuha mitdwa kte tsl he mitsmco, eya. Ukd wo- yitha kt yudkipam witsdku. Ukd iyohakam dpetu tonana, tsihttku hakakta ko he owdsi witaya tpahi, ka itehdyd makdtse wet ekta itsimani ya; ka hen si^d o^dydpi kl o, tdku yuhe tsl owdsT hdutdkunisni. Ev. Luc. 15, 11 13. Remarks. The distinction of long and short vowels is not clearly fixed. The sound of > in s>a, p>a, bot>d,.kap>Z^ kas't^ etc. is that of the Arabic hamza. Of the sounds, which we write , ts, t, p, the Rev. S. R. Riggs, Missionary of the American Board, says in his Grammar and Dictionary of the Dakota lan- guage, Washington, 1852, that they are pronounced "with a strong pressure of the organs, followed by a sudden ex- pulsion of the breath." Others call this pressure a sort of aspiration. We believe that the pronunciation of these sounds is the same as that of the corresponding Khetsua sounds, and write them accordingly (see below). The assibilated palatals ts and ts seem to be mostly derived from gutturals. The letter I occurs regularly only in the Titowd dialect, replacing d or n of the other dialects. Mr. Riggs uses arj , erj, etc., c, k, c, , 50, h, a, s, c instead of our a, e, etc., ts, &, , -*> P> X> ft *> L 298 AMERICAN LANGUAGES. T M I. e h ad / g n X e e o ts - ny - y i i u t d n s z r d e 'I o u p b m f ~ Remarks. See Grammatical della lingua Otomi dal conte V. Piccolo- mini. Roma. 1841. KHETSUA (QQUICHUA, PERUVIAN). a k kh k - e o ts tsh ny i u t th t n p ph p m h h (/) r Specimen. Tsaipatsapi Jesus yatsatsiskankunamanmi nyirkan: unantsa- kuna intipi, khUyapipas koilyurkunapipas kankam; runakumiri kaipatsapi mamakotsap tsaunyiinyinpa pokhtsikenkunap kurn- nyiinyinpas mantsamyinhuan lyakhirayankatn. Runakuna tukui tekkimuiup hdhuaman hamukunap mantsamyinhuan suiainyin- huanpas tsakhikuplinkii . Tsekapunim hanak phatsakunap tckkin- kuna kuiukunka. Tsaipatsapi)' i hatum atipainyinhuan , apu- kainyinhuanpas phuhuiupi hamukhta virgenpa huahuanta ri- kunkam. Ev. Luc. 21, 2527. OTOMI. KHETSUA. KIRIRI. 299 Remarks. The following letters: i; k, M, k, h; ts, tsh, ny , s, ly; th, t, s, s; ph, p, of our transcription are rendered by Goncalez Holguin in his Vocabulario de la lengua Qquichua, Reyes, 1608, and by Torres Rubio in his Arte y Vocabulario de la lengua Quichua, Lima, 1754, y; c (before a, o, u y i) or qu (before e, ?'), k (a, o, u, f) or qqu (e, V), cc or kc t h; ch, chh, n, s, II; th or tt, tt^ $ (a, o, u) or c (, i) or z (before con- sonants), fs; p or pp, pp; by J. von Tschudi, Die Kechua- Sprache, Wien, 1853: t/, c or k, k or k\ c or &', Ji ; ch, ch, n, s*, H', f ?, s, -^; p, p. The letters k, t, p, h, s are pro- nounced with a peculiar contraction of the throat, which we can only compare with the guttural emphasis of the Semitic linguals, and which we render accordingly by the same line underneath. The letter, which we write s seems to be more a sharp aspirated s, than a full s, which latter, however, is substituted in some districts. We are not sure as to the exact pronunciation of the letter written y by former gram- marians and rendered by us z; it seems to be always com- bined with other vowels. e e a OQ o i u a e etc. KIRIRI. k 9 - h ts dz ny s z y t d n s z r p b m - - w U2 300 AMERICAN LANGUAGES. OTOMI. Remarks. P. Mamiani in his Kiriri- Grammar , translated from the Portugese into German by II. C. von der G ab el entz (Leipzig. 1852), writes cc, ~; 1. 12 after (yain) add: mod. greek ya#r>i;; 1. 23 add behind w engl. ?tv?: w germ. Wwc?. P. 80, not. 2, expunge: TKe cannot till -i', and in the last line but one road: work (p. 4. 15. 34). P. 81, note, 1. 5: the clicks before these letters. P. 82, note, 1. 3: declare ourselves. P. 89, 1: into. P. 90, 10: all. P. 94, 8: nasalization; 1. 15: We; 1. 30: view of the ancient. P. 99, 4: wishes; 1. 8: ^ (ks) and U (yn)\ 1. 21 : yazdh or .; , yazzicah\ last 1. but one: KdyatJil. P. 107, 16: which^s derived. P. 109, 4: Balbodh. P. 114, 1: Posts. P. 117 put 1 \ behind -H? eb ; ]. 9: Ligatures. P. 121, 9: -C y. P. 123, 24: Cf. p. 311. P. 124, 9: y,,>; 1. 10: *,,() . /',; /; 1. 11: ,(/),. w; 4 -. P. 125, 3 d 1. from the end: Oatiya. P. 127, 21: the Romance. P. 129, 12: Cf. p. 311. P. 131, 16: (*) d(s). P. 136, Spec. 1. 1. 2: ^, se\ 1. 9: e _e i o. P. 138, 12: fc dres; 6 th 1. from below: Wayya>ax. P. 188, 3: k'a, y'a. P. 189, 2: Himyaritic. P. 191 and 192, write * instead of s. P. 198, 16: drops. P. 200, on the end: Matth. 2, 1. 2. P. 202: t d -. P. 203, 3: descendants. P. 204, 6: w; 1. 15 expunge: s. P. 205: Ta-Maseq (Mdsiy). P. 206, 12: Ta-MaSeq, Ta-Maseyt. 1*. 209, 10: w i; 1. 11: Write (V^ instead of (V), and likewise p. 210; after the Specimen add: Langles, Alphabet Mantschou, POSTSCRIPT. 315 p. 158, sqq. P. 218, 12: r I 1. P. 233, 1: Yeu r thai\ P. 243, 15: it *, if. P. 25*2, 4 th 1. from below: The letters .... where g i is. P. 253. According to Prof. Schiefner's Versuch fiber die Tkusch- Sprache oder die Khistische Mundart in Thuschetien ill the Me- moires de T Acad. des Sc. de St. Petersbourg, VI. Ser. Sc. Polit. etc. T. IX, (in which he writes #, f, p instead of M, th, ph) it seeins that his c and c belong to our third , c and c to our first column. P. 252, 6 f. b.: feeble; 1. 1 f. b.: pronounceable. P. 255, 2 d 1. from below: # d s z. P. 271, 5 f. b.: newly; 1. 2 f. b.: feel. P. 275, 14: Clarke (Baptist Miss.). P. 270278: Yonila instead of Yoritba. P. 280, 3: blew. P. 288, 4; Kon- gara, and likewise 1. 4 from below. P. 292, 3 from below: Miss, of the Amer. Board of Comm. P. 295, 23 : of the Chi- nese, the Tatarian and Slavonic languages. Berlin, printed by C. and F. Unger. A 000687364 o