Standards for Representing Tibetan Pronunciation in Roman Letters Asian Classics Institute, New York February 1998 A. Mission statement 1) In establishing standards for translation it is important first of all to define why we are engaged in a translation at all. The sole purpose of translating Buddhist texts is to create an English version which, when read by the defined audience, will most easily and efficiently inspire them to undertake the necessary action to achieve their temporal and ultimate happiness. B. Defining the audience 1) Different standards may be appropriate for different audiences. Before undertaking the translation of any particular work, it is necessary to define the intended audience. 2) Three different audiences may be defined for translations of Buddhist texts: a) Audience type: professional, English-speaking university scholars Approximate population: about 200 persons in Europe, North America, and Australia Appropriate translation style: frequent use of words from original languages; technical transcriptions; annotations including comparisons to Western scholarly tradition and current Western scholastic opinion; extensive bibliographies; avoidance of devotional language, oral tradition, and vocabulary or ideas which are not currently "politically correct" Accomplishment of desired affect: chances for major conversion of this population are unlikely Example of translations: "heavier" translations of A. Wayman, J. Hopkins, R. Thurman b) Audience type: "professional" dharma students associated with Gelukpa Tibetan Buddhist organizations in English-speaking countries Approximate population: about 3,000 persons in Europe, North America, and Australia; perhaps half of these affiliated in some way with the FPMT organization of Lama Yeshe and Zopa Rinpoche Appropriate translation style: utilization of about 100 words from original languages which have gained usage among this population; technical transcription normally in appendices, otherwise some kind of simplified pronunciation, sometimes with diacritical marks; annotations restricted to explanations of the lineage lamas or accepted commentaries; brief bibliographies and glossaries; devotional language or oral tradition acceptable Accomplishment of desired affect: population is already well- disposed towards Buddhist practice and will likely take some major part of the translation to heart; some danger of "saturation" in given subjects Example of translations: entire series by Geshe Kelsang Gyatso, works by Geshe Rabten or Geshe Ngawang Dargye c) Audience type: persons in English-speaking countries who are seeking a spiritual path but have little exposure to Buddhism Approximate population: perhaps 20 million persons in Europe, North America, and Australia Appropriate translation style: translate foreign words wherever possible; rely on simplified pronunciation that will allow a normal person to make the closest approximation to the original sound possible without the use of diacritical symbols not occurring in normal English; annotate any unfamiliar word or idea with a brief, simple explanation based on authoritative scripture or oral tradition; include brief biographies and glossaries; devotional language and oral tradition acceptable except where they may appear directly critical or offensive in existing cultural milieu Accomplishment of desired affect: perhaps one in five or ten persons from this audience will get any lasting benefit from reading a translation, but this is still a potential of 2 to 3 million people Examples of translations: the Dhammapada in the Penguin edition; approximately 20 translations available of His Holiness the Dalai Lama's lectures and writings 3) The suggestions in this document relate primarily to the "normal," English- speaking, uninitiated audience of the third group above, and somewhat to the second audience. C. System for simplified pronunciation of Tibetan and Sanskrit 1) The goal of simplified pronunciation is to allow a normal, uninitiated English speaker to make the sound which most closely approximates the original word without the use of special marks which are not utilized in normal written English. 2) It is important to distinguish pronunciation from transliteration. The sole object of a pronunciation system is to approximate a sound most closely. A transliteration system must represent every written symbol used in the original word. Many pronunciation systems are defective and fail to produce the closest audible approximation due to a failure to distinguish between these two functions. This failure is responsible for the incorrect spellings for example of the names of Zopa Rinpoche, Khen Rinpoche Geshe Lobsang Tharchin, and the capital city of Tibet, Lhasa. 3) In a case though where a misspelling has become sufficiently standard usage, it should be retained, as in the previous three names. 4) It is important to recognize that a large part of the Tibetan alphabet has no English equivalent (for example, the first five letters of the first column the unaspirated consonants, and the first five letters of the third column the low voiced consonants except in conjunction with a superjoined or prefix letter). This means that different English speakers will "hear" those letters that have no English equivalent as being close to different English letters, and this is precisely the reason for the variety of systems in current usage (for example, Prof. Hopkins using "Dzong-kha-ba" and Prof. Thurman using "Tsongkhapa"). Because in many cases there are no exact English equivalents, there will never be a system which is clearly more "correct" than another one. Nonetheless, it is incumbent upon any translator always to be consistent in the method they choose ultimately to utilize for pronunciation. 5) The system outlined here is based on the most widely-used dialect in Tibet, the Lhasa dialect. The mixed usage of different dialects in the past is responsible for spellings such as "Kanjur" (for Kangyur), "Tenjur" (for Tengyur), and "z" for [ZA]. 6) The following are specific rules for the simplified pronunciation of Tibetan. A practical chart for immediate usage is found at the end of these notes, which present rationalizations for the information found in the chart. Tibetan letters appear in brackets, in ACIP transliteration. 7) The use of an additional h to distinguish between aspirated and non-aspirated consonants is an error. English has essentially no equivalents for the first (unaspirated) column of the Tibetan alphabet ([KA], [CA], [TA], [PA], and [TZA]); all English equivalents are the aspirated versions of the second column ([KHA], [CHA], [THA], [PHA], and [TSA]). Respective examples are the initial sounds in the words "cop," "chop," "top," "pop," and the tsa sound in "pizza." It is therefore illogical to add an h to indicate aspiration with English letters which are by definition already aspirated; conversely, it is meaningless to attempt to indicate that if the h is absent the reader is supposed to pronounce the consonant as unaspirated (which normal English speakers cannot do). Finally, it is inconsistent to add this additional h to t and k, and then fail to add it to p, ch, or ts. Therefore for example we suggest "Je Tsongkapa" rather than "Je Tsongkhapa," and "tarpa" rather than "tharpa." 8) Despite the foregoing, the additional h in the column of Sanskrit letters such as the dh in "dharma" or the gh in "sangha" has become so prevalent that it should be retained, even though it has no functional value (no English speaker pronounces "dharma" any differently than "darma"). 9) The combination tz is used only to distinguish between [TZA] and [TSA] in transliteration and has no place in pronunciation. Therefore "Je Tsongkapa" rather than "Je Tzongkapa." The same is true of [Z] and [S], and s should be used for both of these. 10) The treatment of vowels in Tibetan is difficult and highly prone to the "transcription confusion" syndrome. Following are the closest equivalent sounds in English: [A] as an open vowel or with "neutral" suffix letters such as [GA] or [NGA]: use a as pronounced in "ball" [A] with "umlauting" suffix letters such as NA or LA: use e as pronounced in "get" (please note that "umlauting" suffix letters are defined in the chart below) [I] except as a final genitive additive: use i as pronounced in "sing" [I] as a final genitive additive: use y as pronounced in "way" [U] use u as pronounced in "tune" [E] use e as pronounced in "get" [O] as an open vowel or with "neutral" suffix letters: use o as pronounced in "woke" [O] with "umlauting" suffix letters: use u as pronounced in "tune" (example "sunam" rather than "sonam" for [BSOD NAMS], and "supa" rather than "sopa" for [BZOD PA]) 11) Since English vowels can be pronounced in a variety of ways, and also in order to prevent the mistake of interpreting a final e as a silent English e creating a long vowel in the syllable before it, the following brief summary could be given in the introduction to translations: a is pronounced as the one in the word "ball" e is pronounced as in "get" i is pronounced as in "sing" o is pronounced as in "woke" u is pronounced as in "tune" y is pronounced as in "way" 12) The presence of a "prenasal" sound which is sufficiently noticeable should be reflected in pronunciation. This is normally when they occur as subsequent syllables in multi-syllable words preceded by a syllable with an open vowel (example Ganden for [DGA' LDAN,] or menda ("rifle") for [ME MDA',]). Prenasals occur as specified in the chart below, whenever a third-column letter of the first five rows is preceded by the prefix letter ['A] or [MA], and also in the combinations [ZL] or [LD]. 13) Those few Tibetan words which have special pronunciations outside of any regular rules should be spelled as they are normally pronounced; for example, dorje for [RDO RJE]. 14) Silent suffix letters such as [DA] or [SA], or silent additives like the wasur are obviously not reflected in pronunciation, except insofar as they affect the vowel (such as with the "umlauting" suffix letters in the words le [LAS] or ne [NAD]). 15) Hyphenation should be avoided as much as possible, as should anything that does not resemble a normal English word, distracts the reader, or prevents easy communication. Cases where hyphenation is necessary are outlined in the chart below; that is, where the lack of hyphenation would cause a misreading before subsequent syllables which begin with Tibetan letters requiring more than a single English character. Examples: penyun for [PHAN YON] but ge-nyen for [DGE GNYEN] (which would otherwise be misread as gen-yen); and kunga for [KUN DGA'] but te-ngen for [LTAS NGAN] (which would otherwise be misread as ten-gen). Cases which are not ambiguous do not require hyphens; for example, rangga for [RANG DGA'], since it can only be read one way. 16) For audiences #2 and #3 above, the use of Sanskrit transcription for pronunciation is unnecessary and often misleading; it causes, for example, many beginning dharma students to mispronounce the words shunyata (pronounced as sunyata due to ignorance of the diacritic's function) and bodhichitta (pronounced bodhikitta due to misunderstanding of the single c used in transcription for a non-aspirated palatal). Virtually no Westerner, scholar or not, pronounces the short and long Sanskrit vowels differently from each other, so the use of macrons for pronunciation is meaningless. As such, the transcription should be converted to pronunciation following the rules outlined in the chart below. Transcription is appropriate in glossaries (in conjunction with pronunciation), and in bibliographies, which are primarily intended for the scholar. 17) The pronunciation of mantras in Tibetan texts poses a special problem. In some cases the Tibetan pronunciation is correct but does not match the conventions of modern Sanskrit scholars (for example, radza rather than raja for "king," and wadzra rather than vajra for "diamond"). In other cases the Tibetan pronunciation appears clearly mistaken (for example, the retroflex dental "t" for a whole variety of combinations such as kr or gr). We cannot be completely sure though that the way in which Tibetans pronounce Sanskrit is not the way in which it was pronounced in India when the transmission was being made to Tibet. It would be silly for example to give Tibetan pronunciation using all the silent prefix and suffix letters (as it was most likely once pronounced), rather than as it is actually currently pronounced by native speakers. The current pronunciation of several Indic languages indicates that many of the seemingly peculiar Tibetan pronunciations are more correct than the system of Western scholars, which simply follows the spelling. The solution then is as follows. For texts closely related to the oral tradition (for example, in a work based on an eminent Lama's lectures, or in a ritual meant for recitation in a group), the pronunciation of the living masters of the lineage should be used. If necessary, a short note at the beginning should explain that this is being done. Otherwise Western students would, for example, pronounce phrases differently than the chanting master in a ceremony led by a Tibetan. In more literary translations, including material that was not likely to be recited in a group, the pronunciation could be presented based on the current usage of Western scholars of Sanskrit; that is, following closely the theoretical pronunciation based upon the full spelling. 18) For consistency in phrasing, the Tibetan phrase marker she [SHAD] should be represented with a comma, and with a period to represent termination at the end of a defined verse, major section of a text, etc. 19) Needless to mention, combinations whose pronunciation equates to other letters are treated as if they were those letters; for example, ny for [MY], j for [BY], dr for [BR], shomdre [ZHO 'BRAS], yarka for [DBYAR KHA] SUMMARY CHART OF PRONUNCIATION RULES [K] = k example kawa for [KA BA] [KH] = k example Tsongkapa for [TZONG KHA PA] [G] as main letter, except in a prenasal, subsequent syllable = g example gawa for [DGA' BA] [G] as main letter, in a prenasal, subsequent syllable = ng example renga for [RE 'GA'] [G] as suffix letter = k example Gelukpa for [DGE LUGS PA] [GH] = gh example Maghada for [MA GHA DA] [NG] except as ambiguous beginning letter of a subsequent syllable = ng examples nga for [NGA] and lang for [BLANG] [NG] as ambiguous beginning letter of a subsequent syllable = -ng example ka-ngel for [DKA' NGAL] [C] = ch example chungse for [CUNG ZAD] [CH] = ch example chu for [CHOS] [J] = except in a prenasal, subsequent syllable = j example jarin for [JA RIN] [J] in a prenasal, subsequent syllable = nj (technically this should be nyj, but this would be too difficult and distracting for Western readers) example tanjuk for [MTHA' MJUG] [NY] except as ambiguous beginning letter in a multisyllabic word = ny example nyelwa for [DMYAL BA] [NY] as ambiguous beginning letter in a multisyllabic word = -ny example cha-nyam for [CHA MNYAM] (Sanskrit retroflex series:) [t] = t example tikchen for [t'IK CHEN] [th] = t example kota for [KO thA] [d] as main letter, except in a prenasal, subsequent syllable = d example damaru for [dA MA RU] [dh] = dh example uttarashadha for [AUTTA RA shA dhA] [n] = n example panchen for [PAn CHEN] [T] = t example ten for [GTAN] [TH] = t example tarpa for [THAR PA] [D] as main letter, except in a prenasal, subsequent syllable = d example den for [GDAN] [D] as main letter, in a prenasal, subsequent syllable = nd example Ganden for [DGA' LDAN] [DH] = dh example sindhura for [SINDHU RA] [N] = n examples ne for [GNAS] or len for [LAN] [P] = p example pawo for [DPA' BO] [PH] = p example pentok for [PHAN THOGS] [B] as main letter, except in a prenasal, subsequent syllable = b example bardo for [BAR DO] [B] as main letter, in a prenasal, subsequent syllable = mb example chumbep for [CHU 'BEBS] [B] as suffix letter = p example raptu tsawa for [RAB TU TSA BA] [BH] = bh example Sambhota for [SAM BHO tA] [M] = m examples menpa for [DMAN PA] or rimpa for [RIM PA] [TZ] except as ambiguous beginning letter in a multisyllabic word = ts example tsukpu for [GTZUG PHUD] [TZ] as ambiguous beginning letter in a multisyllabic word = -ts (this is so rare as to be a nearly unnecessary distinction, since almost no Tibetan syllable ever ends in a t sound) [TS] except as ambiguous beginning letter in a multisyllabic word = ts example Tsarchen for [TSAR CHEN] [TS] as ambiguous beginning letter in a multisyllabic word = -ts (equally rare) [DZ] as main letter, except in a prenasal, subsequent syllable = dz example dze for [RDZAS] [DZ] as main letter, in a prenasal, subsequent syllable = ndz example sandzin for [GZA' 'DZIN] [W] = w example wa for [WA] [ZH] = sh example shenpa for [ZHEN PA] [Z] = s example sampa for [ZAM PA] ['] as main letter = vowel alone see vowels below example oma for ['O MA] ['] as suffix letter in a dipthong = subsequent vowel alone, except when [I] example yin pao for [YIN PA'O] ['] as suffix letter followed by genitive additive = y example lamay tuk for [BLA MA'I THUGS] [Y] = y example yarwa for [G-YAR BA] [R] = r examples rimpa for [RIM PA] and tsarwa for [TSAR BA] [L] = l examples lelo for [LE LO] and selwa for [GSAL BA] [SH] = sh example shepa for [BSHAD PA] [S] = s example sipa for [SRID PA] [H] = h example hla for [LHA] [A] = vowel alone see next [vowels:] [A] except before umlauting suffix letter = a example rangwang for [RANG DBANG] [A] before umlauting suffix letter = e example lenpa for [LAN PA] [I] = i example yi for [YID] [U] = u example lu for [LUS] [E] = e example leppa for [SLEB PA] [O] except before umlauting suffix letter = o example gongpa for [DGONGS PA] [O] before umlauting suffix letter = u example sunam for [BSOD NAMS] ("umlauting" suffix letters are defined as: [DA] [NA] [RA] [LA] [SA])