HINDU THOUGHT AND CULTURE COURSE HOME PAGE

USING DIACRITICS TO PRONOUNCE INDIAN WORDS

Note to new students: If you wish simply to get the basics without understanding the whole system, simply scroll down till you see the material in this typeface.

Note to web users: This page has been changed. The diacritic symbols are now inserted graphics (which is why they look a little uneven), so they should display correctly regardless of which fonts you have installed.  If you are familiar with downloading and installing fonts and wish to download fonts with diacritics that can be used in word processing, try the csx+ fonts on Sanskritist John Smith's web site. But note that I don't require diacritics in written work submitted to me.

Note to everyone: With this key, practice words that are in your readings.

As you begin your study of a South Asian religion such as Hinduism or Buddhism, it is important that you learn to pronounce Indian terms confidently and correctly. If, as you read, you start to slide over the multi-syllabic words with the lines over the vowels and the dots underneath, you run the risk of not being able to pronounce the names and concepts you are attempting to learn, or (even worse!) of not being able to learn them at all. In other words, if you don't say and hear the words you won't learn their meaning.

Basically, you will find two systems employed in transliteration (i.e. representation in Roman script)--the informal and the scholarly. The informal method is simply the use of the normal characters of English to approximate the pronounciation of words written in a South Asian script. The problem with this method is that the South Asian scripts are phonetic and perfectly adapted to representing the phonology (sound system) of South Asian languages. They include sounds which, while not particularly difficult to produce, are nonetheless not found in English, and therefore not accurately indicated by the letters of our alphabet. Further, as we all know, the spelling of English words is chaotic, and owes much to the historical development of the language. Think of the different sounds represented by the letter "a" in fat, fate and father. So when you see the letter "a" in an Indian word that is written in the informal system you don't really know how to pronounce it.

These problems are more or less solved by the scholarly system of transliteration, in which each sound in the Indian phonology is represented by only one letter in the Roman script. Since there are more sounds than letters, some letters are altered by marks called diacritics (the lines and dots, etc.) to represent additional sounds. The following is not intended to be a perfect representation of South Asian phonology, but a rough guide to the pronunciation of the transliterated words.

First, those that are likely to deceive, being unlike their usual English pronunciation:

a	like the "u" in "but," never like the "a" in "fat"
	like "a" in "father," never like "a" in "fate"
e	like the "ai" in "train," a longer vowel than in "let," and never like the e in "legal"
c	like "ch" in "search," never the "k" sound of "c" in "cat" or "c" in "city"
g	always hard as in "go," never soft as in "gem"
th	like the "t" in "table," never like the "th" in "the" or "thin
ph	like the "p" in "pin" or "ph" in "uphill." (Only in later Urdu words does it sometimes represents "f")

Letters with underdots are retroflexes; see below. There are three different kinds of s's; also see below.

The rest of the system is as follows:

Vowels
As with a and above, all vowels are short or long.
Others are:

i	like the "i" in "bit"
	like the "i" in "machine," never like the "i" in "fiber"
u	like the "u" in "put," never like the "u" in "but"
	like the "u" in "rule," never with a "y" sound at the beginning as in "future"
	is a vowel, pronounced as a short trill with an even shorter i or u sound following. Thus g Veda is written in informal transliteration as Rig Veda. o, like the "o" in "notation," never like the "o" in "pot"

Diphthongs (vowels that consist of two sounds

Consonants
First, look at these groups of consonants and (except for the warning above about c g th ph) pronounce them roughly as they are written.

(Gutturals)
k kh g gh

(Palatals)
c ch j jh ñ
Note: Besides c always being like our English ch, j is always hard, as in John, it's not like a German j (y) or a Spanish j (h) or a French j (zh).

(Retroflexes, with the tongue turned back to the top of the mouth; the underdot is the sign of the retroflex)

(Dentals, with the tongue right up against the teeth)
t th d dh n

(Labials)
p ph b bh m

Notice that the consonants come as five groups of five sounds. Each group represents a particular place of articulation in the mouth, and may be produced unvoiced (vocal chords not engaged) or voiced (vocal chords engaged), unaspirated (no breath after the sound) or aspirated (little burst of breath after the sound). The fifth member of each group is its nasal. If you reflect on the way we make sounds in American English, you will realize that within each group there are two that we never make.  Whenever we make unvoiced sounds (vocal chords not engaged) they are always aspirated, even though they are not spelled as such.  (Put your hand in front of your mouth when you say "kick" and you'll see that you're really saying "khick"; likewise with "table" or "pill." If you can learn to say these sounds without the little burst of air after the initial consonant, you've got the unvoiced unaspirated sounds--k c t p.) On the other hand, we never aspirate voiced sounds (vocal chords engaged--put your hand on your throat to feel the difference between the sound of a vowelless k and a vowelless g]. Thus when trying to explain sounds like gh or dh to speakers of American English, we must resort to tricks like "bughouse" and "ma[dh]ouse" to demonstrate the sounds in such venerable words as "Bhagavan" (lord) and "dharma" (justice, duty, religion).

The semivowels y l are as in English.
r is a very short trill, really a "flap," not a "rolled r" as in Spanish.
v is somewhere between a "v" and a "w"; in various locales in India it may veer toward one or the other.

There are three kinds of sibilants or s's:

s	the dental, as in English
	the palatal, roughly like "sh" in English
	the retroflex, a "sh" sound, but with the tongue more turned back

( and may be pronounced the same in modern usage.)

h is as in English.

represents a nasalization of the preceding vowel (as in French), but is generally pronounced as a normal "m" when followed by a consonant.

Technically, there is no stress accent in Indian words (unlike English), but we tend to accent the penultimate (next-to-last) vowel, if it is a "heavy" syllables, that is, one which contains a long vowel or ends in two consonants. If the penultimate syllable is "light" (as in Gautama), stress the antepenultimate syllable (the one before the penultimate), whether it is heavy or light. (Thus the name Gautama is stressed on its first syllable, which contains the double vowel "au" and is heavy.)

Practice your pronunciation by saying out loud the Indian terms that occur in the readings, by listening to your teachers, and by asking your classmates who know Indian languages.

These are the sounds of Sanskrit and most of the other Indian languages. There are a few others for Tamil (distinctly Dravidian sounds) and for Urdu (sounds which are the heritage of Arabic and Persian). These are generally involved in some varieties of Hinduism, but not in Buddhism.