Pronunciation of Sanskrit words
Here you will find:
- An essential explanation, which will give you some basic know-how and help you to avoid common mistakes.
- An elaborate explanation, which will provide more details as to how Sanskrit syllables are pronounced.
Sanskrit pronunciation can be quite complex, but if you learn just a few basic rules, you will be able to pronounce Sanskrit words correctly for the most part and will avoid the most common pitfalls. The following two explanations provide some helpful guidelines, but for a complete explanation it would be better to look for a teacher.
Since the Sanskrit alphabet consists of a number of letters and sounds that do not exist in the Latin alphabet, certain additional signs - so-called diacritics - are required in the Latin script for the representation and transliteration of these sounds. In Sanskrit each letter represents one, and only one, sound. In English the letter a for example may indicate many sounds (e.g. fat, fate, fare, far) but this is not so in Sanskrit. Sanskrit follows very consistent rules and pronunciation and contains no “silent letters” as is common in English.
There are five different kinds of diacritical signs:
- a horizontal line on top of some vowel. E.g. ā
- a dot on top of the n. E.g. ṅ
- a dot and a half-moon circle on top of m. E.g. m̐
- a dot underneath some letter. E.g. ṭ
- a tilde (~) for the palatal nasal sound ñ
- an accent for the palatal sibilant ś
- Each syllable receives approximately the same emphasis; vowels are lengthened rather than stressed.
- A horizontal line on top of a vowel (e.g. ā) indicates a long vowel. Long vowels are held for about twice the length than their corresponding short vowels. E.g. a is pronounced like the "a" in "fat", and ā is pronounced like the "a" in "father" or as in "harm". Examples are the words Tathāgata or Padmākara. Here, the emphasis lies on ā.
- The letters e, o, ai, and au are counted as long vowels and hence the vocal length is prolonged as well. An example is the word Vairocana. Thus, ai and o are held longer than the two following short a's.
- The letter ṛ is counted as a vowel in Sanskrit. The sound of ṛ is a combination of "r" followed by a short "ee"-sound, e.g. as in "rich", unlike "reef". An example is the word Amṛta.
- A dot on top of n equals the "ng"-sound in wrong. Examples are the word Saṅgha
- A dot and a half-moon circle on top of m̐ also equals the "ng"-sound in wrong. Examples are the syllables om̐ and hūm̐. Although om̐ takes the m̐ it is nevertheless pronounced "om" rather than "ong". hūm̐ however is thus pronounced as the Tibetans would as "hung".
- A dot underneath for reflection. In the case of the letters ḷ, ṭ, ḍ, ṇ, ṃ, the difference is too subtle, so we can neglect this and pronounce the letter as if there was no dot.
- The letter ḥ is an unvoiced breath following a vowel. An example is the syllable āḥ.
- A tilde (~) for the palatal nasal sound ñ. This sounds equals ny, like in canyon. An example is the word Mañjuśrī.
- An accent for the palatal sibilant ś equals a "sh"-sound, like in fresh. Examples are the words Śūnyatā, Śākyamuni or Śāripūtra.
- The letter ṣ is very similar to ś. An example is the word Śeṣa.
- The aspirated consonants (kh, gh, ch, jh, th, dh, th, dh, ph, bh) are pronounced as the consonant plus a noticeable aspiration of breath.
- An apostrophe (') at the beginning of a word stands for a "half a". It is either pronounced as a very short a sound or dropped. An example is puṇye 'parimita.
- An additional important point for English speakers is that the Sanskrit consonant ca is pronounced like the ch in chip and not like the ca in cat. Examples are the words cāmara or Cakrasaṃvara.
Commonly used conjunct consonants, that is a combination of two or three consonants, are:
- kṣa pronounced kscha. Examples are the words rākṣasa and kṣatriya.
- tra pronounced like the tra in trap. An example is the word mantra.
- jñā is pronounced "j-nya". Examples are the words jñāna (pronounced j-nyana) and prajñāpāramitā (pronounced praj-nya-paramita) 
Sanskrit is made up of 49 phonemes, that is distinct units of sound. These can be grouped into thirteen vowels, thirty-three consonants and two extra sounds.
The Thirteen Vowels:
The vowels, that is, sounds that can be voiced on their own, are: a (as the "u" in but), ā (as the "o" in mom), i (as the “i” in bit), ī (as as the “ee” beet), u (as the “u” put), ū (as the “oo” in pool), ṛ and ṝ (as the “ri” in rig), and ḷ and ḹ (as the “L” in sickle). The diphthongs, that is combined vowel sounds, are: e (as the “a” in gate), ai (as the “ie” in pie), o (as the “o” in go), and au (as the “ou” in loud).
As we have seen like the Roman alphabet, the Sanskrit alphabet has the vowels: a, e, i, o and u. In addition, Sanskrit adds long vowels of ā, ī and ū. Furthermore, ai and au are added. What may come as a surprise to us, is that Sanskrit also has ṛ and ḷ and their corresponding long forms ṝ and ḹ as vowels.
The Thirty-three Consonants:
Similar to the Tibetan alphabet, the Sanskrit consonants are arranged in five groups according to where they are produced, beginning from the throat and moving forward toward the lips:
- The Velars or Gutturals, produced in the throat, are k, kh, g, gh, and ṅ. The first, k, is a hard guttural. The second, kh, is aspirated, because it pushes out air as it is sounded. The third, g, is a softened guttural. The fourth, gh, is soft and aspirated. And the fifth, ṅ, is a nasal guttural that sounds like the “ng” in wrong. This pattern is followed for the subsequent groups.
- The Palatals are produced at the rear of the mouth, by the palate. They are c, pronounced like the “ch” in chip, ch (aspirated), j, jh, and ñ (pronounced as “ny” in canyon).
- The Cerebral or Retroflex consonants are produced by curling the tongue to touch the roof of the mouth. These are ṭ, ṭh, ḍ, ḍh, and the nasal, ṇ.
- The Dentals are produced by having the tongue touch the back of the teeth. These are t, th, d, dh, and n.
- The Labials are sounded with the lips. These are p, ph (as the p-h in cup-handle), b, bh, and the dental nasal, m.
Furthermore, there are four semi-vowels: y, r, l, and v.
Then, there are also three sibilants: ś, pronounced like sh in shade, ṣ which sounds similar to ś, but which is a retroflex produced by placing the tongue at the roof of the mouth, and s. Finally, there is a voiced aspirate, h.
The Three Extra Sounds:
The three extra sounds are: an unvoiced aspirate, ḥ, known as the ‘’visarga’’, which echoes the preceding vowel, ṃ, known as the ‘’anusvāra’’, which nasalizes the preceding vowel, and m̐, known as the "anunāsika", which creates a strong nasalization of the preceding vowel.
- Goldman Robert. Devavāṇīpraveśikā: An Introduction to the Sanskrit Language. Delhi: Motilal Banarsidass, 2011.
- Deshpande Madhav. Samskrta-Subodhini: A Sanskrit Primer. Center for South and Southeast Asian Studies, University of Michigan, 1997.
- Rodrigues Hillary. Hinduism: The Ebook. JBE Online Books, 2006.
- Charles Wikner. A practical Sanskrit Introductory. June, 1996.
- Depending on the area jñā may also be pronounced "gya". Examples would be jñāna (pronounced gyana) and prajñāpāramitā (pronounced pra-gya-paramita)
- This explanation is following the outline of Rodrigues Hillary. Hinduism: The Ebook. JBE Online Books, 2006. If you wish to have an even more detailed explanation you may want to read either Goldman's or Deshpande's introductions to Sanskrit (See Reference Section).
- The vowel ṝ has no English equivalent. It is pronounced like ṛ (as in rig) but the sound is held twice as long.
- Although the letter ḹ is counted as a vowel, it is not used in any Sanskrit word.