ITRANS Transliteration

In the West, Sanskrit is rarely represented in its correct Devanagari form since few people would be able to make any sense of it. Instead, it is usually written in what is called a transliterated form, which means that the actual characters of a given word are converted into an Anglicised form using the Roman alphabet. There are several methods for doing this but the one most frequently used on the Internet is called ITRANS and was devised by Avinash Chopde. His software, and details about the system, may be downloaded from his website. The problem was that the previously most widely used scheme used symbols called macrons (lines above letters) and dots above and below letters, so that it was quite unsuitable for computer keyboards with basic letters and the normal fonts provided with word processors. Accordingly,ITRANS was devised to use only the usual letters of our alphabet, together with the occasional special character such as the tilde ~.

The five basic vowels

The first letter of the alphabet forms the fundamental sound from which all others are derived simply by moving the tongue and lips. It is made by opening the mouth wide and letting the vocal chords operate. The sound which emerges sounds like a cross between the short 'a' in cat and 'u' in but. It is written as 'a' in ITRANS but the correct letter in the proper script, called devanAgarI (meaning city of the gods), is: -

This is the first letter of the alphabet and the first vowel or, to use its correct term, svara, meaning sound.

If the back of the tongue is now raised slightly towards the back of the roof of the mouth, keeping the front of the tongue down against the back of the lower teeth, and the same short movement of the vocal chord is made, a slightly different sound emerges. This sounds a bit like the short 'i' in bit. It is written as i in ITRANS.

The next two vowels seem strange to Westerners but follow the logic of the development. The underlying sound for both of these is the i sound just covered but the tip of the tongue is first moved further forward in the mouth. If you raise the tip of the tongue until it is almost touching the roof of the mouth and then, make the 'i' sound as before, the next vowel sound emerges. Modern students often actually flick the tongue downwards as the sound is made so that the result sounds something like 'ri' in the word rip, though the rolling 'r' beginning is not clearly enunciated because the tongue never actually touched the roof of the mouth. However, strictly speaking this is not correct. It is written RRi in ITRANS (or, in old releases, R^i).

This procedure is repeated but now the tip of the tongue moves further forward still, to just behind the front teeth, before the 'i' is sounded. Again, modern speakers often flick the tongue up towards the roof and down so that the sound that actually comes out is 'lri'. Again, not strictly correct but it hardly matters since there is only one word in the language that uses this letter! It is written LLi in ITRANS (or, in old releases, L^i).

Continuing the development, the emphasis finally shifts to the lips (labial position), having begun in the throat (guttural position), moved to at the back of the mouth (palatal), then to the roof of the mouth (cerebral) and then the teeth (dental). If a circle is formed of the lips but without any tension and the basic sound is made, a short 'oo' sound comes out as in soot or cut. This is the last of the simple vowels, written u.

These five vowels with their characteristic mouth positions effectively head up the five main groups of consonants. Consonants all still sound the basic 'a' but 'stop' it from coming out in that simple way by varying the position of the tongue and lips in the way dictated by the vowel at the head of the group.

The long vowels

The basic five vowels above are all short vowels - hrasva. This means that, when pronounced, the sound is made as short as possible whilst still being distinguishable - really quite short! Each of these vowels can be sounded long. The length is actually very precise. If the short form is treated as one measure, then the long form should be two measures. The long form is called dIrgha. In ITRANS, the vowel is shown as long either by putting two of them, as with the ii in diirgha or by capitalising it thus: - dIrgha. The latter form has been used as standard in this book.

When the vowels become long, the pronunciation naturally changes slightly, too. Thus, the short a becomes aa or A and sounds like the 'a' in calm. The short i becomes ii or I and sounds like the double 'ee' sound in words like sleep. The short RRi becomes RRI (or R^I in old releases of ITRANS). Here there is no option of having two small i's. The dIrgha vowel is sounded as for the hrasva form but with the ending 'ee' instead of 'i'. Similarly, the short LLi becomes LLI (or L^I in the old releases of ITRANS) but, since there are no words at all known to contain it, this hardly seems to matter! Finally, short u becomes long uu or U and sounds like the double 'oo' in root.

The vowels can be sounded for longer than two measures, in which case they are called pluta - prolonged. In this case they are written with a number '3' below and just to the right of the letter, both in devanAgarI and in the Romanised version. This form cannot be represented in ITRANS.

The compound vowels

Now, if the sound a is made and continues to sound while the mouth is slowly closed, the sound made before the lips come together is u. If these two sounds are made together or, more practically speaking, if the sound corresponding to the mid-point between these two is made, the sound that emerges is o (as in 'boat'). This new letter is called a compound vowel. Similarly, when a combines with i, it forms the compound vowel e.

If you sound a prolonged a3, and then raise the back of the tongue towards the i position, but stop before you get there, you should hear the e sound. It's a bit like the 'a' in 'hate' but not as open as we would pronounce this.

If, after making the above sound for e, you relax the tongue back towards the a position but again stop before you get there, there is another sound formed as a compound between a and e which sounds like the 'ie' in 'die'. It is written ai.

In a similar way to that described above, if the mouth moves from the a (open-mouthed) position to the o (partially closed) position but stops half-way, there is a sound similar to 'ow' in 'brown'. This is written au in ITRANS.

These, then are the fourteen vowels but there are two final letters to be added to complete the group of sixteen, so-called mAtRRikA. They are not really part of the alphabet but act as modifications to a preceding vowel. (Note that, because of this, if they are sounded as letters in their own right, they assume an 'a' before rather than after.)

The first of these is called an anusvAra. It is written as M and causes the preceding vowel to be sounded through the nose. In te Devanagari, a single dot is added over the preceding letter. The precise nasal sound is determined by the consonant that follows it, in that it uses the mouth position corresponding to that for the consonant so that the effect is something like the nasal consonant (anunAsika) described below.

The other special letter, not really part of the alphabet, also modifies the sound of the preceding vowel, is written H and it has the effect of adding a brief, breathing out, 'unvoiced' sound after the vowel. It is as though there were a word beginning with 'h' immediately following and you start to sound it as soon as you finish the preceding letter (i.e. without changing the mouth position) but then realise your mistake and stop before the word itself starts to sound. It is called a visarga. In the Devanagari, two dots are added to the right of the preceding letter.

Table of Vowels

So, to recap, the 16 vowels, or mAtRRikA, are as follows: -

The first group of consonants (guttural)

The Sanskrit term for consonants is vya~njana, meaning a 'decoration' (of the basic vowel sound). Twenty-five of these are grouped in five sets of five 'underneath' the five basic vowels described above. They are formed by positioning the mouth (tongue or lips) in such a way as to 'stop' the sound of the vowel in some way. The first group uses the mouth position of the a sound for decorating. This all takes place at the back of the mouth where it becomes the throat - the 'guttural' position.

Strictly speaking it is not possible to pronounce a consonant on its own. It is in itself only a positioning of the mouth to 'stop' the sound made by a vowel. Accordingly, when speaking the alphabet, the sound of a is used by default after each letter. The first consonant of this group is written k, sounded (with a) 'ka' as in cat.

When talking about the letter on its own, the sound 'a' is automatically assumed after it, since it cannot be sounded on its own without a vowel. Clearly it could occur at the end of the word (as 'k' in 'rack' for example). In this case it would have an additional mark under the letter, called a 'halant', which means 'don't make any vowel sound after this'. This used to be written k.h - .h after any consonant in ITRANS means that it is followed by a halant and 'a' is not sounded after it. In fact, it is no longer necessary in ITRANS to do this - if the letter is written on its own, a halant is inserted automatically.

The second consonant in the guttural group is written kh. Its pronunciation is much like the preceding one but with the addition of a slight breathy sound caused by actually letting out some air immediately following the 'k' sound. It is often sounded as though it were 'k-h' in an imaginary word 'k-hat' but there is much too much emphasis in this - it is really more subtle. Consonants such as k are said to be 'with little breath' (alpaprANa) while ones like kh are 'with much breath' (mahAprANa).

The third in this group is written g, sounded 'g' as in gap (alpaprANa). The fourth is written gh and, like 'kh' is mahAprANa and sounds like 'gh' in doghouse.

The final consonant in this group is the first of the type mentioned briefly above - anunAsika, meaning that the sound is made through the nose. It is written ~N. There are four n-related sounds; hence the need for the tilde and capitalisation. It has the sound of 'ng' made at the back of the throat and sounding through the nose, like 'sing' but with the ending further back in the throat like someone being strangled rather than singing!

The second group of consonants (palatal)

This second group forms the sounds in the back part of the mouth but not the throat. Based on the i vowel, these use the back of the tongue and the rear of the mouth; they are called 'palatal'. They follow the same pattern as the previous group (as do all five of these groups of consonants, you will be pleased to know!) in that the first and third members are alpaprANa, the second and fourth are mahAprANa, and the fifth is anunAsika. The first, then is written ch and is sounded like the 'cha' in chap but, whereas English pronounces this by using the front of the tongue near the front of the roof of the mouth, Sanskrit uses the rear parts.

The second character is written Ch (or chh in older versions). Since I have already said that the pattern of the first group is repeated in the others, you might guess that this is sounded pretty much like ch but with some added breath - and you would be right! Just remember not to make it too pronounced so that it comes out like 'ch-ha' and it should be fine.

The third is written j and pronounced more or less as would expect, like the 'ja' in 'jam' spoken as far back in the mouth as you can without injuring yourself. It uses minimum breath again as for all in this third 'row' of the main consonants. The fourth is written jh and you can work out now how it should sound - like 'j-ha' but not too much so.

The final letter in this group the second of the anunAsika characters (the 'n' type sounds made through the nose). It is written ~n and has a sort of 'ny' sound, as in canyon. However, whereas the latter is made by the front of the tongue at the front of the mouth, you have to try to make this sound with the back of the tongue at the rear of the mouth.

The third group of consonants (cerebral)

This third group has now moved the mouth position another step forward so that the tip of the tongue is used, pointing up to the roof of the mouth. To construct the main consonants, the tongue actually touches the roof. It is called the cerebral position. The first is written T. All of this third, (middle) group are written as capital letters to differentiate them from the fourth group. (In the Romanised transliteration, the letters have a dot beneath them.). T is pronounced as the 't' in tub but instead of having the tongue forward of the roof of the mouth, put it right up to the roof as you say it. That should have been spoken 'with only a small breath' as usual (alpaprANa). The second letter is the same but with more breath as you make the sound (mahAprANa), a bit like 'po-th-ole' (pothole). It is written Th.

The next is written D and pronounced like 'd' in dot but, as before, with the tip of the tongue right up in the roof of the mouth. The fourth letter sounds the same as the third but with more breath (e.g. go-dh-ead) and is written Dh. The last in the group is another 'na' sound but with the tongue in the roof of the mouth. Written, as you already know, N.

The fourth group of consonants (dental)

This group of consonants are sounded just behind the teeth and called, unsurprisingly, dental. The first is t. It is sounded just like our t, as in 'tip'. Then comes the equivalent letter, but with more breath (mahAprANa), th, as in 'butthead'. Next is d as in 'dog' and the breathy equivalent dh, as in 'redhead'. Finally, in this group, is the one sounded through the nose (anunAsika) n, as in .er 'nose'.

The fifth group of consonants (labial)

And so, at last, the final group of the main consonants sounded at the lips and called labial. The first is p just like our p, as in 'put'. Then comes the corresponding breathy ph, as in uphill. Next is b, as in 'bad' and the mahAprANa bh, as in 'clubhouse'. And finally the anunAsika in this group is m, as in 'man'.

Table of basic consonants

The table of the five groups of consonants, with the corresponding vowel shown in Column 1 for reference, is as shown below: -

The semi-vowels

There are two small groups of letters left. The first of these is the group of four so-called 'semi-vowels'. They are formed by combining the four main vowels other than 'a' with 'a'. Thus, if you sound i and then immediately move to the a sound, what emerges sounds like 'ya' and this is the first semi-vowel or antaHsthA: - y as in 'yap'. If you sound RRi and move to a, you get r as in 'rap'. If you sound LLi and move to a, you get l as in 'lap'. Finally, if you sound u and move to a, you get v as in 'wag'. Note that Americans seem to prefer to ignore this logical derivation and pronounce it as 'va' in 'van'. Since it is somewhat illogical to write it as beginning with a 'v' while sounding it as a 'w', I suppose both sides of the Atlantic have a case.

The sibilants

Almost last of all, there are three sibilants or sss-sounds. (A sibilant is called UShman in Sanskrit.) These are in the palatal, cerebral and dental positions. (In theory there are also ones in the other two positions but these are so rare that they are usually ignored.). In the palatal position, there is sh sounded by making a shh sound in that mouth position; it comes out like the ending of a soft German 'ich' with the default a ending of course. The second, in the cerebral position, is Sh (or shh in older releases) made by sounding 'sha' with the tongue up to the roof of the mouth. Finally is the dental s, sounding like the normal 's' in 'sand'.

h

This leaves the last letter in the alphabet, h, sounding, as you would expect, as 'h' in 'hat'. It is sometimes considered to be another sibilant and is also called UShman, which literally means 'heated'.

The complete alphabet

The order of the alphabet, if you want to look up a word in the dictionary, is pretty much the order used here in introducing the letters. The 16 mAtRRikA are at the beginning, followed by the basic consonants - guttural, palatal, cerebral, dental and labial. Then come the 4 semi-vowels (antaHsthA) and the three sibilants and finally h.

Further study

As you will certainly have realised by now, in order to learn how to sound these correctly, you really need to listen to someone who knows. Since it is unlikely that you will be sufficiently interested to go into it this deeply, however, just follow the instructions and don't worry about how it feels. We are unused to making full use of our mouth and tongue in speaking and, since Sanskrit makes almost scientific use, we will find much of it peculiar and initially uncomfortable.

If you want a more thorough introduction, with little in the way of expense or commitment, there is a truly excellent one available off the Internet. You can download this free of charge and the only cost is the subsequent printing. It is called 'A Practical Sanskrit Introductory' and was produced by Charles Wikner. You can access it from one of the sites devoted to Sanskrit. As well as describing the alphabet in much more detail and teaching the Devanagari script, this introduces the grammar and some vocabulary.

Finally, if you are really serious, an excellent book - 'DevavANIpraveshikA - An Introduction to the Sanskrit Language' by Robert P. Goldman and Sally J. Sutherland can be purchased from the Center for South and Southeast Asia Studies, University of California, Berkeley.

Conclusion

If you read through this a few times and refer to it when in doubt, you should be able to make a creditable attempt to pronounce most words. Of course, such a brief introduction cannot give you any real idea of the language, which is extremely beautiful in more ways than one. Here is what the Vedic prayer, given in the Introduction, actually looks like in the original Sanskrit: -

Note that all material extracted from the book is -
© Dennis Waite 2003