Comparison of PTS and RS


PTS Pali
PTS Pali was invented during the days of hand typesetting and treadle platen presses. The Buddhist booklets were printed in small jobbing presses. Those days, lead types were sold by the weight in measures of fonts. Accented types were rare and even more expensive. (It seems like language evolved: types, fonts?)

The task was to transcribe Singhala phonemes into Latin letters. It is important to understand the distinction between transliteration and transcription. Indic languages cannot be transliterated to Latin because there are thousands of letters forming as consonant clusters combine with each vowel and in the case of Pali, that set combing with 'ng' sound ending to form an additional set. Singhala hodiya is a phoneme chart. It is the extension of the Sanskrit phoneme chart. Each position in it is phonetically defined. The following is the Pali subset of the mixed Singhala hodiya. Mixed Singhala (mizra síhala) means pure Singhala plus Sanskrit.

aaaiiiuuu
eo 
á 
kkhggh 
cchjjhç 
tthddhN 
þþhððhn 
pphbbhm 
yrlv 
shL 

They were assigned Latin letters (the PTS alphabet):
aāiīuū
eo 
 
kkhggh 
cchjjhñ 
ṭhḍh 
tthddhn 
pphbbhm 
yrlv 
sh 

There is an interesting rationale for selecting these letters. As said earlier, the presses were poor (the world was) and could not afford accented type. The monks were penniless mendicants -- beggars! But, in the old tradition, the monks and merchants had a symbiotic relationship. (Recall that in the past, the monks went with the merchant caravans carrying the 'relics' so that they could set up the road show where the merchants sold their wares and the monks their 'awares'.)

The printers suggested that they'd improvise the accent marks provided, they are not angular. They cut brass rules (cheap then) and made bars and dots and placed them above and below the individual types inside the leading space (line gaps).

As we see, the long (double-mora) vowels got the bars and the extra consonants got dots under. Enye (ñ) came from the foundries. This alphabet is good to write but conveys misleading pronunciation cues. The t and d with dot under are actually pronounced the same way as English t and d. The regular t and d in PTS are dental t and d closer to the sounds of the English digraph 'th'. The reason for this seemingly counter intuitive decision is that the dental t and d occur ten times or more in the Pali language than the alveolar / retroflex ones. Obviously, it was done to avoid the high typesetting cost. M with dot under is same as the English digraph 'ng'. Dotted n was already assigned and using 'ng' was not feasible. That makes the reader to think it is an alternate m.

Even today with advanced digital technology, some of the PTS letters cannot be shown on a web page without using HTML entities. We have tested the bar and dot diacritics on a keyboard that will allow you to type the accented PTS letters. We are still not sure how portable they are, though. If you check the HTML code of this page, you'll see that several letters needed to be entered as HTML entities.


Romanized Singhala
This is the Pali subset of the Romanized Singhala alphabet:
aaaiiiuuu
eo 
áíúéó 
kkhggh 
cchjjhç 
tthddhN 
þþhððhn 
pphbbhm 
yrlv 
shL 


Every letter in the above table could be typed directly using the already available US-International / US-Extended / Dead-key key layouts available with Windows / Mac / Linux machines. They are all single-byte characters that could be typed on email messages and sent across the network. They are within ISO-8859-1 character set.

As you see, we are overcoming the pronunciation confusion and using the two English letters þorn and eð for the dental t and d. ('th' and also 'y' were substituted for þ and ð when printing was introduced to England.) We made a fast Singhala keyboard that is intuitive to type Singhala. A touch-typist will find it quite easy because it is very close to the English keyboard.

Converting the existing Pali pages is easy with the program we wrote, which is available at: LovataSinhala.com/pali.php. The biggest advantage of romanized Singhala is that it is dual-scripted when the partially developed Singhala orthographic smartfont is employed. (It is partial in the sense that it does not make all the possible ligatures and not typographically consistent. It is only a proof-of-concept.)

Objections to Singhala smart fonts
There is an opinion out there that this font runs against the 'spirit' of Unicode. The argument is that Singhala letters stand in the positions of Latin characters. As far back as the earliest Unicode versions, the standard specifically stated that the character codes do not imply shapes of glyphs (though their names are like, LATIN SMALL LETTER A). It also cited Fraktur and Gaelic as examples where the glyphs are different from the expected shapes of Latin letters, but use the code positions assigned to Latin letter names.

Another spurious argument is that if Singhala uses character positions with names of Latin letters, the language of the text would not be able to be determined. We say that this is firstly romanized Singhala and further, determining the language using heuristics is more reliable, and how do you differentiate between dozens of languages that already share Latin letters? (English, French, German, Icelandic etc.)

Basically, Singhala (along with Sanskrit and Pali) are romanized through transcription of phonemes. This is how English was romanized from fuþorc (runic). There is a perfect equivalence there. As for the shapes of glyphs in the orthographic font, the service it provides is marvelous. We can now inspect Pali text converted back from Latin to verify their accuracy by simply reading the text draped with the Singhala font. Unicode failed with Indic. Orthographic smartfont technology is the only practical and comprehensive solution not only for Singhala but all Indic languages. The font provides the proper shapes for the phonemes typed maintaining the script forward in time.