Syloti Nagari — 1550? AD, Bangladesh

Posted on April 26, 2011 by ducky

Syloti Nagari “ka”

Syloti Nagari, also called Sylheti Nagari, was used in northern India, what is now the Sylhet region of Bangladesh, from around 1550 AD until the 1970s.  The region where it was used has been Muslim for a very long time, which might have had interesting effects on the writing system: Syloti Nagari “feels” much more like one of the Semitic abjads than any of the other Brahmi-derived languages.

Sylhet was conquered by Shah Jalal, a Muslim from Yemen, in 1303 AD.  More Muslims — mostly Afghans and Central Asians — immigrated to that area as first the Hindu Hemu took over, then lost northern India to the Mughal (Turkish-speaking Mongolian Muslim) Akbar.  Even during the British rule, there were still pockets of people who spoke Persian as their mother tongue in Sylhet.

Sylhet treats its vowels more cavalierly than most of the Brahmic languages.  Most of the Brahmi-derived languages have a virama (also called a halant in some languages) to “kill” vowels, but Syloti Nagari did not traditionally have a virama.  A virama crept into some — but not all — documents after about 1950 AD.

Another way that Brahmi-derived writing systems show that there is no vowel is to smoosh two characters together horizontally in a ligature called a conjunct.  In many Brahmic writing systems, most glyphs have a vertical line; in the smooshed characters, that line is removed when the vowel is removed.  In other Brahmic writing systems, the two characters are smooshed together vertically instead of horizontally.  Either way, the idea gets acrossed that the vowel gets smooshed out.  While the Sylheti language doesn’t have many consonant clusters in the spoken languages, it does have some; they don’t always use the appropriate conjunct, however.

  • Usually conjuncts aren’t used for Arabic or Persian loanwords.
  • Usually conjuncts aren’t used across morpheme (word piece) boundaries.  For example, in the word “beekeeper”, “bee”, “keep”, and “er” are all morphemes.
  • Frequently conjuncts are used for English loan words.

Syloti Nagari is completely unique among Brahmic writing systems in sometimes forming vowel+syllable ligatures.  While most Brahmic scripts will only conjoin glyphs with consonants (e.g. “ka” + “ra” to make “kra”), Syloti Nagari will sometimes join a leading vowel glyph with a syllabic glyph, e.g. “a”+”ma” to make “ama”.  Syloti Nagari will even (rarely!) smoosh together a vowel diacritic with a syllable glyph!  Smooshing vowel glyphs or vowel diacritics and syllable glyphs doesn’t make any sense if you have the Brahmic idea “you smoosh glyphs together to kill vowels”, but it does make sense if you have the Arabic/Persian idea “you smoosh glyphs together in order to make the writing more smooth or beautiful”.

Finally, Syloti Nagari poetry sometimes has characters smooshed together which by rights should NOT be smooshed together — where the vowel is NOT killed.  It appears that this is done in order to make the line width in the poetry work out.

All of these features suggests a Semitic attitude towards the reader, one coloured by exposure to abjads: that the reader will be able to figure out by themselves where the vowels go.

Links: Wikipedia, Omniglot, Unicode proposal, Climbing the Mountain of Languages

Posted in Abugida, Rating: 5 "Whoa!!" | 2 Comments

Mithilakshar — 1300 AD, India

Posted on April 25, 2011 by ducky

Mithilakshar "ccha" conjunct

Mithilakshar — also called Maithili, Mithilaksara, and Tirahut — has been used since the 14th century in the northeast part of India, although it has limited use now.  It is very similar to Bengali, so much so that early 20th century printers sometimes augmented set of Bengali metal type with newly cut Mithilakshar characters.  It became distinct from Bengali by the 14th century.

Many of the glyphs have the same form in the two writing systems, but many do not.  Some of the glyphs look the same, but are pronounced differently.  (The Bengali “ba” and the Mithilakshar “ra” use the same glyph shape, for example.)

The two languages also have very different conjuncts — ligatures of two or more glyphs that most (all?) of the Brahmi-derived languages have in abundance.  (Conjuncts are a way of both economizing on space and reducing the need for viramas to “kill” vowels when there are consonant clusters.)  For example, the consonant cluster “ccha” exists in Mithilakshar as a single glyph (see the glyph at the top of this posting) with the two parent glyphs smooshed together vertically; in Bengali, they are smooshed together more horizontally.

Mithilakshar, like many of the northern Indian scripts, fell out of common use in the 20th century due to the rise of Devanagari.

Links: Wikipedia, Unicode proposal

Posted in Abugida, now ceremonial, Rating: 4 "Huh, interesting!" | Leave a comment

Kaithi — 1500 AD, India

Posted on April 24, 2011 by ducky

Kaithi “nna”

Kaithi is Brahmi-derived, descended from Nagari. Because of the fuzziness in when Bengali split from Nagari, it’s not really clear whether to say if Kaithi is descended from Bengali or from Nagari; this means Kaithi is either sibling to Devanagari or niece/nephew to it.

Both Gujarati and Kaithi are normally written without a headline.  However, Kaithi sometimes uses a headline for emphasis, particularly of titles, in much the same way that English uses underlines.

Like Gujarati, Kaithi was used for practical matters, but instead of being used by merchants, it was originally used by governmental scribes in North India.  These scribes filled a niche filled today by stenographers and photocopiers.  Both the merchants using Gujarati and the scribes using Kaithi were interested in writing quickly, so losing the headline might have been parallel discoveries of a way to save time, or perhaps Kaithi was influenced by Gujarati.

Kaithi was used widely in northeastern India in the 1600-2000 AD timeframe, used to write many spoken languages.  In 1875, two British-run areas — Bihar and the North-Western Provinces & Oudh — standardized the Kaithi script and used it for administrative purposes.  In the 20th century, however, it was displaced by the growing popularity of Devanagari in India.  It persisted for some time — and in fact might still persist — in the Indian diaspora community for private use.

Kaithi used word separators and sentence separators, but inconsistently.  Usually the word separator was something that looked like a ‘-‘; usually the sentence separator looked like an “=”.

Kaithi scribes also sometimes used a “danda”, a form of punctuation common to most Indic scripts.  The single danda is a vertical bar in most scripts  used to separate sentences much like the “.” in Latinate writing systems.  A double danda was used to delimit the ends of verses.  (Kaithi is slightly unusual among Brahmic scripts in that its danda has serifs at the top and bottom.)

Links: Wikipedia, Omniglot, Unicode proposal

Posted in Abugida, Rating: 3 "I did not know that" | Leave a comment

Oriya — 1100 AD, India

Posted on April 23, 2011 by ducky

Oriya "to"

Oriya probably descended from Bengali, though some say Kalinga (a script so obscure I can’t find out much about it, and which unfortunately shares a name with a language spoken in the Phillipines).

Oriya looks quite different from the other Brahmi-derived scripts that I have talked about so far: it is much rounder and swoopier; instead of a straight, horizontal headline, many of its glyphs have a semi-circular-ish hoops at the top of the glyph instead.

The curvaceousness of Oriya is another example of technology influencing the writing system.   Oriya was mostly written on palm leaves, and straight lines will tend to tear or split palm leaves.  Hence, they made their lines curvy.

While I am not an expert on the distribution of palm trees in India, I note that the Orissa province of India (where Oriya is used), is used farther south than the previous Brahmi-derived scripts that I have talked about previously.  I will be blogging soon about the scripts of Southern India and Sri Lanka soon, and you might agree with me that they are even more curly and swoopy.

Links: Wikipedia, Ancient Scripts, Omniglot, Technology Development for Indian Languages: Oriya script details also Ancient Scripts Kalinga

Posted in Abugida, Rating: 5 "Whoa!!", technology influenced | 4 Comments

Bengali — 1100 AD, Bangladesh

Posted on April 22, 2011 by ducky

Bengali "cho"

Bengali script, sometimes called Bangla, evolved gradually from Nagari, which makes giving it a starting point difficult.  Certainly Bengali was a separate script by 1778, when the first metal type was cut for it, but the date sources say it diverged from Nagari ranges from 1000 to 1200.

Worse, some people define an intermediate script “Eastern Nagari” between Bengali and Devanagari; others say what I am calling “Bengali script” is “Eastern Nagari script”.  (I suspect that is for political reasons, as there are two minor variants of this script used to capture spoken languages that are not Bengali.  English is written in a script that type (font) geeks call “Latin script”; people don’t mind much.  I bet there would be serious objections if that script was called “Italian script” instead.)

Like all of the Brahmi-derived scripts, Bengali is an abugida, where most characters are syllables with a consonants plus an implied vowel.  Unlike most of the other Brahmi-derived scripts, Bengali script’s implied vowel is pronounced as “oh”, while most other Brahmi-derived scripts pronounce it somewhere between an “ah” and “uh”.  This would be similar to if, say, Spanish pronounced the letter “A” as “ah” and Swedish pronounced is as “oh”.

Note that even though the implied vowel is pronounced as an “oh”, the characters are described or named in many places with the “a” vowel.  For example, Omniglot describes the character at the top of this post as the “cha” character.  This makes a certain sense: it is related to the character that all the other Brahmic scripts pronounce as “ah”/”uh”.  This is similar to how we English-speakers describe a descendant of “a” as “a-umlaut” even though Gremans pronounce a-umlaut close to the way Spanish speakers pronounce the letter “e”.

Links: Wikipedia, Omniglot, Ancient Scripts, Banglapedia, Tripod

Posted in Abugida, Rating: 4 "Huh, interesting!" | 5 Comments

Modi — 1600 AD, India

Posted on April 16, 2011 by ducky

Modi "la"

Like Gujarati, Modi is a variant of Devanagari that was developed in about 1600 AD. Like Gujarati, it was used mostly for accounting, then later for administration.

It looks very similar to Devanagari, but with fewer ligatures, rounder/”swoopier” glyphs that are faster to write (and harder to read).  Like Gujarati, Modi got rid of Devanagari’s inconsistent headline, but unlike Gujarati, Modi always has a headline instead of never.  If you look at examples of old texts in Modi, it looks almost like the writer drew the headline before or after the rest of the text.

Links: Wikipedia, Omniglot, Ancient Scripts, Unicode proposal

Posted in Abugida, Rating: 2 "Not all that interesting" | Leave a comment

Gujarati — <1600 AD, India

Posted on April 15, 2011 by ducky

Gujarati "tha"

Gujarati is a direct descendant of Devanagari, used mostly but not exclusively for writing the Gujarati language.

Like the Landa scripts, Gujarati was initially used mostly for commerce. It is sometimes called “banker’s script”, “merchant’s script”, or “trader’s script” in the native language(s).

Like the Landa scripts, it was purportedly used in part to make it more difficult for outsiders to read documents in it, since Gujarati wasn’t terribly well known (at first).  Unlike the Landa scripts, it does not appear to have developed lots of regional variations.

One of its advantages for commerce is that it could be written faster than Devanagari, in part because Gujarati did away with the horizontal line (the “head” line) at the top of the characters.  Devanagari characters sometimes have a headline and sometimes don’t, so you have to be a little bit careful about where you draw the headline.

Links: Wikipedia, Ancient Scripts, Omniglot

Posted in Abugida, Rating: 3 "I did not know that" | 2 Comments

Devanagari — 1000 AD, India

Posted on April 14, 2011 by ducky

Devanagari long "u"

Devanagari developed from Gupta via an intermediary script called “Nagari” which has very little information about it available.  “Nagri” means “urbane”, so presumably it was used for commercial purposes.  “Deva” means “deity”, so calling the derivative script “Devanagari” presumably was to recognize that this script was used for both commercial and religious purposes.

Nagari became distinct from Gupta in about 700 AD; Devanagari became distinct from Nagari in about 1000 AD, but didn’t get a separate name until later.

Devanagari is closely associated with Sanskrit, the language of both Buddhism and Hinduism.  Many people think that Devanagari is “the” writing system of Sanskrit, Buddhism, and/or Hinduism, despite the fact that Sanskrit, Buddhism, and Hinduism all predate Devanagari.  People used whatever writing system happened to be around.  (And people even now still write Sanskrit with the Tibetan script.)  There are also many languages other than Sanskrit that were and are written with Devanagari, including many Indian languages, most notably Hindi.  (While Devanagari readers are in the minority of all of the world’s readers, it’s a big minority.)

Devanagari is a pretty standard Brahmi-derived abugida, but has an unusual punctuation symbol, the avagraha. The avagraha, which looks like a Latin “S” is used to show vowels that have gotten lost by being smooshed together, and also for vowels that have gotten extended.  If English used “S” as an avgraha and if English spelled things like they were pronounced, you might use the avagraha in the first way to write “ISscream” to show that even though it is commonly pronounced “I scream”, you really mean “ice cream”.  For the second use case, imagine a kid yelling for his mother: “MaSSSSSSSS!”

 

Links: Wikipedia, Ancient Scripts, Omniglot

Posted in Abugida, Rating: 3 "I did not know that" | 2 Comments

Tibetan — 630 AD, Tibet

Posted on April 13, 2011 by ducky

Tibetan "kroo"

According to Tibetan tradition, Thonmi Sambhota went to India in about 630 AD to study writing, and devised a script based on Gupta or Siddham for the Tibetan language.

Tibetan has some differences from mainline Brahmi-derived scripts.  For consonant clusters, Tibetan doesn’t use a vowel-killing virama as much as other Brahmi-derived scripts.  Instead, they frequently place small versions of the “extra” consonant above or below the parent syllable.  Predictably, the stacked characters frequently (usually?) turn into a ligature (i.e. what looks like a single character and sometimes abstracted slightly).    For example, the syllable glyph adorning has in the centre a “ka” glyph, with a “oo” diacritic on the top and a “r” on the bottom, making “kroo”.  Note: this stacking is done even more when used to write Sanskrit than native Tibetan.

For punctuation, Tibetan uses a dot (a “tsheg”), placed relatively high up (around where the dot of an i lies) to separate syllables.  The single dot might have derived from a “double tsheg“, shaped like a colon. Note that this has the effect of marking when the implied vowel should be retained, and when the vowel should be dropped and the character should represent only the consonant.

Tibetan also use a short vertical stroke (a “shad”) that looks like a very skinny cuneiform wedge, that functions similarly to the Latin comma and/or the word-continuation hyphen.  While Tibetan uses a space after shads, they do not have a word separator.  There are also several characters used to denote the beginnings of passages, verses, chapters, and (in some instances) pages.

There are two versions of the script, one with a horizontal line at the top (the “head” line), which is mostly used in printing, and one without, used mostly in handwriting.  If you hear someone talking about headless Tibetan, they are probably talking about the latter script and not decapitated corpses.

Tibetan is an extremely conservative orthography.  They still spell things the way they were spelled in about 900 AD, despite the spoken language having moved on.  (To make an English-language analogy, it would be if we used Late Old English spelling now.   While it would mean that we could read Beowulf in the original, it would make it even harder to learn to read!)

Links: Wikipedia, Omniglot, Ancient Scripts, Formatting rules for Tibetan

 

Posted in Abugida, Rating: 4 "Huh, interesting!" | 1 Comment

Siddham — 650 AD, India

Posted on April 12, 2011 by ducky

Siddham “ja”

Buddhists in around 400 AD wrote Sanskrit with a version of Gupta script that eventually diverged into Siddham script.  At first, writing down the Sanskrit was used mostly as a memorization aid, but when Buddhism spread to the much-more literate China, apparently things needed to be written down to be taken seriously.  Problem: to recite the mantras properly, you needed to be able to pronounce them properly, and the logographic Chinese script did a really bad job of showing how you should pronounce syllables.  Thus the Siddham script moved from being useful to being critical.

A pair of Japanese monks, Kukai and Saicho, traveled to China to study in the early 800s. They brought Siddham texts back to Japan with them and founded highly influential Buddhist schools.  About 40 years later, Emperor Tang Wuzong repressed Buddhism in China quite effectively, cutting Japan off from other Buddhist influences.  Left undisturbed, Siddham script just happily stayed in Japan long after the Buddhists in India switched over to Devanagari.  Siddham is still in use in Japan today, 1200 years later, although it is now called Bonji script.

This makes Siddham relatively old among modern continuously-used scripts, and means that Japan has five scripts that are currently used, not four like you thought.

Links: Wikipedia, Omniglot, Visible Mantra

Posted in Abugida, now ceremonial, Rating: 4 "Huh, interesting!" | 2 Comments