Hebrew — 300 BC, Israel

Posted on February 18, 2011 by ducky

Hebrew "sh"

Hebrew is a difficult writing system to shoehorn into this blog format.  For starters, when did the Hebrew script come into existence?  Unlike Cree and Cherokee, which had very distinct release dates, the Hebrew script evolved over thousands of years.  Of major* modern languages, only Chinese has been written down for longer than Hebrew.  (Given how much Chinese characters have changed since it started being written down, perhaps we should be surprised at how little Hebrew has changed!)  There have been several significant changes in the script along the way.

As I mentioned in the Samaritan post, the oldest written Hebrew was in a script that looks to me like Phoenician.  Some time during the Babylonian exile (597 BC to 538 BC), the Jewish people started writing in the Aramaic script.  After they got back home, their dialect of the Aramaic script gradually developed its characteristic squarish form.

As Aramaic became more and more popular, Hebrew started to fall out of favour.   By 200 AD, Hebrew was not spoken as a day-to-day language, but was still used as a liturgical language (similar to how Latin held on as the Catholic Church’s primary religious language until 1963 AD).   The Hebrew script was used by Jewish communities for various non-Hebrew Jewish vernaculars, including Karaim, Judæo-Arabic, Ladino, and Yiddish.  In around 1880 AD, a movement began to reintroduce Hebrew as a home language.  That movement ultimately was very successful; today Hebrew is the native language of about 7 million people, mostly in Israel.

Like its Aramaic parent, the Hebrew square script does not have separate vowels, but three consonants sometimes also act as long vowels.  In around 700-1000 AD, scholars got uncomfortable with the ambiguity, and developed various systems of diacritics to augment the consonants with diacritics, with the niqqud system winning out.  With the niqqud system, dots are placed in different places above, below, or even inside the consonants.  However, the niqqud are usually only used for people learning the language, e.g. children and immigrants, or in dictionaries.

The lack of acceptance of using the niqqud meant that there was still significant ambiguity, so there have been several spelling reforms to make it clearer when a glyph is a consonant and when it is a vowel, most recently in 1996.

One innovation that Hebrew made was to have a different form of some of the characters when they were at the end of a word.  This is quite a clever invention: it helps readers recognize word boundaries without taking up more horizontal space on expensive papyrus or vellum.

Hebrew also developed a method for expanding the palette of sounds, e.g. for loanwords.  The geresh — which looks very much like an apostrope — is placed in front of the consonant it changes.  For example, a ‘ in front of a “g” (as in “gap”) sign turns it into a “j” (as in “Jupiter”).

* There are about 200 native speakers of Aramaic.  Greek has been written for longer than Hebrew has been written in the Hebrew square script, but not longer than Hebrew has been written.  Samaritan (which has about 700 native speakers) writing shares an origin with Hebrew, so has been written for the same length of time.

Links: Wikipedia, Ancient Scripts, Omniglot

Posted in Abjad, Rating: 4 "Huh, interesting!" | 2 Comments

Cree — ~1840 AD, Canada

Posted on February 17, 2011 by ducky

Cree "lii"

Missionary James Evans developed a romanization for the Ojibwe language in around 1830 AD, but found that Ojibwe students had difficulty switching between the two very different mappings of Latin characters to pronunciation.  Inspired by the stunning success of Cherokee a few years earlier, Evans started to develop an Ojibwe syllabary, but his superiours were hostile to the idea, as it ran counter to their goals of assimilation.  Evans was not allowed to announce his syllabary.

In 1840 AD, he was redeployed into an area of Manitoba that spoke Swampy Cree and where he had superiours who were supportive of his ideas.  He dusted off his Ojibwe syllabary and adapted it a bit for Cree.

Evans found inspiration in three different writing systems from three different continents: Cherokee from the USA as mentioned above, Pittman shorthand from England, and Devanagari from India.

Evans’ syllabary was an abugida, which he was familiar with as he had previously learned Devanagari, one of the many Brahmi-derived writing systems of India.  The shapes Evans used for the consonants which started syllables are very similar to the corresponding Devanagari syllables.

Cree also needed consonants at the ends of syllables, and Evans derived the terminal consonants from Pittman shorthand.  They aren’t exactly the same, but they are very close.

Something else that Evans took from Pittman, and one of the striking features of his syllabary, was rotation.  In Evans’ script, you change a syllable’s vowel not by decorating it with extra lines, but by rotating or flipping it.  Rotating a syllable ending in “-a” to the left by 90 degrees turned the vowel into a long “-e”.  Flipping an “-a” syllable vertically gave the corresponding “-o” syllable.

Evans gave his script to the world, and the world took it and ran like hell with it.  Like happened when Cherokee was introduced, literacy amongst the Cree zoomed up in only a few years, surpassing their white neighbours’ literacy.  White folks got a bit nervous about the Cree’s literacy for a time; the Hudson’s Bay Company actually refused to import and sell a printing press to Evans.   Evans had to resort to building a printing press himself.

Evans’ script was not only popular with the Cree.  The descendants of Evans’ Cree syllabary are used all across Canada to write a large number of First Nations languages, and to a lesser extent in the USA.

Postscript: It is quite possible that the abugida was only invented once, in the form of either Kharosthi or Brahmi.  One of those two clearly spawned the other.  All other Indic abugidas, past and present, are unambiguously descended from Brahmi.  All the Canadian First Nations syllabaries are unambiguously derived from Evans’ Cree syllabary, which was clearly influenced by a Brahmi-derived script.  The only other abugidas in the worlds, past or present, were Ge’ez (whose vowel markings showed up so much after Brahmi-descended writing systems became widespread that it is quite likely that the Ethiopians got the idea from the Indians) and a few forms of shorthand (who might have gotten the idea from Indic scripts).

Links: Wikipedia, Ancient Scripts, Omniglot

Posted in Abugida, inventor known, Rating: 5 "Whoa!!", Syllabaries | 3 Comments

Shorthands — <300 BC, Greece?

Posted on February 16, 2011 by ducky

Pittman “th”

Shorthands — forms of writing that sacrifices accuracy and/or shared orthography for speed — are very old.  The earliest example of shorthand comes from Greece, and was sort of an inverse abugida: the vowels were primary, and consonants were noted by decorating the vowels.

English-speakers have been particularly keen on shorthand, for perhaps understandable reasons.  Four different shorthand systems were published in the UK from 1588 to 1626.  The latter, by Thomas Shelton, became quite popular.  This was sort of like an abjad, sort of like an abugida: each consonant had a glyph, and the arrangement of the consonants gave the vowel.  Placing the second consonant above the first gave an “a”, below gave an “e”, etc.

Several other Englishmen and one German came up with reasonably popular shorthands, including Samual Taylor, who cut redundant consonants and most vowels.

In 1937, Sir Isaac Pitman published a shorthand which was popular enough that it still is used now and got him knighted then.  Pittman, unlike his predecessors, was a phonetician, and his script is not only a phonetic script but took advantage of his phonetic knowledge.  For example, voiced consonants (e.g. “s”) and unvoiced (e.g. “z”) differed by the thickness of the stroke; it uses straight lines for  plosives and arcs for fricatives.  Some glyphs are rotations or flips of glyphs for similar sounds.  The result is that things that sound similar, look similar.  Pittman also has some logograms — single glyphs that represent common words.

In 1888, John Robert Gregg published a shorthand which now pretty much shares the English “market” (and much of the non-English Latin-script market as well) with Pittman. Gregg shorthand is reputed to be the fastest shorthand.   Both are phonetic, but where Pittman uses light weight as a dinstinguishing element, Gregg uses line length.

Shorthand has been popular in Japan as well; no fewer than nine pen shorthands are in use there.

Links: Wikipedia (English), Wikipedia (French), Pittman how-to, Flavin’s Corner (extremely long and detailed discussion of shorthands)

Shelton borrowed heavily from his predecessors, especially Edmond Willis. Each consonant was represented by an arbitrary but simple symbol, while the five vowels were represented by the relative positions of the surrounding consonants. Thus the symbol for B with symbol for T drawn directly above it represented “bat”,
Posted in Abjad, Abugida, Alphabet, inventor known, Logograms, Rating: 5 "Whoa!!" | 3 Comments

Cherokee — 1819 AD, USA

Posted on February 15, 2011 by ducky

Cherokee "o"

The development and adoption of Cherokee are two hugely impressive accomplishments.  Chief Sequoyah, who was illiterate himself, single-handedly created a script for his language, and within eleven years, 90% of Cherokee were literate in their language.  Stop and marvel about that for a minute.

Sequoyah was impressed by the white folks’ “talking leaves” and spent about twelve years coming up with a writing system.  At first, he tried one glyph per word, but realized that would be a boatload of characters.  He then tried a syllabary, and apparently worked hard at figuring out which sounds there were, enlisting friends to help him tease out fine distinctions.

He spent twelve years on the task, but the script was extremely well-done when he finished.  Aside from some minor changes in the shapes of the glyphs to make it easier to use with a printing press, there have been almost no changes.  Stop and marvel again, I can wait.

Sequoyah owned a Bible, which he was unable to read, but he derived many of the shapes of the letters from the shapes he saw there; in his 86 characters, about 20 have glyphs that look just like a Latin character.   However, there is no correspondence between the meaning of the Latin letter and the similar Cherokee letter.  The symbol that looks like Latin “A” is the syllable “go”; “B” is the syllable “yv”, and “4” is the syllable “se”.  (The other characters tend to be curvier than Latin.)

Part of the immediate spread of literacy among the Cherokee was probably due to the technology for writing being readily available.  Sumerian cuneiform would have spread much more quickly if paper, pencils, a postal service, and printing industry were available from the moment they developed writing.  However, Sumerian cuneiform was very complicated and would have been very difficult to learn.  Even if the Cherokees’ push for literacy was boosted by national pride, even given the technology availability, there is no way that they could have become so literate so quickly if Sequoyah had not done such a fine job developing his writing system.

Links: Wikipedia, Ancient Scripts, Omniglot

Posted in developed by illiterate(s), inventor known, National pride, Rating: 5 "Whoa!!", Syllabaries | 4 Comments

Ge’ez — 400 BC, Ethiopia

Posted on February 14, 2011 by ducky

Ge'ez "ci"

Ge’ez, aslo called Ethiopic, is the only Old World abugida outside of Southest Asia and the only abugida that is not clearly derived from Brahmi.  (Aside from Kharosthi, of course, which maybe spawned Brahmi.)

However, it took a long time for the vowel decorations to appear.  For the first 500 years of Ge’ez’ life, it was a rather ordinary offshoot of South Arabian (except for the fact that it is the only offshoot of South Arabian).  In around 200-300 AD, however, they started decorating the consonants with vowel information, with credit generally being given to Saint Frumentius for that invention.  (He also is credited with the first translation of the Bible into Ge’ez, the spoken language of the area, and spreading Christianity widely in what is now Ethiopia.)

Frumentius’ life story doesn’t show him spending time in India, but Ethiopia was an ocean country, and Frumentius might have come into contact with Indic abugidas via Indian traders.  (Frumentius was the boyhood tutor for the king Ezana, and coins with Ezana’s name were found in India.)

Ge’ez started out being written right-to-left, like most of the other Semitic abjads, but switched to left-to-right, perhaps because of the influence of Greek.

Ge’ez is still used in Ethiopia to write the national language, Amharic.  With a population of 85M in Ethiopia (for comparison, Germany has 80M people), Ge’ez is a significant writing system.

Ge’ez uses something that looks like a semi-colon to indicate word breaks.

Links: Wikipedia, Ancient Scripts, Omniglot, Ethiopic.com

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

Brahmi — 400 BC, India

Posted on February 13, 2011 by ducky

Brahmi "sa"

Brahmi is sort of the Phoenician of East Asia: almost all the non-logographic scripts in East Asia come from Brahmi, including almost all of the scripts used in India, Tibet, Nepal, Mongolia, Thailand, Indonesia, Cambodia, and Laos.  Brahmi was a big deal.

Like Kharosthi, Brahmi is an abugida.  Either Kharosthi or Brami was the first abugida in the world, but it isn’t clear which was first — it’s difficult to date either of them well.  There were many examples of Brahmi in around 250 BC from the stone pillars of the Edicts of Ashoka, and there was some geographical variation already, but not much.  There are some examples of what might have been Brahmi on some pottery fragments from 500 BC, but not many and it isn’t completely clear that they are Brahmi.

The glyphs in Brahmi and Kharosthi are very different, but the mechanics are very similar: consonants are assumed to have the “a” sound attached to them; if it is a different vowel, a diacritic is added to the consonant.  However, leading vowels are treated differently.  In Kharosthi, there is a sign meaning “vowel at the beginning of a word” which is presumed to be an “a” unless it has vowel diacritics on it.  In Brahmi, there are five different glyphs for five different leading vowels.

Brahmi also deals with consonant clusters in a slightly different way than Kharosthi: both  smoosh the consonant signs together, but Kharosthi smooshes horizontally, while Brahmi puts one of the consonants under the other.

Brahmi might have descended from Phoenician via Aramaic.  Many of the glyphs look similar to the Aramaic glyphs of similar sounds, although sometimes flipped (which is common when the writing direction changes, and Brahmi is a left-to-right language) and sometimes upside-down.  To my eye, the upside-down-ness makes it easier to decorate those glyphs with the vowel diacritics.

While it is clear that Kharosthi developed in Pakistan, which definitely used Aramaic as part of the Achaemenid Empire, it’s hard to tell where Brahmi developed: the definitive records burst onto the scene with Ashoka (whose dynasty originated in eastern India), while the disputed texts come from southern India — neither of which were under Achaemenid rule.  To my eye, the Brahmi glyphs look more like Aramaic than Kharosthi, despite Kharosthi coming from a Aramaic-writing area.

Links: Wikipedia, Ancient Scripts, Omniglot, very cool interactive Indic scripts comparison Tool from Ancient Scripts

Posted in Abugida, previous script didn't quite work, Rating: 4 "Huh, interesting!" | 12 Comments

Kharosthi — 350 BC, Pakistan

Posted on February 12, 2011 by ducky

Kharosthi “cha”

The Persian Achaemenid Empire conquered huge tracts of Asia in around 500 BC, and held it until about 330 BC.  They spread the use of their official language, Aramaic, and with it the Aramaic writing system.  Near the end of the Achaemenid Empire, the Gandhara culture of Pakistan/Afghanistan developed a writing system that probably came from Aramaic, but with additions to handle the different sounds of their language.

It was deciphered using bilinual Greek-Kharosthi coins.  Yes, Greek.  (Shortly after the development of Kharosthi, Alexander the Great swept through, bringing not only war and destruction, but also Greek colonists in his wake.)

While Kharosthi has enough clues to make scholars think it came from Aramaic, it is a very different form of writing system: an abugida.  Abugidas write vowels, but the vowels are written as decorations upon the consonants as opposed to being letters in their own right.  In addition, abugidas usually have a “default value” vowel which does not get diacritics.  (This is different from the vowel pointing that modern Hebrew and Arabic have, in that vowel pointing is optional and Kharosthi diacritics are not.)  The default vowel in Kharosthi is an “a”: glyphs for “consonant plus ‘a'” were undecorated, but otherwise the other vowel’s diacritic would be added to the glyph.

Kharosthi also has a character for an isolated vowel (which only happens at the beginning of a word).  If it is an “a”, then the vowel glyph doesn’t have any decoration; if it is one of the other vowels, it gets a diacritic.  Consonant clusters were shown by smooshing the characters together closely, maybe even making it a ligature.  In addition, there are a few diacritics for modifying vowels (e.g. to lengthen the vowel) and a few to modify consonants (e.g. to aspirate).

How did they come up with this idea?  Perhaps they came up with it themselves, but perhaps they were influenced by Old Persian cuneiform, which died out in around 300 BC.  Old Persian also had “a” as a default-value vowel, but it had separate glyphs for the non-default vowels instead of them being decorations on the consonants.

Unlike other Indic scripts, but like Aramaic, Kharosthi is written from right to left.

Links: Wikipedia, Ancient Scripts, Omniglot, Unicode proposal (which has some discussion of the writing system mechanics, including the diacritics)

Posted in Abugida, previous script didn't quite work, Rating: 5 "Whoa!!" | 2 Comments

South Arabian — 800 BC, Yemen

Posted on February 11, 2011 by ducky

South Arabian "s"

Proto-Sinaitic split into two branches: a northern one which spawned almost all the writing systems of the modern world, and a southern one that did not.  Perhaps it is fairer to say that one branch of the script went to Yemen and stayed there unmolested.

South Arabian is the only child in the southern branch.  (If you look at a map of Yemen, it is kind of remote.  It isn’t that far from the Levant by boat now, but before the Suez Canal, you couldn’t get from the Mediterranean to Yemen without going though a bunch of desert.  It is not at all far from Somalia by boat, but you do have to have a boat.)  South Arabian happily sat in Yemen and evolved.  By about 800 BC, the South Arabian script had developed quite distinct glyph shapes.

Even in the far-off reaches of Yemen, however, South Arabian was able to borrow good concepts from other languages.  It adopted the matres lectionis idea, presumably from Aramaic, and the alphabetical ordering from Ugaritic.  However, the alphabetical ordering it got from Ugaritic was NOT “abg…” that turned into the “abc” of Latin, but the other Ugaritic ordering, which went “hlħmqw…”

One innovation that they never did get was vowels.  South Arabian went to its metaphorical grave without markings for vowels (aside from a few double-duty matres lectionis consonants).

South Arabian could be written left-to-right or right-to-left, with the symbols reversed according to the direction.  It also had a cursive form (Zabur) that was used for more casual communications.

Links: Wikipedia, Ancient Scripts, Omniglot

Posted in Abjad, Evolved slowly from parent, Rating: 3 "I did not know that" | Leave a comment

Paleohispanic scripts — 450 BC, Spain

Posted on February 10, 2011 by ducky

Celtiberian "bu"

One of the places the Phoenicians colonized was the Mediterranean side of Spain, and their writing system spread around that peninsula.  Like in Italy, there were quite a few different, mostly related scripts.  Unlike in Italy, there wasn’t a hugely successful culture in the vicinity which talked about its neighbours and whose manuscripts got copied and recopied by monks in later years; their own writers also didn’t leave as much behind.  Less is known about the tribes on the Iberian peninsula than the Italian one.

The scripts that are best understood are the Levantine scripts, which descended from Phoenician.  There is a northern set and a southern set, but their differences are — scholars believe — mostly just in glyph shape, not in the fundamental structure.  (The Celtiberian branch in the northeast dropped two characters.)

While these scripts are clearly the children of Phoenician, Phoenician might not be all that eager to claim them.  Indeed, they might look like Phoenician, but they act very differently.  Instead of being an abjad like their parent, instead of being a forward-looking syllabary or even one o’ them avant-guard alphabets, these scripts are a bastard love-child combination of syllabary and alphabet.  They have vowels!  They have consonants!  They have syllables!

All the vowels have their own glyphs.  All of the consonants that you can say continuously (the continuants, like “m”, “l”, and “r”) are individual letters.  All the ones that you can’t (the occlusives, like “b”, “k”, and “t”) only get a glyph if they are paired with a vowel.  For example, you have glyphs for “a”, “m”, and “ba”, but not for “b” or “ma”.

Greek script was also used on the Iberian peninsula, with almost no change.  They left some letters out and included a short vertical mark to the “r” symbol to make it “trilled r”.  It’s most striking feature is how few letters it has: sixteen, fifteen if you consider the trill mark as a modifier and count the trilled “r” glyph as the same as the untrilled “r” glyph.

Links: Wikipedia, Ancient Scripts, Omniglot, Iberian Epigraphy Page

Posted in Alphabet, Rating: 4 "Huh, interesting!", Syllabaries | Leave a comment

Tifinagh — 400 BC, Tunisia or Libya

Posted on February 9, 2011 by ducky

Tifinagh “s”

Nobody is quite sure where the Berber script, used by the nomads of Northern Africa, came from.  English sources are pretty certain that Tifinagh evolved from the Phoenician script that settlers brought with them when they founded Carthage in about 1000 BC.  French sources admit the possibility that it evolved from a much older, unknown, native script.   English scholars think that the name of the modern script derives “from” (“ti-“) + “punic” (the name for the Carthaginian Phoenicians).  French sources suggest that it means “discovered” (“tifi“) + “ours” (“nnagh”).

Both languages’ sources say that there was an eastern dialect and a western dialect of the script.  Both agree that Tifinagh was used until about 200 AD, when the artifact record died out, but there is disagreement on the dating of the start of its use.  Because of a small number of bilingual texts and its continuation into the modern era (more on that later), scholars are pretty sure what phonemes the glyphs correspond to. The English sources say that, like Etruscan, they cannot read the language.  French sources say that the Eastern dialect has been deciphered.  Because of who was in the area, they are both pretty sure it was a language of the Berber nomads.

Unusually for scripts of that era, Tifinagh was usually written vertically, and even more unusually, most commonly from bottom-to-top.  Only some glyphs were allowed at the beginning of lines.  These could be used to tell which direction to read the writing in.

Update: Hanunó’o, a writing system in the Philippines, is also written bottom-to-top.

English links: Wikipedia, Ancient Scripts, Omniglot

French links: Wikipedia, Ennedi

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