Syllabaries

Chapter 3 ended with Japanese borrowing Chinese characters. Japanese then added something Chinese does not have: a set of symbols for syllables. This chapter looks at the syllabary type through three examples from three continents and three periods: Japanese kana, the Cherokee syllabary and Linear B.

Why Syllabaries Fit Some Languages

• Simple syllables. The Syllabary article says that because Japanese uses mainly CV syllables, a syllabary is well suited to write it. Modern syllabary languages tend to have simple phonotactics, with a predominance of single-mora CV syllables.
• Complex syllables are a problem. The article says English would need over 10,000 separate glyphs as a pure syllabary, so alphabets handle complex syllables better. Syllabaries often use "dummy vowels" for coda consonants instead.
• Often from logograms. The article says syllabaries often begin as simplified logograms, with katakana as its example.
• Not an abugida. The article separates syllabaries from abugidas (Chapter 7): in an abugida related syllables share graphical elements in a regular way, while in a syllabary the similarities are not systematic.

Japanese Kana

From Wikipedia's Kana article (first 100,000 of 150,752 characters read).

• Two sets. Hiragana and katakana, both developed around 800 CE per the Syllabary article. The Kana article says they are syllabaries used to write Japanese morae.
• Origins. Early Japanese was written in kanji. Man'yōgana, kanji used for their sound, represented grammatical particles and morphemes. Hiragana developed from cursive man'yōgana; katakana from abbreviated parts of regular-script man'yōgana, as a glossing system for Buddhist sutras. Kūkai is traditionally said to have invented kana, which the article says is not fully established.
• Division of labour. Hiragana is the default syllabary: native words without kanji, particles and okurigana (inflection endings). Katakana covers words of foreign origin, foreign names, onomatopoeia, emphasis, technical terms and some branding. Furigana are small kana beside kanji to show pronunciation.
• Moraic, not strictly syllabic. A CVC syllable needs two kana; the only coda kana is ん (n). The Syllabary article says kana are more correctly called a moraic writing system.
• Voicing marks. Voiced consonants (g, z, d, b) reuse the unvoiced kana plus a voicing mark (dakuten); p uses the h kana with a small circle (handakuten).
• Not fully phonetic. Kana do not mark pitch accent, so homophones like はし ("chopsticks" 箸 and "bridge" 橋) need context, which kanji help.

For learning the kana themselves, the site has the Hiragana and Katakana courses; this course asks only what kind of system they are.

The Cherokee Syllabary

From Wikipedia's Cherokee syllabary article (first 100,000 of 205,413 characters read).

• Inventor. Sequoyah, who was illiterate until he created it, began work around 1809 and finished in the early 1820s after about twelve years, dropping or changing most of his original characters.
• Size. The modern system has 85 characters. Sequoyah's original had 86; the 86th was dropped in 1828 when Elias Boudinot adapted the chart for printing.
• Borrowed look, different sounds. Some characters resemble Latin, Greek, Cyrillic or Glagolitic letters but do not represent the same sounds.
• Fast uptake. The article says it became popular almost immediately; by 1825 most Cherokee people could read and write in it, and the Cherokee Phoenix adopted it after the 1828 changes.
• True syllabary. "Each symbol represents a syllable rather than a single phoneme", though some distinctions (voicing, vowel length, tone) go unmarked.

Linear B

From Wikipedia's Linear B article (first 100,000 of 223,272 characters read).

• Language. Mycenaean Greek, the earliest attested form of Greek; c. 1450 to 1200 BC, from palace archives at Knossos, Kydonia, Pylos, Thebes and Mycenae. It disappeared when Mycenaean civilisation fell in the Late Bronze Age collapse.
• Signs. Around 87 syllabic signs and over 100 ideographic signs; the ideograms have no phonetic value.
• Uses. Mostly confined to administrative contexts at palace sites.
• Decipherment. Michael Ventris deciphered it in 1952, building on Alice Kober's work; John Chadwick helped, and the two did the bulk of it between 1951 and 1953. It was adapted from Linear A, an undeciphered script perhaps used for Minoan.
• A mismatch. Greek has complex syllables, but Linear B is a syllabary of simple signs. The Syllabary article says complex onsets were written with two signs or simplified, and codas generally ignored: ko-no-so for Knōsos.

Kana and Cherokee in Unicode

My own example; I ran it and copied the output. The first lines show a regularity: each katakana sits exactly 0x60 above its hiragana twin. I found this by experiment and have not seen the article state it.

The fourth line shows that が (ga) can be split into か plus a combining voicing mark: the dakuten from the Kana article, visible as a separate code point. The Cherokee block (U+13A0 to U+13FF) and Linear B syllabary block (U+10000 to U+1007F) are the articles' figures.

Try It Yourself