What If a Writing System Had No Letters? One Sign per Syllable

An alphabet assigns signs to individual consonants and vowels, so a syllable is assembled from two or more of them, and a language with a few dozen sounds needs only a few dozen letters. A syllabary assigns one sign to a whole syllable, so the sign for a particular consonant followed by a particular vowel has no visual relationship to the sign for the same consonant with a different vowel. The number of signs required therefore equals the number of distinct syllables in the language, which is the property that determines whether the approach is practical. Learners of such a system are not taught to sound out letters, since there is nothing smaller than a syllable to sound out.

The fit depends on the shape of the syllables in the language:

• •Japanese has a small inventory of simple syllables, mostly a consonant plus a vowel
• •Around fifty signs therefore cover the language, which is easily learned
• •English permits complex clusters at both ends of a syllable
• •Counting distinct English syllables gives a figure in the thousands
• •A syllabary for English would be unusable, which is why none exists
• •The system suits languages with simple syllable structure and nothing else

Several such systems are in current use or were until recently. Japanese uses two parallel sets of the same syllables, one for native grammatical elements and one for foreign words and emphasis, alongside thousands of borrowed Chinese characters, which makes it among the most complicated writing arrangements in use. Cherokee acquired one in the 1820s devised by Sequoyah, who could not read any script and invented the entire system from scratch after recognising that writing was possible, which is a documented case of one person creating a working script for a language that had none. Inuktitut uses a syllabic system adapted from one devised for Cree. The Linear B script of Bronze Age Greece was a syllabary, and deciphering it revealed it was recording an early form of Greek.

Placing this category properly means setting out what the alternatives are, since the usual division into pictures and letters is wrong. A logographic system gives each sign a word or meaningful unit, which is how Chinese characters work, requiring thousands of signs and making the script usable across languages that sound nothing alike. An alphabet gives separate signs to consonants and vowels. An abjad writes consonants and leaves vowels to be inferred or marked optionally, which is how Arabic and Hebrew work and suits languages where the consonants carry the root meaning. An abugida attaches vowels to a consonant sign as modifications rather than separate letters, which covers the scripts of India and Ethiopia. Most real writing systems mix more than one of these.

Neither approach is superior in general and the advantages pull in opposite directions. A syllabary requires memorising more signs but removes the problem of blending sounds together, which is genuinely difficult for beginning readers and is where a substantial part of reading difficulty concentrates in alphabetic languages. Each sign maps to a pronunciation reliably, so spelling and reading are consistent, unlike English where the relationship is notoriously irregular. Against that, the system cannot represent syllables it lacks signs for, so foreign words and names get distorted into the nearest available shapes. And a larger sign inventory takes longer to learn to write. Literacy rates offer no clean comparison, since too many other factors differ between the societies concerned.

Each sign stands for a whole syllable, so the number of signs needed equals the number of distinct syllables in the language, which makes the approach workable for Japanese and impossible for English. Japanese, Cherokee and Inuktitut all use one, and the Cherokee example was invented outright by a person who could not read. The trade is more signs to memorise against no need to blend sounds and reliable pronunciation.