Writing systems
Alphabet, Abjad, Abugida, Syllabary: Reading the Four Kinds of Script
What each type encodes, one worked example each, and how to tell which you are looking at.

Four types of writing systems encode sound in different ways: alphabet, abjad, abugida, and syllabary. To decipher an unfamiliar script, your first move is to determine what unit it uses: phonemes (sounds), consonants, consonant-vowel pairs, or syllables. Read on for how to diagnose each category on sight.
What a script encodes
An alphabet encodes speech sounds (phonemes) with graphemes, like our Latin script. Abjads also mark consonants, but leave vowels to be inferred unless specifically marked. Abugidas write vowelized consonants but modify them for other vowel meanings. Syllabaries assign one character to each syllable, with no common base across syllables.
Abjads: consonants first
Abjads prioritize consonants. If you see a script that writes consonants only, except for a few optional marks, it's likely an abjad. Look for a pattern where the shape denotes a consonant and a secondary mark for a vowel:
Arabic | Hebrew | Syriac ----- | ------ | ----- ك | כ | ܟ (#unit meaning /k/) كا | כַ | | كي | קי | | كُو | כוֹ | |
Arabic, Hebrew, and Syriac are the classic examples. Inscriptions.
Abugidas: consonant plus default vowel
Abugidas are consonant-based, but include a default vowel to guide pronunciation. The wrinkle is that the base letter assumes a particular vowel, with the other vowels marked by diacritic. That default vowel can be neutral (as in Cherokee), short (as in Burmese), or long (as in Devanagari).
Unicode | Example in Devanagari ----- | ----- क | /ka/ का | /kā/ (长∗∗) की | /kī/ (长∗∗) कु | /ku/ (短) त | /ṭa/
In Cherokee, the default vowel is /a/: Ꭴ /u/ is marked with an overdot, but Ꭷ /ka/ assumes the a built into the consonant shape. In Indic abugidas like Devanagari, the vowel a is the default (क्र /kra/).
Most readers will recognise many abugidas from Thai or Kannada:
| Thai Consonants with Low Class | Thai Consonants with High Class ----- | ----- ก | เ ๊a: | ข | เ ก | ค | ง ใ ษ | กา ษ | ม | น | ค | น "; ค "| <------ | </tr> <------ | </tr> <------ |
To sum up, abugidas are written from vowelized consonants, which modify shape for vowel changes. They write base vowels separately when occurring alone or at the start.
Syllabaries: one sign per syllable
Syllabaries seem the simplest, but they are the exception in the world's writing systems. Kana, Yi, and historic Egyptian hieroglyphics are some of the few successful syllabaries. In a syllabary, each character represents a whole syllable, usually consonant-vowel:
| Kana | Hiragana | Roman | ----- | ----- | ----- あ | /a/ | a| い | /i/ | i| う | /u/ | u| え | /e/ | か | /ka/ | ka| ち | /ti/ | ti| と | /to/ | to|
Kana writes kana, a, i, u, e, o and their combinations. Each symbol has a sound value.
Alphabets: consonant and vowel sounds
Most Europhone readers know the alphabet, where each letter represents a unique vowel or consonant sound. Unlike the other systems, the component is what's important, not the consonant-vowel conjunction. In Latin, Greek, Cyrillic, and similar alphabets, vowels and consonants work on the same footing. The vowel letter /i/ represents the vowel sound /i/.
| Greek | Cyrillic | Latin | IPA ----- | ------- | ----- | ----- Α | А | A | [a]/[æ] | Β | Б | B | /β/ | Γ | В | C or K | /g/ | Δ | Г | D | /d/ | Ε | Д | E | [e]/[ɛ]|
Note that Latin is one alphabet and Greek one with quite different phoneme values. The crucial test is whether consonants and vowels are written by separate units.
How to classify a script at a glance
To classify a script, check the components:
1. Are there only consonants and sometimes extra vowel marks? Abjad.
2. Is every letter or glyph a consonant plus one implied vowel, with signs for other vowels? Abugida.
3. Does the script write a character for each syllable without a shared base across syllables? Syllabary.
4. Do the letters mark vowel and consonant sounds on an equal footing? Alphabet.
... and in real life? Ask to describe or link to an essay already published in a standard reference. (Language typology handbook or encyclopedia).]
Don't rely on your letter for script classification.
Where this sits in the Register
Written by the Linguasphere desk. Classification data referred to here comes from the Linguasphere Register — see sources and attribution.