Carian Alphabet Table
| Letter | Unicode name | Value & origin |
|---|---|---|
| ๐ | a | a /a/ — from Greek Α |
| ๐ก | p2 | b /ʙb/ — perhaps a ligature |
| ๐ข | d | d /ð?/ — from Δ |
| ๐ฃ | l | l /l~ɾ?/ — from Λ |
| ๐ค | uuu | y /y/ — perhaps a modified ƒ |
| ๐ฅ | r | r /r/ — from Ρ |
| ๐ฆ | ld | lambda /lː~ld?/ — Λ plus diacritic |
| ๐ง | a2 | city-variant codepoint of a |
| ๐จ | q | q /kʲ/ — from ζ qoppa |
| ๐ฉ | b | b /β?/ — Greek-shaped variant |
| ๐ช | m | m /m/ — from archaic Μ |
| ๐ซ | o | o /o/ — from Ο |
| ๐ฌ | d2 | variant codepoint — value provisional |
| ๐ญ | t | t /t/ — from Τ (via the cursive-restructure path) |
| ๐ฎ | sh | sh /ʃ/ — perhaps from sampi |
| ๐ฏ | sh2 | variant codepoint of sh |
| ๐ฐ | s | s /s/ — from san |
| ๐ฑ | c-18 | unknown value |
| ๐ฒ | u | u /u/ — from Υ |
| ๐ณ | nn | n-tilde /n̩/ — syllabic nasal |
| ๐ด | x | k-circumflex /c/ — maybe a modification of Κ |
| ๐ต | n | n /n/ — from archaic Ν |
| ๐ถ | tt2 | variant codepoint of tau-sound |
| ๐ท | p | p /p/ — from Β |
| ๐ธ | ss | s-acute /ç?/ — perhaps from sampi |
| ๐น | i | i /i/ — from Ε, ΕΙ, or heta |
| ๐บ | e | e /e/ — from Η or heta |
| ๐ป | uuuu | y-umlaut /ʏ/ — not a Greek value |
| ๐ผ | k | k /k/ — perhaps Ψ rather than Κ |
| ๐ฝ | k2 | variant codepoint of k |
| ๐พ | nd | n-d /ⁿd/ — perhaps a ligature of ΔΔ |
| ๐ฟ | uu | w /w/ — from ƒ |
| ๐ | g | gamma-sound /ŋkʲ?/ — not a Greek value |
| ๐ | g2 | variant codepoint of gamma-sound |
| ๐ | st | z /t͡s/ or /st/ — not a Greek value |
| ๐ | st2 | variant codepoint of z |
| ๐ | ng | eng /ŋk/ — perhaps a ligature of ΓΓ |
| ๐ | ii | j /j/ — perhaps related to Phrygian /j/ |
| ๐ | c-39 | unknown value |
| ๐ | tt | tau-sound /t͡ʃ/ — perhaps from sampi |
| ๐ | uuu2 | variant codepoint of y |
| ๐ | rr | r-acute, l-acute /rʲ?/ — used in Egypt for Greek rho-rho |
| ๐ | mb | m-b /ʙb/ variant — Egyptian-variant row of b |
| ๐ | mb2 | variant codepoint of the b row |
| ๐ | mb3 | variant codepoint of the b row |
| ๐ | mb4 | variant codepoint of the b row |
| ๐ | ld2 | variant codepoint of lambda |
| ๐ | e2 | variant codepoint of e |
| ๐ | uuu3 | variant codepoint of y |
The Carian alphabets are a group of regional scripts used to write the extinct Carian language of western Anatolia, the peninsula opposite Rhodes. They consist of some 30 alphabetic letters - with several geographic variants in Caria itself and a homogeneous variant attested from the Nile delta, where Carian mercenaries fought for the Egyptian pharaohs - and they were written left to right in Caria (apart from the Carian-Lydian city of Tralleis) and right to left in Egypt. What makes this chart unlike its Lycian and Lydian neighbors is the mismatch at its core: most Carian letters resemble Greek letters, but their sound values are generally unrelated - a, o, s and u match their Greek models, and the rest of the alphabet reads like a rearranged keyboard. This chart lists all 49 Unicode codepoints, because Caria’s city-to-city graphic variation was extreme enough that some variants earned their own codepoints - the base-plus-suffix names (A2, SH2, MB2...) are city-specific shapes, not new sounds.
Carian was the last of the Anatolian trio to be read, and the chart still carries the scars: decipherment ran from John Ray’s 1981 work on the Egyptian-Carian bilingual tomb inscriptions to the Carian-Greek bilingual found at Kaunos in 1996, which confirmed personal names for the first time; many values remain provisional and debated, and a few letters (the C-18 and C-39 rows here) are wholly unknown. The leading explanation for the Greek-shape mismatch is Ignacio J. Adiego’s: the script was adopted from cursive Greek, then restructured for monumental carving by imitating the most similar Greek print forms without preserving their sounds. Unicode encodes the whole system at U+102A0 (version 5.1, 2008), this chart’s verified source.
How to use
- Read the table in block order; each row gives the letter, its official Unicode name, and the scholarly value with its suspected Greek origin. The rows to trust first are the four that behave themselves - a, o, s and u match their Greek shapes - then the named consonants (m, n, p, t, r, l, k) whose values the Kaunos bilingual pinned down via personal names. Treat every question mark in the value column as exactly that: scholars mark these values as provisional, and this chart preserves the uncertainty rather than smoothing it over.
- Use the suffix names as a city atlas: A2, SH2, K2, G2, ST2, TT2, MB2-4, LD2, E2 and the UUU family are variant shapes of the base letter used by particular cities of Caria, and the MB row is the Egyptian-variant form of the b row. If you are reading an inscription from a specific city, matching its letterforms to the right codepoints is half the work of identification. The two C-numbered rows (C-18, C-39) have no identified value at all - the honest boundary of current scholarship.
- Practice with the bilinguals, which are the only safe ground truth: the Egyptian tomb texts let you check names against hieroglyphic spellings, and the 1996 Kaunos Carian-Greek inscription verifies personal names letter-for-letter. The standard loop mirrors the Lycian one on this site - pick a name in the Greek column, transliterate it from the Carian column using this chart, compare. Watch the direction: Caria runs left to right (Tralleis excepted) while the Egyptian mercenary texts run right to left, so orient yourself before decoding.
Frequently asked questions
Why was Carian so hard to decipher?
Two reasons, one visible in this chart. First, the sound values of most letters are unrelated to their Greek shapes - unlike neighboring Lycian or Lydian, you cannot lean on Greek to guess values, because a Carian letter that looks like a Greek one usually does not sound like it. Second, the bilingual evidence is thin: progress waited on Egyptian-Carian tomb bilinguals (John Ray, from 1981) and a single Carian-Greek inscription found at Kaunos in 1996, which is why many values are still marked provisional and a few letters remain wholly unknown.
What did the 1996 Kaunos bilingual change?
Everything about verification. The Carian-Greek bilingual inscription discovered at Kaunos in 1996 confirmed personal names for the first time - names are the ideal test case because the same name spelled in Greek gives you a nearly complete sound-value crossword. The readings of John Ray and subsequent scholars were largely confirmed by it, and it remains the anchor text students use alongside this chart to check letter values.
Why does one sound have several codepoints?
Because the cities of Caria wrote differently - graphic variation between cities was extreme enough to have separate Unicode characters. The base-plus-suffix names on this chart (A2, SH2, K2, ST2, MB2 and so on) are city-specific shapes of a base letter rather than new sounds, so the block’s 49 codepoints cover roughly 30 alphabetic letters. Reading a Carian inscription starts with matching its city’s letterforms to the right codepoints - the Kaunos alphabet is thought to be the most complete set.
Did Carian really get written in Egypt?
Yes - and in the opposite direction. Carian mercenaries served the Egyptian pharaohs, and their communities in the Nile delta left a homogeneous variant of the script attested far from Caria, written right to left while the home-country inscriptions run left to right (the city of Tralleis is the exception there). Those Egyptian tomb bilinguals are precisely the texts that started the decipherment, so the diaspora row of this chart is also its rescue rope.
How is Carian related to Lycian and Lydian?
They are the three indigenous alphabets of Iron Age Anatolia, all ultimately rooted in the Phoenician-Greek tradition, and Carian is the odd one out: where Lycian extends the Greek alphabet transparently and Lydian stays close to it, Carian keeps the Greek shapes while scrambling most of the values. This site has all three charts - the Lycian, the Lydian and this one - plus the Old Persian cuneiform chart, so you can compare how one coastline and one century produced four very different answers to writing local languages in alphabet form.