Akkadian Cuneiform Translator

How it works

Akkadian is a syllabary. A sign stands for a, ba or ab — never for a bare consonant. So English is first reduced to sounds, then cut into syllables, then written one sign per syllable. Three facts about the sign inventory shape everything you see.

Sounds Akkadian doesn't have

There is no o f v j ch sh th. Each folds onto the nearest sound Akkadian does have, which is why fox becomes buks and jumped becomes zumped.

Closing consonants, and the vowel-killer

Signs for vowel-plus-consonant exist only patchily: Unicode has ab ib ub ad ud ig ak al el il im um an en in un ir ur and nothing at all for syllables closing in h, p, s, t, z or q. Where a sign exists it is used, because that is how Akkadian really wrote it. Where none exists, the consonant is written with a consonant-vowel sign followed by a vowel-killer mark , meaning “the vowel before me is silent”. That mark is this page's own convention, not an ancient one. It is what lets the reverse direction be exact instead of guesswork.

Where it loses information

Cuneiform back to sounds is exact. Sounds back to English cannot be: the spelling was thrown away on the way in. quick goes out as k-u-i-k and returns as kuik. There is also no pu sign, so it borrows bu — Akkadian wrote p and b alike — and a few Ce values borrow their Ci sign. Those borrows are one-way, so jumped returns as zumbid.

The signs

Every sign used, with the syllable it writes.

Source

The Python original, the test suite and this JavaScript port are on GitHub. Sign values were assembled from the Unicode names plus common Akkadian readings; the British Museum's how to write cuneiform is a good place to check them.