The cuneiform script was first deciphered by the
teacher Georg Grotefend (1775-1853), whose work was completed by
Henry Rawlinson (1810-1895), working with the
Achaemenid Persian inscriptions of
Darius I. The first results were published in 1850, and since then
Assyriologists and
Sumerologists have worked their way backwards, to the earliest
Sumerian pictographic signs.
Because of their early pictographic nature, most signs represent more than one ideographic or phonetic value, usually of no more than three letters. Thus, the ka sign, originally a stylised picture of a head and neck, meaning "mouth", also serves for gù, "to call, name", and inim, "word, speech".
The signs were only formalized over time, taking their final form in the Neo Assyrian period. The first signs are obviously representative, and often allow for curved elements, with individual strokes often freely disappearing or re-appearing according to the individual scribe. Eventually, the script was reduced to combinations of 5 strokes: an horizontal with the angle beginning at the left, a downward sloping, an upward sloping, a "winkelhaken" (basically a wedge opening towards the right and no real linear extension), and a vertical with the angle beginning at the top.
Since the days of Grotefend and Rawlinson, many new phonetic values have been discovered for each sign (the current sign list numbers around 500 individual signs). In publishing a text, a transliteration is included from which it should every time be possible to reconstruct the cuneiform text. Since many signs have a value ka (it is assumed that there were originally slight differences in pronunciation, though these are impossible to reconstruct), every time a sign is discovered to have this value, it is transcribed with the value and a sequential number. The first value is thus transliterated as simply ka, the second ká (with an acute), the third kà (with a grave), and the rest numbered with subscripts (thus, e.g., ka4, etc.).
Sign lists are constantly being updated, though it is a slow process. The standard sign lists are Labat's Manuel d'epigraphie akkadienne, which serves as an excellent epigraphic reference though fails in the sign-values, and Borger's Assyrisch-babylonische Zeichenliste. Both are organised according to the neo-Assyrian values.
I'll add an addendum (03-18-01), just because it's 5:00 in the AM and I'm still up working on dear Hammurabi's
law codes (I'm around paragraph 65 or so). This stuff can be quite
maddening, really...signs blur, and without a
transcription, it can take hours to piece together the correct values (is it a
ga? a
ka? a
qa? each would give a different verbal root, and then you realize, hey! they're throwing in Sumerian just for the fun of it!). There's a sort of
academic urban legend about an assyriologist who went dotty as a doughnut in his later years, and started translating the bird-tracks on his frosted window sill. It's pretty much what I feel like right now.