In the Hittite text CTH 591, we see this curious passage:

9-an  ŠI  9-an  SIG₇-an          9-an  GAŠAN-TI MU-ḪI.A-uš  a-aš-šu-ša-aš  
9-ACC eye 9-ACC yellow.green-ACC 9-ACC lady- ?? year-PL-ACC good.ACC.PL

On the face of it, this doesn't make any sense. But the logogram ŠI "eye" looks like a 10 next to a 100, and thus can also be used to mean "1,000"; we know the Akkadian word for "thousand" is līmu, so this sign has also been given the reading LIM.

SIG₇ "yellow-green", then, looks like the sign LIM with some extra strokes decorating it, probably being used here to mean 10,000. We don't know how the word for "10,000" was pronounced, but conveniently it looks the same as the logogram for "yellow-green", so we can use that name for it.

The signs GAŠAN-TI together, then, look like SIG₇ with even more decoration added. So presumably (at least according to Hoffner), this should actually be read as a single sign, probably meaning "100,000". In which case, this whole passage is actually:

9-an  LIM      9-an  SIG₇-an         9-an  ?????           MU-ḪI.A-uš  a-aš-šu-ša-aš  
9-ACC thousand 9-ACC tenthousand-ACC 9-ACC hundredthousand year-PL-ACC good.ACC.PL

That is, "nine hundred ninety-nine thousand good years". This makes a lot more sense than "eye", "yellow-green", and "lady".

But if this is the case…what can we call this sign for "100,000"? It never seems to be used outside of Hittite, and we don't know a Hittite, Akkadian, or Sumerian word for this number.

What is the standard in Assyriology, when we know the meaning of a sign through context, but don't know any viable reading? What do we call the sign in that context? (The classic Hethitisches Zeichenlexikon sidesteps the issue by refusing to acknowledge this as a single sign, but in a system I'm working on, I think it deserves inclusion…which means it needs a name.)

1 Answer 1


A relatively common convention (see e.g. the ETCSL sign list and Wikipedia) is to notate such "juxtaposed" compound signs by joining the component sign names with a period (.). That is, your compound sign made up of GAŠAN juxtaposed with TI would be named simply "GAŠAN.TI".

One potential drawback of this notation is that it can conflict with the (also somewhat common) use of periods to separate the signs of a multi-sign logogram like HI.A.* If you want to use both conventions, you'll need some way of disambiguating them.

In ATF notation (see the AFT Inline Tutorial, ATF Quick Reference and CDLI ATF Primer on Oracc), compound signs are delimited with vertical bars, so your sign would be written as |GAŠAN.TI| in ATF. Of course ATF also does not use periods as delimiters except in compound signs, but still, this would be one way to avoid the ambiguity.

In any case, if you really have no phonetic reading to suggest for the sign, and aren't planning to publish your own sign list (in which case you could in principle refer to the sign with an abbreviation of your sign list's title and its number in the list), I think this is just about the best you can do.

Ps. If possible, it's also worth checking the original tablets and/or tracings of them to see what the actual signs on clay look like. For such "dubious compounds", it's not uncommon for the actual sign shapes to differ in some detail from the conventional shapes of the supposed component signs. (See for example the note about the orthography of KEŠ3 in one of my earlier answers here.)

If they do, you might want to reflect this in your transcription; for example, if the "TI" part of the sign turned out to be consistently missing wedges or looked unusual in some other way compared to a normal Hittite TI, you could consider it an allograph and notate that e.g. as |GAŠAN.TI~v| in ATF.

FWIW, such variations in form, if they indeed exist, would IMO also support your claim that this should really be read as a single distinct sign, rather than as merely a sequence of GAŠAN and TI.

*) Of course the distinction between a compound sign and a multi-sign logogram can often be quite blurry in the original cuneiform text as well.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.