In our context, transliteration means representing cuneiform signs in the Roman alphabet, with the addition of a few non-Roman letters (š, ĝ/g̃ and ḫ), using hyphens and spaces to indicate sign boundaries (more about this in the document on hyphenation practices). A letter or a sequence of letters in lower case is called a (transliteration) value, and it may represent a word (logogram) or a grammatical morpheme (phonogram). Take the sign (NI), for example. It can be transliterated i3 with the meaning 'oil'. However, it can also be transliterated zal and mean 'to pass' as in 'to pass the time' (ud zal). This one-to-many relationship between sign and values is a real challenge for anyone transliterating Sumerian. It seems to have been so for the Mesopotamian scribe as well, and may be one of the reasons why determinatives are used so frequently. Determinatives are semantic classifiers written before, or sometimes after, another sign to show the meaning or semantic category of the following (preceding) sign. The most frequent determinative is (AN) written before a deity's name. Another is (URUDA) written before objects of copper. Determinatives are thought not to have been pronounced, and are therefore marked in a special way in transliteration. On the ETCSL web site, they are displayed in superscript or as entities, bounded by & and ; (semicolon), e.g. &urud;.
There are a number of problems associated with reading cuneiform writing, e.g. interpreting the handwriting of the individual scribe or deciphering the signs on a tablet in poor condition, which we will not enter into here. The main problem, however, when transliterating cuneiform signs is their polysemous nature. Just like a word in English can have more than one meaning, e.g. 'might', so can a Sumerian sign. (UD) for instance can stand for 'sun', 'day', 'storm', 'white', and 'to shine'. In Sumerian, as in English, the context will decide which meaning is intended. However, the more meaning potential a word or sign has, the more difficult it is to assign the correct interpretation. When transliterating a Sumerian text, we also interpret it by assigning different values to the individual signs based on what we believe the most likely meaning is in a particular context. Sometimes the several values of a sign have related meanings, making it even more difficult to decide on a value even with an extended context. One such sign is DU , which has the values du (a form of the verb 'to go'), de6 ('to bring/carry'), gub ('to stand'), and ĝen ('to go') among others.
Another challenge for the aspiring Sumerologist is the fact that different signs are used to code the same phonetic value, that is homophony. To distinguish between homophonous signs subscript numbers have been introduced. gu ( = GU), gu2 ( = GU2), and gu3 ( = KA) are good examples. An English parallel would be the word 'bust'. If you look this word up in a dictionary you will see that there are two entries, (often) distinguished by superscript numbers, although the words are pronounced the same. This is because they have different, unrelated meanings, just like gu = 'cord/net', gu2 = 'neck/bank', and gu3 = 'voice'. In the example with gu/2/3 the same phonetic value stems from different signs AND have different meanings. This is not always the case. Sometimes different transliterations (graphic values) have the same meaning whether they stem from the same sign or not. This is called heterography. The values ba, bad, be2 are cases in point. They can all mean 'to open'.