The current suggested methodology for dealing with cuneiform merger and split trees is to encode just the roots of the mergers and leaves of the splits. I have always objected to this and here is why.
If you look at the attached example of the KU series of mergers (taken from D. O. Edzard's Reallexikon der Assyriologie "Keilschrift" article), the current suggestion would be to encode only the 6 roots - DUR2, KU, TUG2, NAM2, ESH2, and HUN. The problem with this is the ambiguity that results in plain text from such an approach. If I input "KU" how do I know in plain text which KU this is? The one with 6 ancestors, the one with 2, or the one with none? The reality is that, due to the diachronic nature of cuneiform, there is only one KU sign, then there is a KU/DUR2 sign, and finally a KU SERIES sign. There are not 6 characters in this tree; there are 10, and that's what we should encode. The complexity can be handled easily by input methods, one for each level in the tree. So when I am working on a later text I will input the character sequence "k" + "u" + SYLLABLE TERMINATOR and the input method for that period will enter the CUNEIFORM KU SERIES character; when I do the same thing for text in the earlier period the input method for that period enter the CUNEIFORM KU AND DUR2 character. Respectfully, Dean A. Snyder Assistant Research Scholar Manager, Digital Hammurabi Project Computer Science Department Whiting School of Engineering 218C New Engineering Building 3400 North Charles Street Johns Hopkins University Baltimore, Maryland, USA 21218 office: 410 516-6850 www.jhu.edu/digitalhammurabi
<<attachment: KU-series.gif>>
