Re: meaning of "two-level" in Han unification

From: John H. Jenkins (
Date: Mon Nov 17 2008 - 11:13:51 CST

  • Next message: Julian Bradfield: "Re: meaning of "two-level" in Han unification"

    On Nov 17, 2008, at 6:35 AM, Julian Bradfield wrote:

    > The Unicode Book describes the criteria for Han unification in terms
    > of a "two-level classification". Unfortunately, it doesn't explain
    > what the "two-level" means - does it mean the two dimensions of
    > abstract and actual shape, or does it mean that the ideograph
    > component tree is only examined for the top two levels?
    > In particular, I don't understand the 6th row of Table 12-5.
    > Why do the characters have the same abstract shape, when one has a
    > Claw radical and the other doesn't - they even have different indexing
    > radicals.

    The "two-level classification" has to do with the three-axis model
    and, as you surmise, means that the analysis of the y- and z-axes are

    As for the two characters in row six of Table 12-5 (U+70BA and U+7232,
    which both, BTW, have the same indexing radical, as they're both found
    under the fire radical), they're probably not the best example since
    they're separately encoded, owing to the source separation rule, if
    nothing else. And, arguably, they have a different abstract shape,
    too. They are, however, treated as z-variants in CCCII. Probably the
    best thing to do is to simply ignore the example because it's a bad
    one. In fact, just ignore Table 12-5 altogether because it really
    isn't making a clear point.

    A better pair of characters with the same abstract shape but different
    actual shape would be U+8AAA and U+8AAC. (They're separated because
    of the source-separation rule.)

    John H. Jenkins

    This archive was generated by hypermail 2.1.5 : Mon Nov 17 2008 - 11:18:15 CST