Kana (Hiragana/Katakana): Two (essentially) iso-phonic(?) systems, where each symbol in one set has a corresponding symbol in the other set, both denoting the same sound value.
The set of forms are historically unrelated. There is little overlap in the forms. Competent readers will know both sets, but will lean them separately. Convention decides which set to use, but innovative uses are known that flout these conventions. Use of Katakana for foreign words is conventional. Having longer texts (book length) available in both forms, however, is very uncommon (never say "never"). The rules of layout are identical, spelling rules differ in the demarkation of vowel length. Widespread daily modern use Monofont support is a practical everyday requirement Encoded as two scripts Latin (Fraktur/Roman/italic): Three isophonic systems Forms historically related Some overlap in the forms (Some forms of "Fraktur" have what I call 'embellished roman' capitals, instead of true Fraktur shapes.) Knowledge of roman/italic only is widespread, but reading Fraktur can be self-taught. Convention decides which one to use, when they occur together, but innovative uses of Fraktur are common for names and titles, and misuse of italics is rampant. Use of roman for foreign words is a common feature of Fraktur texts. Use of italic for emphasis is a common feature of roman texts. Books published in Fraktur, have commonly been republished in roman style, as Fraktur has fallen out of common use. The rules of layout (ligating, hyphenating, etc.) are different. In Fraktur, emphasis is denoted by s e p a r a t i n g the letters, whereas in many languages w/o a Fraktur tradition, italics have taken on this role, and character spacing is used to justify lines. For languages with Fraktur tradition, separation is still used with roman, and automatic use for character spacing is an example of poor localization (!). No longer widespread use. Limited to attention grabbing (titles, names) and specialize (math) uses. Monofont support is not an everyday requirement, except in specialized notation (mathematics). Encoded as one script plus extension for mathematics. I think this is a complete summary. My belief is that if Fraktur was still common today, and more commonly used together with roman, and/or if Japanese usage rules for Kana were somewhat different, then the resulting encodings might well have been different in each case. From a purely rich-text point of view there is nothing that prevents treating the Kana as a single script. On the other hand, the layout rule differences make a simple font substitution awkward for Fraktur text of any length. So does the use of length mark for vowels in Katakan, vs. vowel doubling in Hiragana. No such issues exist for roman/italic. A./ PS: The set of 'scripts' unified with Latin is in fact a bit larger, if manuscript and handwriting styles are considered as well. Some handwriting styles (Suetterlin) are so different that considerable training is required to read them. PPS: I don't care to distinguish between 'conventions' and 'rules'. A tendency of considering conventions as/in tersm of rules, is quite conventional in Germany ;-)