The dynamic flow unit word must be interpreted as being dependent upon
the language or writing system of the affected content. If the language
or writing system is unknown or unspecified, then word is interpreted
1. If the affected content consists solely or mostly of Unified CJK
Ideographic characters or of characters of another Unicode character
block that are afforded similar treatment to that of Unified CJK
Ideographic characters, then word is to be interpreted as if
character were specified.
2. Otherwise, word is to be interpreted as denoting a sequence of one
or more characters that are not interpreted as an XML whitespace
Noting the "must" which is a testable conformance requirement, do the
following paragraphs contain one word or two?