XHTML M12N: Tokenization of "public text description"

Dear HTML Working Group,

  Section 3.1 of the XHTML M12N SE WD requires

[...]
  2. The schema which defines the document type must have a unique
     identifier as defined in Naming Rules that uses the string "XHTML"
     in its first token of the public text description.
[...]

It is not clear to me what is considered a "token" in a "public text
description", please state this more clearly in the next working draft.
http://www.w3.org/MarkUp/SGML/productions.html#prod87 would suggest that
tokens are separated using RS/RE/SPACE, is this correct? Or would they
be split at RS/RE/SPACE/Special? Or yet something else? Do I understand
correctly that "NOTXHTML 1.0" is considered to "use" the string "XHTML"?

regards.