The real problem here is that Currier chose the words "language" and
"hand" in his description of statistical similarities. Much like my
personal problem with "character" and "letter", "language" and "hand" mean
different things to different people, even when the context for their use is
evident as statistical, not real. "Language" in these arguments does not
imply a difference in spoken or written language, and Rene makes this point
as well. "Hand" does not signify that someone has actually determined that
this was another writer. These are Currier's definitions for statistical
"language" and apparent "hand", nothing more. I sincerely wish we could
come up with better and less confusing terms for these things. I've already
introduced "glyph" to narrow the wide variation of perception about the
actual transcription of a "character", maybe someone else can find something
to mark differences in "language"?