On 5/1/2012 11:19 AM, Michael Everson wrote:
It does not matter if sideways text can be read as words, or just as gibberish.
Good practice and typographic design will not rotate syllabic text because of
the inherent confusability.
Michael has a generally valid point. Rotating *small* snippets of Syllabics
text would be inherently confusing, because the entire script is built out
of a relatively small repertoire of geometrically shaped base letters, for
which many of the rotations and reflections turn them into distinct letters
with different readings. See the first example that Suzuki-san provided a
link for, which prominently has a KA a NU and a NI syllable, where these
3 fall into a rotational set. This is the Latin pbdq problem, but much more
inherent to the graphical structure of the glyphs.
However, I take issue with Michael's immediate trashing of the typography
of these two examples. The second example, in particular:
http://www.cley.gov.nu.ca/pdf/Documentary%20Art%20Project_Inuk.pdf
is quite clearly part of the typesetting of something which could then be
folded into thirds as part of a hand-held piece of collateral -- a
little pamphlet.
And the rules for what is well-set on a pamphlet are rather different that
what makes sense on a unrotatable computer screen display, precisely
because the pamphlet is in your hands and can be--in fact is designed to
be--fiddled with and rotated manually. In such cases laying out text boxes
with arbitrary rotations is a perfectly reasonable thing to do.
On the other hand, that Project Inuk pdf example then fails as a real
example
of vertical typesetting in the sense intended by TR 50, I think, because it
isn't *meant* to be read vertically.
Other examples of cases like this might be the printing of spines on
publications,
which also are often not meant to be read vertically.
--Ken
