one more response

[Arnold Zwicky]

mark mandel:
 >The basic reason is that lowercase Roman lettering is rich in
 >ascenders and descenders, so that many words have outlines that are
 >much more distinctive than the uniformly rectangular outlines of
 >Roman uppercase. Fluent readers have learned to use this
 >information, along with many other cues...

but this is just the the-more-familiar-is-easier argument again.  i
don't think anyone's arguing with that, though you might question
whether the difference in ease (measured by, say, reading speed and
comprehension) is big enough to care about.

but if there's an *intrinsic* advantage to mixed-case printing, then
there should be a difference in the performance of people who are
experienced and fluent readers of all-caps (but not mixed-case) text
vs. people who are experienced and fluent readers of mixed-case (but
not all-caps) text.  and people who grew up, so to speak, in one
scheme but are trained on the other should take longer to shift in one
direction than the other and should still show a difference in
performance in the two schemes after training.  a claim of intrinsic
difference *is* testable, just not easily testable.

arnold (zwicky at csli.stanford.edu)