On Monday, 23 January 2012 at 10:54:15 UTC, Walter Bright wrote:
On 1/23/2012 2:22 AM, Nick Sabalausky wrote:
Although I disagree with phonetic being *necessarily* better
than
ideographic. I do agree with the benefits of phonetic you
describe -
essentially "easier to learn". But the benefit of ideographic
is that they
can be quicker and easier to use *after* you've learned them.
I find that very difficult to believe. But I don't know Kanji.
Children and non-native speakers are taught the phonetic
alphabets first
(hiragana and katakana), because they're easier to learn and
can handle any
word with a small number of simple symbols. Then learners move
on to the
ideographic ones (the Chinese kanji). I only ever learned a
few kanji, but
you notice pretty quickly that once you've learned a kanji you
can read it
much more quickly than the phonetic equivalent. (It also helps
your brain
divide a sentence into words, since Japanese doesn't use
spaces, but that's
not really relevent here).
I've seen the same books written in both Kanji and English. The
English ones were smaller, significantly so. I suspect the
problem was the Kanji font had to be considerably larger in
order to be legible, which negated any compression advantage it
might have.
I think a big part of the reason kanji is easier to read (once
you've
learned it) is that your eyes don't have to move nearly as
much, and there's
much more visual distinction between words (since there's so
many more basic
patterns). The fact that they originate from images is
irrelevant since they
don't really retain much of the resemblance they once did (a
few of them do,
like "mountain" or "gate", but only if you already know how to
"see" it -
like being told the "box of kleenex" is a printer). It really
is exactly the
same as reading "42" instead of "fourty-two". Or the standard
VCR-control
icons instead of "fast-forward", "next chapter", etc. Totally
obscure if you
don't already know them, but much quicker and easier to read
then the
english words if you do.
As far as ability to look things up: Other ideographic
languages may be
different than this (and this certainly doesn't apply to
computer icons
either), but most of the Japanese kanji (ie, Chinese
characters) are
constructed from a smaller number of common building blocks,
the "radicals"
(around 100ish-or-so, IIRC?). As such, there actually is such
thing as kanji
dictionaries where you can look up an unknown symbol. (I
almost bought one
once...)
Getting back to software, I like the words when I'm learning a
program
(whether they're tooltips or labels) since the icons are
initially
meaningless. But once I learn what the icon means, I often
prefer to not
have the words because, compared to the icons, they're just
indistinct
visual clutter (and they take up that much more screen real
estate). The
color in icons also adds yet another dimension for your eyes
to lock onto
which text labels just don't offer, at least not as naturally.
I agree that color can help, but it helps just as well with
text. That's why we have color syntax highlighting editors.
Another thing to note: While the connection between an icon
and it's meaning
may not (ever) be close enough to initially teach you what it
does, the
metaphor (even for non-physical things) is usually close
enough, or logical
enough in its own way, to help you *remember* what it does
after you've
initially learned it.
I still can't remember which of | and O means "on" and "off".
Ever since the industry helpfully stopped labeling switches
with "on" and "off" my usual technique is to flip it back and
forth until it goes on. Is it really progress to change from a
system where 99% of the world knows what it means to one where
2% know? I suspect it is driven by some people who feel guilty
about knowing english, or something like that.
I remember in the 1970's when the europeans decided to
standardize on a traffic "stop" sign. They bikeshedded so much
over this, the compromise selected was the american octagonal
STOP sign. Nationalistic egos prevented selecting one from a
european country.
Bring up Adobe's pdf viewer. It has a whole row of icons across
the top. I defy you to tell me what they do without hovering
over each. Nobody has ever figured out a picture that
intuitively means "save", "send" or "print". Some icons do have
meaningful pictures, like scroll arrows. But the rest is an
awful stretch that is driven by some ideology <shatner>must ---
make --- icon</shatner> rather than practicality.
Back to Thunderbird email. The icon for "Spell" is ABC over a
check mark. That is not smaller or more intuitive than "Spell".
A few additional points:
# Microsoft allegedly does a lot of usability research and they
came up with the upcoming Metro design which relies on text
instead of icons. # Regarding the English language - Icons are
supposed to be universal so it saves money for companies to
localize their software. Localized UIs do present a trade off in
usability: It depends which terminology is more common, the local
or the foreign (English). E.g. "print" is easy to translate and
would be intuitive for non techies but "bittorent" probably isn't.
