Path: utzoo!attcan!uunet!lll-winken!ncis.llnl.gov!helios.ee.lbl.gov!pasteur!ucbvax!decwrl!labrea!rutgers!bellcore!geppetto!duncan
From: duncan@geppetto.ctt.bellcore.com (Scott Duncan)
Newsgroups: comp.society.futures
Subject: Re: New Input Devices (in lieu of Keyboards)
Message-ID: <13549@bellcore.bellcore.com>
Date: 20 Jan 89 12:33:52 GMT
References: <8901121714.AA26346@multimax.encore.com> <400011@hpdsla.HP.COM>
Sender: news@bellcore.bellcore.com
Reply-To: duncan@ctt.bellcore.com (Scott Duncan)
Organization: Computer Technology Transfer, Bellcore
Lines: 75

In article <400011@hpdsla.HP.COM> garye@hpdsla.HP.COM (Gary Ericson) writes:
>
>...what is it we're trying to do by typing on a keyboard?  ... communicating,
>both with the computer and with other human beings (sometimes indirectly
>through written documents and directly with email, etc.).  We don't have to
>use a natural-looking language to talk to the computer (in fact, a lot of what
>I type for the computer looks like garbage to me
	[...]
>Without a computer, how do I communicate with others?  By talking (English),
>writing (English), gesturing (body language), and drawing pictures
>(graphics[?]).
	[...]
>                             This makes me think that maybe we're too myopic
>with respect to input devices, considering only textual input.  True, most
>computer systems today lean heavily on textual communication (even windowing
>systems with pull-down menus tend to just allow the user to enter text by
>selecting words from a list - it's still text, you just are able to avoid
>explicitly typing it), but I think that's just historical, caused by the lack
>of alternate input devices resulting from limited technology.  I wonder if
>things would change much if DataGloves and other exotic devices were cheap and 
>available.
	[...]
>To incorporate drawings into communications, I think the interface would have
>to allow you to *fluidly* (is that a word?) move between text and drawings.
	[...]
>When you write, you record the individual letters of words, and thus we have 
>invented keyboards that let you enter those letters quickly.  But when you 
>talk (and when you read, especially if you're a speed-reader), you communicate
>whole words/sentences/ideas - you don't spell out each letter (in talking, you
>do pronounce most letters, but not all of them, e.g., "silent" letters and 
>combinations like 'ch').  That's why talking is typically faster than writing.
	[...]
>Is there any way to come up with a new device that will allow you to input
>thoughts at this kind of higher level, rather than letter-by-letter?  My
>impression is that shorthand does this as does court reporting.

I think one important issue here is that we have managed to more easily come up
with computer applications which can process text more effectively than they
can process other forms of information.  It is not simply a matter of getting
computers to accept (and reproduce upon output) other forms of information than
text characters, it is also a matter of representing this data internally and
having some agreed upon rules (syntactic and semantic) for processing it.

In the above discussion, one aspect of why talking, and gesturing, and drawing
are faster than writing is that the human processor does a darn good job at
comprehending the meaning of the data coming into it!  We do NOT have widely
available, efficient software to do lots of the processing of images and voice
input that the human brain has (matphorically speaking regarding brain software
that is).

At a local ACM meeting a few years ago, Brian Kernighan was asked about various
workstation technologies as compared to the 'little languages' approach.  He
made the point that, once you introduce the need to hve a human being interact
with the computer, i.e., the computer cannot (or is not supposed to) process
on its own, then you reduce some of the power of the processing that could go
on and diminish the ability to link together tools as effectively.  (All this
is my interpretation of what I heard, so you can ignore that I claim Kernighan
said it and just accept the opinion on face value if you like.)

I guess what I'm getting at is:  talk about new input and output technology is
fine, but what do we expect the computing system, apart from human interven-
tion, to DO with the other kinds of data?  It was have to be translated into
something the computing system can manipulate and that's still a serious prob-
lem since images, gestures, etc. are more open to interpretation as to their
meaning than text.  (Yes, I know the myriad interpretations that can be made
of a given natural language expression, but we have, at least, agreed on some
subset of these for programming languages, etc.  People still argue over what
icons should mean and whether dragging a disk icon to a trash can is proper.
We need to agree to some of this before we can expect the computing system to
deal with raw graphic input I would imagine.)

Speaking only for myself, of course, I am...
Scott P. Duncan (duncan@ctt.bellcore.com OR ...!bellcore!ctt!duncan)
                (Bellcore, 444 Hoes Lane  RRC 1H-210, Piscataway, NJ  08854)
                (201-699-3910 (w)   201-463-3683 (h))