Alan
Kay
Learning
Research
@POUgs
Xerox Palo Alto Research Center
Palo Alto,
Californis,
WA
Invited Paper:
Meeting
on
20 Years of
Compueei-
Science
Institute
di Elaborazione
d&a
Pn/ormazione,
PBS&
ITALY
-/
DATE
12-JUN-75
10:19:38
I
I
PAGE 2
/’
J’
Invited Paper:
/’
Meeting on 20 Years of Computer Science
Znstituto
di
Eloborazione
della
Znformazione,
PI&I,
ITALY
Personal Computing
by
Alan Kay
Learning
Research
Croup
Xerox
Palo
Alto Research Center
Introduction
Imagine having your own self-contained knowledge manipulator in a portable package the size and
shape of an ordinary notebook. How would you use it if it had enough power to outrace your
senses of sight and hearing, enough capacity to store for later retrieval thousands of page-equivalents
of reference materials, poems, letters, recipes, drawings, animations, musical scores, waveforms,
dynamic simulations, and anything else you would like to create, remember, and change?
Several years ago, we crystallized these long-held desires into a design idea for a personal dynamic
medium called the Dynubook. We felt that the hardware power for the Dynabook would be
inevitably available in ten years time, but there was no reason to believe that it would be easily usable
by its millions of potential owners if progress in man-computer communication continued at its present
rate.
In particular, we wanted our range of users to include children from age 5 or 6 and ‘noncomputer
adults’ such as secretaries, librarians, architects, musicians, housewives, doctors, and so on. We felt
strongly that the major design problems of the Dynabook lay in the area of communication
rather than in new hardware architectures.
Since it is very difficult to design such an integrated and comprehensive system ‘Aristotle fashion’
(from one’s hammock); we, with others at PARC, designed and built a number of stand-alone
self-contained
ZrLterint
Dynabooks in order to have a solid test-bed for our ideas. These machines
are the environment for our experimental communications medium, Smalltalk. Both the
Interim Dynabooks and Smalltalk have been used by children and adults.
The Dynabook Mockup
The Interim Dynabook
DATE
12-JUN-75
10:19:38
Design Background
‘The
first attempt at designing this kind of higher-level
personal
n~ etan~ ediunr
was the
de:relopment
of the FLEX Machine in
1967-69[1,2,31.
Much
of the
hard:rnre
and software was
successful from the standpoint of computer
scier,ce
state-of-the-art research but,
a?
is so often the
case, lacked sufficient expressive power to be truly useful to
a11
ordinary user. At that time started to
appear
Papert
and
Fcurzeig’s[g,lO,i
l]
pioneering
lvork
having to do with helping kids learn
how to think by giving them an environment in which thinking is both fun and rewarding.
A
Drawing of the FLEX
Machine
A
Turrle
Ena..
,-
“$I
Shown On Its Own Display Screen
*c
‘T.L.,,
8-=
x-4
ca.
1968[3]
&s&-
?‘hey
chose a time-shared computer and devised a simple though
comprehensive
language,
called*;
__
v:hich
combined some of the best features of
.JOSS[
161
and
LISP[
18).
Using LOGO, the
children
(ranging in age from
8-
12 years) learned to control a number of exciting activities: a robot
turtle which can draw, a faster CRT version of the turtle, and a simple music generator.
The LOGO work radiates a
comptllirr,
u
excitement in several directions.
First. the children really can program the turtle and the music box to do serious things. The programs
use
symbols to stand for objects, contain loops and recursions, require a fair amount of visualization of
alternate strategies before a tactic is chosen, and involve interactive discovery and removal of ‘bugs’ in
:heir
ic?eas.
As
Papert
points out,
the
children are performing real mathematical acts of a kind,
scope, and level not ever achieved by many college graduates.
Second,
the
kids love it! The
interactivq
c
nature of the dialogue, the fact that they are in control,
the feeling that they are doing real things rather than playing with toys or working out
‘sc’hool’
problems, the pictorial and auditory nature of their results, all contribute a tremendous sense
cf accomplishment to their experience. Their attention spans are measured in hours rather than
minutes.
After seeing the faces of
children
suddenly
drscovering
that they are “doers” acting on the world,
rather than “things” being acted upon, it
WCLS
clear that the next attempt to design a
person&
con?puter
should be done with children strongly in mind.
First, having children as users would throw sharply into focus the expressive communication problems
x;hich
had caused difficulty with the FLEX Machine. In addition, it might be possible to discover why
the
LOGO kids
had
certain difficulties of their own
hnvin
g
to do with naming and parameters, with
partitioning their problems reasonably, and why they
appeared
to reach a plateau: they could design
and write certain kinds of constructive programs but never quite got to the point where they could do a
reai
system.
Second, children really need more computing and expressive power than most adults are willing to
settle
for when using a time-sharing system;
the
best that time-sharing can offer directly is slow
control of crude wire-frame green-tinted graphics
aml
(even cruder) square-wave ‘musical’ tones. The
kids, on the other hand, are used to high-bar:dwidth media such as as finger-paints, water colors,
color TV, real musical instruments, and hi-fi records. If the ‘medium is the message’, then the
message cf low bandwidth time-sharing is ‘blah’!
J
DATE
12-JUN-75
10:19:38
PAGE 4
Our Approach
At the outset, we decided to admit that the design of a truly useful dynamic medium for everyday use
was a hard but
extremely
worthwhile problem which would require both many years and several
complete interim hardware/software systems to be designed, built,
and
tested. Approach:
1. Conceptualize a “Holy Grail” version of what the
eventual
Dynabook should be like in the future,
This image will provide a rallying point and goal which will remind us of what
we
are trying to do
while the sometimes grubby spadework of producing intermediate systems is in progress.
An extrapolation (and compression) of the FLEX
hIachine[4].
2. Do the research in human factors, psychology of perception, physics, and language design which is
prerequisite to any serious attempt at an interim system.
Very few displays have been designed using any knowledge of the human visual system nor have many
artificial languages been developed on
non-Indo
European models.
An overview of what the Dynabook should be like, including display needs
and principles for language design, are found
inC4].
3. Design an interim version of the Dynabook, and build a considerable number of them.
We felt that this step and the next one are the critical ones in our research. We had to get to the kids
and adults as quickly as possible so as not to be led astray by our own assumptions and hopes.
4. Make the medium of communication as simple and powerful as possible.
It should be qualitatively sirnpler and qualitatively more powerful than (say) LOGO. It should be
qualitatively more expressive than the best state-of-the-art “grown-up” programming language for
serious
system design. It should be as “neutral” as possible to all conceivable simulations.
5. Explore the usefulness of such a
system
with a large number of short range projects involving many
users,
ages
4
to
60,
from
vurying backgrounds, and with different needs and goals.
This phase would involve developing all manner of simulated media, both old and new; finding ways to
teach the ideas in the system; do user studies, experiment with peer-group teaching (e.g. 13 year olds
teaching 12 year olds), and so on.
Some of the projects we have undertaken arc explored in
[6,7,8].
6. Re-extend the system in the light of our previous study and start to think about tlre next interim
version.
One of our guiding philosophies has been to do many working versions rather than to attempt a
long-range ‘complete solution’ in one fell swoop and risk not ever getting a working system.
7. Our current plans are to set up a community resource center containing a number of the interim
Dynabook systems for both open- and closed-shop use near school and playground ‘traffic’ patterns.
In order to find out some of the things we would like to know about how children and adults think
about their world, we really need to
concluct
a series of longitudinal studies which investigate
hoiv
daily and casual use of a dynamic medium affect people’s way of doing things
(and
their lives).
We currently expect to start setting up this facility in the very
mar
future.
DATE
12-JUN-75
10:19:38
The
Interim
Dynab~k
and
Smalltalk
.
The Interim
Dynabooh
is
8 completely self-contained system,
dcsigrwd
simulations and human factors experiments. To the
u;er,
it
nr,l~,lrs
a>
inserted a disk memory containing about
:500
lunge-c~ll;ivalt:nti
of mar
very crisp high resolution B&W CRT or a
!owcr
resol.ition
high-cluality
Other
input
devices
hclutle
a standard
typowriter
keyboard,
a ‘chord’ keyboard (for sending
simultaneous signals),
a
‘mouse’
(which inputs position as it is moved about on the table), and
varieties of organ-like keyboards for playing music. New input devices such as these may be easily
attached, usually without
buildin,
q
a hardware interface for them.
Inserting A Disk Memory
The Music Keyboards
The Typing Keyboard
The Mouse
Chord Keyboard on Left, Mouse on Right