Simulation and Dissimulation Severo M. Ornstein CPSR/Palo Alto

The word simulation, though not yet exactly a household word, is
nonetheless beginning to find its way
into common parlance. More than anything this results from the
harnessing of computer animation by
the mass mediaÑ particularly, but not solely, television advertising.
When the real thing is
unavailable, as in a fly-by of Neptune, or when the reality needs
sanitizing, as in the visceral impact of
aspirin, simulation comes to the rescue.

Of course animation is not new and has long been used to make
artificial images more convincing. Today
the art of simulation has developed to the point that it has become
necessary to identify television
simulations as artificial, so that we won't think we're seeing the
real thing. But, like the Surgeon
General's warning on an advertisement clearly designed to sell
cigarettes, disclaimers in fine print at
the bottom of the screen are somewhat hypocritical, since in fact we
are being asked to believe that the
simulation represents what really does, or might, happen. In fact
modern techniques of animation and
simulation have reached such a level of sophistication that real and
artificial images are often
deliberately and cleverly blended in just such a way as to blur the
distinction between them. This
allows reality to be seasoned with illusion and thus expanded. We find
this in movies such as Star Wars
and 2001 which take us on hypothetical voyages with an astonishing
degree of verity.

Today's urge to escapism fits well with the intentions of the
illusion-makers. When employed for
entertainment purposes, illusion is for the most part harmless enough.
But because simulation can be
quite compelling, it constitutes a powerful vehicle for selling both
products and ideasÑespecially those
that cannot be truly demonstrated or proven.

As consumers, we have all become somewhat immunized to routine
advertising, but simulation of a
more sophisticated kind is being used increasingly to sell things at a
far grander scale. When employed
for political purposes, illusion becomes diabolical and deception
becomes downright dangerous.
Numerous artificial images are today routinely employed in debates
about a whole host of important
issues. Star Wars, for example, was sold to the general public through
some of the most outrageous
television propaganda ever aired. We are all thus increasingly at the
mercy of those who perpetrate
convincing images and thus compel our attitudes and opinions.

As computer simulations take on growing importance in our lives, and
as their results now contribute
to reaching important conclusions and making major decisions, it
behooves us to stand back for a
moment and consider where we are going with this technology and ask
what are its strengths and
weaknesses. To help understand what computer simulation is all about,
it is useful to explore briefly
its psychological roots. We human beings manifest a strong urge to
comprehend the world in which we
are immersed. This is not just a matter of intellectual curiosity, but
rather an essential part of our
machinery for survival. If you understand how something works, you can
cope with it betterÑstay out
of its way, or harness it for your own purposes. Comprehension enables
prediction and clearly
prediction, even in limited quantities, constitutes a vital tool for
getting along in life. Without it, our
actions would be utterly whimsical, random, and hence ineffective.

Thus, though it is often unconscious, we are constantly seeking and
forming mental models for how
things workÑ what the rules are that govern the behavior or the
appearance of the things we confront
as we go through life. I have a rough model, for instance, of how my
automobile works. No doubt it is
inaccurate and incomplete, but it allows me to drive successfully. On
the other hand I recently acquired
a modern thermostat and although I programmed it carefully for the
behavior I wanted, it seemed only
approximately to be following my instructions. When I called the
service headquarters to find out what
was wrong I was told, "Just be patient, it's studying your house and
in a few days it will zero in on an
optimum strategy for satisfying your demands." My mental model had
been far too simplistic.

Sometimes our search for explanatory rules lights on a coincidence
which we incorrectly interpret as
causative, and then we get superstitious behavior. A charming instance
was presented by a rat being
trained by an experimental psychologist to wait for a fixed interval
between taps on its feeding lever.
For a long time it tapped away randomly and got fed only when the
interval between taps chanced to
match the desired interval. Then, at one point after a tap, the rat's
tail apparently itched, so it turned
around, gnawed its tail briefly, and then gave the lever another swat.
Lo, food. The rat pondered this
momentarily and then, turning, bit its tail again, and hit the feeding
lever, with the desired result. We
observe similar instances of superstitious behavior as people attempt
to manipulate some of today's
complex gadgetsÑparticularly computer programs.

Amusing as such stories may be, what they really illustrates is our
determination to superimpose
understanding and order on our worldÑrules that help us to get along
in that world more successfully.
It is this force which underlies the building of computer simulations.

A friend of mine is fond of saying that science is prediction. He's
right of course: science tries to
understand how things work, and really understanding how a thing works
means that you should be able
to predict its behavior. We have to use the rather general term
"thing," because science is omniverous
and tries to understand everything it encounters. The "things" we are
talking about here are any and all
collections of interacting parts, and what we are trying to understand
is the rules that govern their
behavior and interactions. The word system is better than "thing" but
the reader should remember that,
as used herein, it doesn't mean anything more specialized than just
some collection of interacting parts,
and that under that definition, virtually everything is a system.

We can describe the behavior of systems in a number of different ways.
Scientists, being fond of
precision and generality, like to do it with equations, which provide
a very terse yet complete means of
describing behavior. When our understanding of a process is
sufficiently complete that we can write
down an analytic equation which describes it, one that always appears
to hold true, then we feel quite
confident of our understanding and consequent predictions.

The bad news is that life doesn't limit itself to tractable problems
and as systems become more
complex, so do the rules that govern their behavior. Thus for many
systems we are simply unable to
write down suitable equations. The good news, however, is that even
without equations, we are often
able to specify rules for how something works. Although such rules may
be more complex and thus less
gratifying than a nice clean equation, they may nonetheless predict
behavior accurately. A collection of
rules which purport to describe the behavior of some system constitute
a model of the system. We
verify the accuracy of a model by trying it out and comparing the
predicted results with the actual
results. Thus for example we might have a model of the weather that
predicts tomorrow's weather based
on information about today's weather and other relevant factors. We
can test the validity of such a model
by seeing how often it makes correct predictions.

Computers come into the act because the rules that people dream up,
which they believe govern
systems, are often so complicated and have so many variables that
figuring out their combined effects
requires a lot of the sort of dreary computation and manipulation that
computers are so good at. This
process of exercising a set of rules that purports to describe the
behavior of an actual system is called
simulating the system, and when a computer does it, it is called
computer simulation. Thanks to more
and more powerful computers, such simulation has become extremely
sophisticated and today provides
a powerful means for demonstrating how a theoretical model of some
system works. This facility is
applied to everything from predicting the outcome of elections to war
gaming. A computer simulation
was central to the prediction of ``nuclear winter" which has aroused
worldwide concern.

But in employing computer simulation, it is vital to remember that the
only thing the computer does is
to say, "Here's what your rules predict the system will do." The
rules, of course, may be utter nonsense
and/or they may be improperly implemented in a particular computer
program. So before we pay
attention to the results of a simulation, it is vital that we test and
verify its validity by comparing its
results and conclusions with behavior of the actual system it purports
to model in a statistically
significant number of cases. All too often this validation step is
slighted or bypassed altogether. People
tend to take simulation results on faith, without questioning the
validity of the model, partly because
they emerge from a computer, and computers are felt by many to be some
sort of infallible authority,
and partly because in any case it is extremely difficult to analyze
and challenge someone's else complex
model.

If a model proves correct over a large number of cases, we may then
begin to have sufficient confidence
in its validity to act upon its predictions without further
verification. Thus, although no one would be
foolish enough to base some crucial action on next week's weather
prediction, we do tend to build
bridges and buildings based on models of architecture and strength of
materials that have developed over
a long period of time and have proven trustworthy for the most part.

How do we look at the "results,, of a simulation and decide whether or
not they are reasonable? Often
the output is just sets of numbers or tables that are exceedingly
difficult to comprehend or to compare
with real system functioning. Consider, for example, a weather
prediction model which produces
sequences of tables showing predicted weather parameters at a large
number of locations. We can, of
course compare these numbers with actual measurements, either by hand
or through a program that
does statistical comparison for us. But in the process we begin to
have to place increasing trust in
sophisticated computer programs, not only for properly implementing
proposed models, but in
weighing the results. Where operations take place so much "behind the
scenes," it becomes increasingly
difficult to be certain that the things we think are happening are
actually happening. We would like,
instead, to "observe" the computer model operating in the same sense
that we observe the real-world
functioning so that we can compare their overall behavior. Modern
computer graphics provide just
such capabilities. These days the results of sophisticated simulations
can often be rendered to a viewer
in ways that closely mimic the appearance of the real system, thus
facilitating evaluation and
comparison with the real system.

This appears to be an unmitigated blessing, but there is a downside.
If a model is grossly inaccurate in
some way and if the real system being simulated is one with which we
have substantial familiarity,
then any mismatch between the theoretical model and the actual system
will likely be quite obvious.
However, without both of these preconditions, we may be fooled into
believing that a model is valid
when in fact it contains serious flaws. The very technology which
provides us with a seemingly
convenient means of evaluation, substitutes subjective observation for
objective measurement and
comparison. While certain kinds of errors may thus be highlighted,
others can slip by unnoticed, and
will invalidate later predictions and render actions based on those
predictions totally inappropriate.

Often, of course, a realistic view is precisely what is wanted. For
instance it is important that flight
simulators present as realistic a view to the pilot trainee as
possible. But it also important that the
physics and mechanics represented in the simulator match those of the
real world and the real aircraft,
otherwise the trainee can acquire dangerously incorrect responses.

In fact the very realism of today's computer simulation is the root of
an even more serious problem. It
is all too easy to begin to believe that what you are seeing on the
computer screen represents the
behavior of the actual system, as opposed to some theoretical model.
Furthermore, we often need to
understand the potential behavior of systems whose actual behavior,
for one reason or another, it is
impossible or impractical to test. This is particularly true of large
proposed man-made systems. For
example, the building of a large dam may have widespread consequences
which are important to
understand before deciding on construction. Simulation provides a tool
for studying the potential effects
of such proposed systems, but what is all too often forgotten in this
process is that the predictions are
only as good as the underlying model and that in such cases the
understanding of the model builders is all
we have to go on since the process of comparing theoretical with
actual results is impossible.
Overlooking the need for this important step is a little like
mindlessly moving from secure to doubtful
scaffolding. Nonetheless, we often take this step without seeming to
care, apparently because computer
simulations have proven valuable in the past (albeit on fundamentally
different, verifiable, problems)
and because they look so convincing on the screen.

And of course it is, as usual, a short step from simple misuse to
willful abuse. Computer modeling and
simulation are today being used increasingly to lend weight to
arguments in many kinds of debates. They
offer an extremely convincing means for presenting an argument. The
military regularly uses
simulations as a means of selling programs to Congress, and the
"simulation" of Star Wars' "protective
shield" was broadcast to sell the program to the public. Often these
simulations have no more foundation
than an "artist's conception," but the vehicle of computer graphics
can lend fantasies an air of realism
that is frequently sufficient to generate funding.

Identifying deliberate distortion is difficult enough in any
complicated situation, and can be virtually
impossible when the distortion is cloaked inside a computer model
containing some subtly biased
assumptions that produce the desired rather than the actual answer.
The fact that simulations can now
be presented in an extremely compelling fashion makes them an
attractive tool for those wishing to
advance a particular viewpoint about how a proposed system will work.
It's the old case of "figures lie
and liars figure." The person with a computer simulation on his screen
which proves his point, is a
long way toward winning an argument today, regardless of the validity
of the point. The opposition is in
no position to unravel the steel wool of simulation code and expose
its inadequacies. Serious arguments
take place today of the form "my simulation" vs. "your simulation,"
and decision makers, even with the
best will in the world, are utterly out of their depth in deciding
where the truth might lie.

This is not to suggest that all simulation is deliberately misleading.
In fact the real culprit is
complexity, which can confound honest and dishonest alike. It is not
clear that there is any easy fix to
this problem. Increasing complexity seems to be a fact of modern life,
and even the most sincere efforts
at understanding must confront our limited powers of comprehension.
The computer offers a powerful
tool for dealing with complexity, but we must remember that it is not
a magical one and that the rules
embodied in its programs are in fact nothing more than the
speculations and best guesses of the usual,
very fallible, human beings. Until we come to accept this fact, and
stop bowing down before computers
as though they were some sort of instrument of divine wisdom, we will
continue to be suckers for every
charlatan with a clever hacker in his employ.

Severo M. Ornstein is one of the co-founders of CPSR and former
chairman of the Board of Directors. He
is retired, having worked for many years at Xerox Palo Alto Research
Center.

Wargames Revisited Charles A. Meconis

"Interesting game. The only way to win is not to play."

That memorable computer-generated line appeared at the end of the
popular 1983 movie Wargames.
Computer professionals scoffedÑjustifiablyÑat the film's implausible
premise: a teenage hacker
breaks into a Pentagon computer whose artificial intelligence can
initiate global thermonuclear war on
its own.

The movie greatly overstated its case but made a valuable point about
the increasingly controversial
role of computers in controlling nuclear forcesÑa point reinforced by
revelations about false alerts and
real hackers getting into DoD computer systems.

The Hollywood emphasis has always been on "bolt out of the blue"
nuclear war (called a BOOB attack by
defense experts who know how unlikely that is). Largely overlooked is
the critical role that
computer-assisted or entirely controlled wargames now play in
justifying military strategies,
operations, tactics and force structure that may actually cause or
escalate crises that lead to nuclear
war.

In his book Wargames: The Secret World of the Creators, Players, and
Policy Makers Rehearsing World
War III Today, (Berkeley paperback, $4.95) Thomas B. Allen has
provided a fascinating and readable
introduction into this field that any educated person can understand.

After a long examination of the history and current role of wargames,
Allen concludes: "The value of a
wargame is that it shows the results of a war when it is played as a
game, with rules. . . . The nuclear
threshold is the place where war by the rules ends. Beyond that
threshold no war game can go, for
beyond that threshold there are no rules."

Unfortunately, this is not the conclusion reached by the military-
industrial-academic "iron triangle."
Because of the increasing expense and complexity of actual military
exercises and hardware, and the
increasing availability and (relative) affordability of computer
simulations, there is greater reliance
on computer simulation in the U.S. military and defense industry today
than ever before.

For example, the AAI Corporation has an advertisement in the August,
1989, issue of the United States
Naval Institute Proceedings. Under the headline "Great Pretender," we
see a Perry Class Frigate at
pierside with a white U.S. Navy truck-trailer connected to the ship by
several cables. The text for the
ad reads, in part:

From within the quiet unassuming trailer parked on the Navy pier,
battle rages. AAI's pierside Combat
System Trainer is providing the most realistic, economical alternative
to at-sea exercises. in the ship
docked nearby, radars, sonars, EW [electronic warfare] sensors,
communications, decoy and weapons
systems act and react. The crew experiences and handles simultaneous
threats from surface ships,
aircraft, missiles and submarines. Even adverse weather and sea
conditions. Without ever leaving the
pier. Very real. Yet very simulated. And extremely cost effective.

So what? one might ask. Here's what: the crew of the Ticonderoga class
ultra-computerized Aegis
cruiser U.S.S. Vincennes went through the most complete computer
simulation and exercise training the
Navy could deviseÑ and in the heat of combat in the Persian Gulf its
crew misidentified and shot out of
the sky a civilian airliner, killing nearly 300 noncombatants. Suppose
the plane had been a Soviet
airliner during a superpower crisis?

The publicly-released Navy report on the Vincennes incident pointed
out that, a) this was the ship's
first actual combat involvement; b) the ship was being attacked by
Iranian gunboats; c) one of the
ship's gun turrets had jammed, forcing the Vincennes into sharp, steep
turns to let the remaining gun
shoot at the Iranians; and, d) that was why the crew failed to read
the accurate data that showed that the
aircraft was emitting a proper ID signal and was gaining, not losing
altitude.

The Navy report laconically concluded that the Aegis display user
interface could use some
improvements, and that the Navy should do more research into "human
stress." Accidents will happen.

Because U.S. naval forces are almost always the first to be deployed
in any crisis, they are the most
likely to be involved in the escalation of a crisis into war, and into
nuclear war. (See Desmond Ball's
"Nuclear War at Sea," in the Winter 1985/6 issue of international
Security.)

There are many examples of naval forces being the catalyst for
conflict: "Remember the Maine;"
"Remember the Lusitania; "The Ballad of the Reuben James;" "Remember
Pearl Harbor;" the Cuban
Missile Crisis naval blockade; the Vietnam Gulf of Tonkin "attack;"
the Pueblo Incident; the Israeli
attack on the U.S.S. Liberty; the U.S. aircraft carrier and battleship
raids on Lebanon and Libya. As this
is written, French and U.S. naval forces are maneuvering in the
eastern Mediterranean in response to
the hostage crisis in Lebanon.

In fact, no aspect of the Reagan military buildup years is more
troubling than the provocative,
aggressive, and potentially disastrous "Maritime Strategy" so well
articulated by the brilliant and
acerbic former Secretary of the Navy, John F. Lehman.

In brief, the Maritime Strategy declares that at the very first sign
of hostilities anywhere in the world
between the U.S. and the U.S.S.R., the U.S. Navy will attack Soviet
naval forces and the Soviet homeland.

American nuclear attack submarines will enter waters close to the
Soviet coasts and begin sinking their
nuclear ballistic missile submarines. American aircraft carrier battle
groups will attack key Soviet
air and naval bases on Soviet territory. American amphibious battle
groups will land Marines, possibly
on Soviet territory along their Pacific coast. (see the January 1986
issue of the U.S. Naval Institute
Proceedings).

The strategy has many critics, not a few of them in the Navy itself.
How is this planÑand the 600 ship
Navy to implement itÑjustified? Here, in John Lehman's words, is the
answer:

We successfully established a process continually to test, refine, and
update the strategy .... We test it
repeatedly in war games and simulations, especially the annual global
war games at the Naval War
College in Newport, Rhode Island. It is used as the planning base for
our major exercisesÑnavy, joint,
and multinationalÑ and the exercises themselves become the crucible
for testing its execution and
challenging its assumptions.

In other words, computer war simulations form the basic justification
for the U.S. Maritime strategy,
and the buying of lavishly expensive hardware.

This is exactly where an organization like CPSR has a crucial role to
play. Who knows better the
usefulnessÑ and the real limitationsÑof computer simulations? Up to
this time, CPSR has focused quite
appropriately on the most visible "high end" of the war games world:
the strategic nuclear command and
control system and "Star Wars," the Strategic Defense Initiative.

That work must continue, but there is a great deal of important work
yet to be done at the operational
and tactical levels of what is euphemistically called "conventional
warfare." This is doubly so at this
critical moment of opportunity for arms control: the Conventional
Force Reduction talks in Europe
resume on September 5, and the Strategic Arms Reduction Talks (START)
began again in August.

Two objections immediately arise. First, aside from occasional, brave
"whistleblowers" like David
Parnas, CPSR people without security clearances are shut out by the
veil of secrecy that surrounds
military computer work. Nor do we in any way intend to push good
people with access into unsolvable
moral dilemmas. Second, the military has access to such technically
superior and expensive hardware
and software, that we can't hope to challenge them on their own
grounds.

False on both counts, as Allen has revealed in his book Wargames.
While the objections may hold at the
very highest level within the military-industrial-academic complex,
they do not for the vast majority
of computer simulation work used to justify military plans,
operations, and equipment procurement.

At the highest professional level, Joshua Epstein and William W.
Kaufmann of the Brookings Institution
and Mark Sakitt, then of the Stanford Center for International
Security and Arms Control, have
published several important computer-assisted works criticizing
military strategy and procurement
plans, especially the Maritime Strategy (Epstein, The 1988 Defense
Budget, and The Calculus of
Conventional War, Kaufmann, A Thoroughly Efficient Navy [both
available from Brookings]; Sakitt,
Submarine Warfare in the Arctic: Option or Illusion? Stanford Center
for International Security and
Arms Control.)

These studies have seriously challenged key aspects of the Navy's
strategy and procurement plans. And
never has naval. arms control been more important, since Gorbachev's
top military adviser has
publicly stated that "Reaching final agreement on radical cuts of
armed forces in Europe and making
them defensively oriented would remain in doubt without initiating
talks on naval cuts . . . Sooner or
later these negotiations will have to be conducted. Otherwise the
entire arms control process will
break" [The New York Times, July 25, 1989].

Because of the computer industry's involvement in the simulation
market, the very same simulation
software used by many branches of the military is available on the
open market and can be run on
anything from an IBM-AT compatible on up. An example is CACI
corporation's Simscript II.5 with
Simgraphics. Training courses are available as well. Neither the
hardware nor the software is out of
reach financially.

The most often heard put-down of non-military people concerned about
the arms: ace is that they are
"armchair critics." When it comes to knowing the results of combat
between nuclear-armed
superpowers, however, everyone is an armchair critic. Physicians for
Social Responsibility made an
invaluable contribution by debunking the notion that nuclear war is
"winnable" in any human sense.

Computer Professionals for Social Responsibility have played an
equally valuable role in pointing out
the dangers of relying on computers in the nuclear command and control
system, and the infeasibility of
a "Star Wars" defense.

When it comes to the actual likelihood of a crisis escalating into
nuclear war, and the enormous expense
of so-called "conventional" forces and strategies, especially naval,
we have barely scratched the
surface.

The opportunities are many. The world of commercially available
computer simulation games within
the $60 range has progressed so rapidly that some of them, according
to active or retired military
personnel, rival or surpass what the military themselves are using.

For example, Strategic Simulations Inc.'s game "Atlantic '86" and
Simulations Canada's "Grey Seas,
Grey Skies" both received favorable reviews in Proceedings, as Allen
notes. More recent games have
received excellent reviews by active, reserve, or retired military
officers in such publications as
Computer Gaming World. At the Institute for Global Security Studies,
we have played four naval
simulation games that offer important insights into the dangers of the
Maritime Strategy. While we
decidedly do not recommend these games as light-hearted entertainment
for children, they are
professionally produced and revealing, within all the limits of the
simulation genre.

At the strategic level, Simulation Canada's "Seventh Fleet" claims to
simulate a 1995 crisis in which
China invades Vietnam (again) and a superpower crisis ensues. The game
is text-only, with
accompanying maps and counters.

Two scenarios are offered: "early warning," in which U.S. and Soviet
forces have been alerted and are
already in position, and "standard," in which forces enter the crisis
area at a much slower rate. No
matter which scenario is chosen, three lessons are apparent: 1 ) the
longer a crisis, the more likely
that hostilities will break out, even inadvertently; 2) even U.S.
"major" or "substantial" victories
come only after appalling losses; and, 3) the Soviets begin using
nuclear weapons at sea when they seem
destined to "lose."

Playing the game is a shocking and sobering experience. While the game
is very sparse in
documentation and thus rather "opaque," one suspects that its Canadian
origin injects an element of
objectivity, even though the game is clearly based on the premises of
the U.S. Maritime Strategy as put
forward in journals like Proceedings.

Electronic Arts' Strike Fleet purports to simulate modern tactical
surface warfare with a wide variety
of scenarios raging from the Persian Gulf to WWIII. After a few games
one will have more sympathy for
the people in the Combat Information Center of the Vincennes--and
deeper doubts about the wisdom of
deploying that kind of ship there. Of particular interest is the
honesty displayed in the designers' notes
at the end of the documentation:

Simulating warfare is a tricky business. It's important to remember
the difference between a game and
the real thing. Although it seems obvious to state, what can be a lot
of fun sitting at your computer
changes when shooting and dodging real bullets. Our hope is that
Strike Fleet gives one some
appreciation of the issues of war without the dangers of actually
trying to live through one.

There are no innocent airliners in the game, but it does teach one
thing: the speed, destructiveness and
complexity of modern naval weapons systems make serious mistakes
virtually inevitable. True, after
the tenth time through a game scenario, players have got it down pat.
Real war is not like that. In real
war, every time is the first scenario run-through. This does not
support early, aggressive, forward
deployment of large numbers of sailors on expensive naval ships in a
crisis. Yet that is exactly what we
do now.

The last two games to be mentioned here attempt to simulate what is
probably the most potentially
dangerous and provocative of all naval operations: submarine warfare.
It was submarine warfare that
brought the United States into World War I, and moved U.S. opinion
toward involvement in World War
II.

Under the Ice by Lyric Software is a very complex strategic game which
depicts contemporary anti-
submarine warfare (ASW) in the North Atlantic. The official U.S.
Maritime Strategy indicates that
nuclear attack submarines will play a key role in the very earliest
phases of a crisis. This game
quickly dispels the notion that a major ASW campaign can be
"controlled"Ñdespite the fact that the game
assumes currently unavailable "blue-green" laser communication between
submarines underwater. It
will be "Cry havoc, and let loose the dogs of war."

Electronic Arts' 688 Attack Sub claims to put you in command of a U.S.
or Soviet nuclear attack
submarine in a variety of scenarios ranging from peacetime exercises
to crisis confrontations, to open
warfare. A careful examination of the documentation reveals that the
game was developed with the help
of "the Naval Postgraduate School Computer Club, Monterey, CA."

The game features very advanced graphics and is gripping enough to
cause the jitters. If World War III
ever begins, the opening shots will probably be fired underwater.

The game's documentation is a veritable unclassified manual of ASW,
that would take a lot of digging to
put together on your own. 688 Attack Sub, for all its
oversimplification, teaches one thing: nobody
really knows what will happen in an undersea war, because there has
never been a real oneÑjust
computer simulations and "exercises."

Submarine warfare is particularly isolated, complex and unstable. The
August 6, 1989 issue of The San
Diego Union revealed that the captain and other senior officers of a
U.S. nuclear attack sub had been
accused by fellow officers and crewmembers of nearly colliding with a
Soviet sub while chasing it and
also intruding into Soviet territorial waters inside the twelve mile
limit without authorization.

Naval arms control of both nuclear and "conventional" forces is now a
critical aspect of the arms
control process. At the Institute for Global Security Studies in
Seattle, Washington, we are committed to
exploring and employing computer-assisted or generated simulations of
contemporary naval warfare in
order to examine and critique the existing maritime strategies and
forces of the superpowers. We would
welcome any suggestions, help and cooperation on the part of the
members of CPSR. Contact: Peacenet:
cmeconis or Compuserve: 75410,45, attn to C. Meconis; or write Attn:
C. Meconis, lGSS, 732 18th
Ave. E. Seattle, WA 98112, phone (206) 324-9810, FAX (206) 329-9499

Charles Meconis is the Research Director of the Institute for Global
Security Studies, an independent,
nonprofit research organization in Seattle, Washington.

Congress Considers "High-Performance" Computing Initiative

Marc Rotenberg CPSR Washington Office Director

At the end of the last Congress, Senator Albert Gore (D-TN) introduced
legislation to develop a national
supercomputer network. The bill was reintroduced this Congress. It
calls for the allocation of $1.75
billion over five years to support the development of a broad-band,
fiber optic network to link
government, industry, and university research sites across the
country. It also calls for funding a
national digital library and for research in artificial intelligence,
communications, and
supercomputing.

Senator Gore's bill has now been matched by an Administration plan
that calls for a "high-performance
computing" initiative. This effort is the result of a long study on
computing priorities conducted by the
White House Office of Science and Technology Policy, under the
auspices of its inter-agency Federal
Coordinating Council for Science, Engineering, and Technology, which
is known colloquially as the
"Fixit" committee. This group had released a series of recommendations
for supporting "high
performance" computing in November, 1987. The FCCSET report stated, "A
strong domestic high
performance computer industry is essential for maintaining U.S.
Ieadership in criticial national
security areas and in broad sectors of the civilian economy."

Two hearings have been held on the supercomputer legislation. The
first hearing focused on computer
networks, and featured John Rollwagen (Cray), Sheryl Handler (Thinking
Machines), Craig Fields
(Defense Advanced Research Projects Agency), and William Wulf
(National Science Foundation). The
second hearing considered the software section of the bill,
particularly visualization, artificial
intelligence, and advanced computer software. The witnesses were Jacob
Schwartz (DARPA), William
Wulf (NSF), James Clark (Silicon Graphics), David Nagel (Apple), J.
William Poduska (Stellar), Raj
Reddy (Carnegie-Mellon University), and Karl-Heinz Winkler (University
of Illinois).

A third hearing examined the plans for a national digital library. At
the hearing were Robert Kahn of
the Corporation for National Research Initiatives, Ted Nelson of the
Xanadu Project, and
representatives from the U.S. Geologic Survey, the National Library of
Medicine, the Library of
Congress, IBM, Xerox, and UMV Microsystems.

CPSR has been concerned about the military dominance of federal
computer funding for a long time. As
an independent organization of computer professionals, CPSR should
develop "a program in which the
critique of military projects is directed to issues of conversion and
reallocation of priorities, not just
the immediate dangers." The supercomputer initiative is likely to be
the primary forum for policy
discussions about funding priorities in computer science for several
years to come. For these reasons,
CPSR should play an active role in the development of the legislation,
steering it toward those projects
and goals that will have the most beneficial social impact.

The supercomputer initiative faces several obstacles, the primary one
is the tight budget. With
Congress looking to tighten it's belt, all new funding for new
programs is coming under close scrutiny.
However, other big science projects, such as the superconductor-
supercollider are moving forward,
and with both Congressional and Administrative backing for the
proposal, some legislation seems likely.

Senator Gore has requested comments from CPSR on the impact of the
proposed legislation. CPSR
members are welcome to submit comments to the Washington Office that
could be included in the
statement from the organizations.

CPSR has scheduled a panel of finding priorities in computer science
for the October board meeting in
Washington, DC. The panel discussion should provide an opportunity for
CPSR members to learn more
about the impact of the supercomputer initiative on the computer
science community.

My Game--Your Simulation? Richard Healey

Materials reviewed in this article:

"Analysis or Propaganda? Measuring American Strategic Nuclear
Capability, 1969-1988," Michael
Salman, Kevin J. Sullivan, and Stephen Van Evera; in Nuclear
Arguments, (eds.) Lynn Eden and Steven
E. Miller, Cornell University Press, 1989.

"SIOP, A Computerized Nuclear Exchange Model for Civilian Defense
Analysts," Kevin J. Sullivan. (runs
on IBM and MS-DOS compatibles).

"Balance of Power, the 1990 Edition," Chris Crawford, Mindscape Inc.,
1988. (Runs on IBM and MS-
DOS compatibles and the Macintosh).

"Hidden Agenda," Jim Gasperini, Springboard Software, Inc., 1989.
(Runs on the Macintosh).

I was the Director of the Coalition for a New Foreign and Military
Policy from early 1982 through
1985. The Coalition, since deceased, lobbied Congress and worked with
grassroots groups on foreign
policy issues. We put enormous energy into our program on disarmament,
which

centered on the "Campaign Against First-Strike Nuclear Weapons." We
were the only national
organization at the time that was concerned with the Trident D-5
missile, though we were also
interested in the MX missile and the so-called Euromissiles, the
Pershing 2s, the Lance missile, and
others.

I remember too well following Howard Morland, who headed the campaign
for us, from one
congressional office to the next, talking with "defense aides" about
the problems with first-strike
capable weapons and with the D-5. A large number of the aides didn't
know what we meant by first-
strike weapons. When we explained, their reaction was often that the
United States obviously needed
such weapons, since the Soviets had or would soon develop comparable
weapons; sometimes their
reaction was great enthusiasm for the idea of a pre-emptive first-
strike, and they wanted to know how
many Trident D-5's were needed to wipe the Soviets out. What Howard
and I needed was to leave them a
copy of "Analysis or Propaganda" and the accompanying computer
program, SIOP, and return in a week
or two and try again. (SIOP is short for "Sullivan's Instructional
Operations Program," and a pun on
the Defense Department's plan for nuclear war, the Single Integrated
Operational Plan.)

Salman, Sullivan, and Van Evera (or S2V for short) have performed a
significant service to analysts
and activists with their article and program. The heart of their
argument is that "military strength
should be assessed by measuring the capacity of forces to execute
strategy," which requires that first a
strategic mission is articulated. The article maintains that most
arguments about the strategic nuclear
balance between the United States and the Soviet Union are based on
fallacious methodologies; they quote
approvingly defense analyst William Kaufmann's comment that the
prevalent static comparisons are
"virtually worthless as a basis for evaluating or designing U.S.
forces."

Most of us interested in the nuclear balance have employed one or more
of these other methodologies, so
it is worth spending a moment to note "the games analysts play." The
first is "static indicator
comparison games," which come in two varieties: number games, or "bean
counting," and trend games,
where the analyst compares trends as opposed to cross-sectional data.
The second category is "flawed
dynamic analysis," in which unreliable or bizarre data are used, or by
playing "operational games," by
assuming that "national leaders operate their nuclear forces in
improbable ways." There are "strategy
games," in which the missions, against which forces are assessed, are
distorted to make the forces
appear strong or weak. The authors provide stinging examples and
criticisms of these games, as well as
a further set of games, including equivalence games, perception games,
nonstrategy games, and one of
my favorites, ex cathedra games, in which analysis and explanation is
replaced by reliance on
authority.

I wholeheartedly agree with S2V's criticisms and, with one
reservation, with their standard for
assessing the strategic nuclear balance. My reservation is that their
argument and methodology
implicitly assumes a traditional arms control contextÑthat is, that we
need to have and can threaten to
use nuclear weapons to deter the Soviets. A few heterodox analysts,
such as Morton Halperin, and many
activists, would reject that assumption. But for those who want to be
part of the mainstream discussion
about nuclear weapons, including lobbying Congress, this is the
dominant assumption. In addition, I
want to add a word in defence of those who haven't been up to the
authors' standards. Their methodology
is extremely demanding, requiring data and computational abilities
beyond the reach of many non-
governmental agencies; it requires what we might call an exemplary
simulation of a nuclear strike.
Their criticisms, while stated with unusual clarity and bite, are not
especially original, as they
acknowledge. What is original and most important is that they have
developed and provided us with a
tool so that almost anyone interested in these questions can come up
with a rigorous assessment of the
current or any alternative strategic nuclear balance.

For example, from my days at the Coalition I continue to be interested
in first-strike weapons and in
particular the Trident D-5 missile. With SIOP, the simulation, I was
able to ask a series of "what-if"
questions that I could not have answered previously in any sensible
way: what difference would it make
if we kept the older C-4 missiles, not replacing them with the D-5?
What is the outcome of various
scenarios with 500 D-5s, or 600, or 1,000? How many missiles have to
be added to make a qualitative
difference in the outcome? I won't tell you my conclusionsÑget this
program and try it for yourself!

SIOP has several components: the main program, five data sets, a file
giving sources and documentation
for the data, and an editing program for modifying the data sets or
creating new ones. The data sets
describe Soviet and U.S. strategic nuclear forces in 1983, in 1987,
and as they would exist under
various treaty restrictions. Running SIOP (under MS-DOS only) is
straightforward. You choose a data
set; then the program takes you through a series of questions about
the war to be waged and choices
about related assumptions about force capabilities. At the end

of the simulation, SIOP calculates what has been destroyed and the
remaining forces that are still
reliable and deliverable. The latter are used to calculate each side's
residual ability to destroy
population, industry, and remaining military forces, measured against
three strategic alternatives:
assumed destruction, first-strike counterforce, and second-strike
counterforce.

SIOP is a "very serious" simulation, in that it is intended to have
policy implications for the real
world, as opposed to being only fun, educational, or worst of all,
advertising. Thus it is especially
important to ask about its "validity" as a model. Severo Ornstein's
first criterion for validity, that we
test the simulation `'by comparing its results and conclusions with
behavior of the actual system it
purports to model," is of course not possible. (see Ornstein, p. 1 in
this Newsletter). I think Ornstein
implicitly suggests a second method for testing validity when the
first is not possible, which is to check
every element of the simulation as closely as possible, thinking about
the consequences if various
assumptions or operational elements fail to hold; I hope Ornstein will
say more about this, as it poses
hard problems.

The authors of SIOP are to be commended for making the task of
checking the validity of the simulation
as easy as possible. They have provided us with the sources of their
data, made assumptions as explicit
as possible, and given the user choices about operational decisions
when appropriate. They highlight
possible controversial assumptions, places where they found little or
inadequate data, and the
limitations of their model. I hope that the Defense Department and
other professional analysts examine
the model, challenge its assumptions, rules of operation, and data,
and then provide us with improved
models.

My own conclusion is that the authors have created a robust model of
strategic war-gaming scenarios.
We should be clear, as the authors are, that it is those scenarios
that are what is under discussion. In
that context, the simulation provides a rich set of conclusions that
deserve to be taken extremely
seriously. I will quote only their summary statement:

The strength of American nuclear forces depends on the missions they
are measured against. American
second-strike counter-value capability is enormous, American first-
strike and second-strike
counterforce capabilities are minimal, and American forces cannot wage
controlled warfare for
coercive or bargaining purposes unless both sides limit their attacks
severely, perhaps not even then.
Soviet capabilities are similar, so the nuclear "balance" is
essentially "even."

Hidden Agenda and Balance of Power: "Serious Games"

Hidden Agenda and Balance of Power are games; to use Clark Abt's
helpful oxymoron, they are "serious
games:" they have explicit and carefully thought out educational
purposes. SIOP can of course be used
for education; it is already being used in courses on nuclear
strategy. But it is designed to produce real-
world policy analysis and recommendations; it isn't a gameÑand it
isn't fun. It might make more sense,
however, to think of simulations as falling on a continuum from vague
analogies to tightly defined
models with immediate operational implications.

Hidden Agenda focuses on a composite country, Chimerica, whose creator
describes it as composed of the
"body of El Salvador, neck of Nicaragua, claws of Cuba, head of Haiti
. . .that together form a beast
representative of all." It is an interactive game, in which the player
becomes the new president after
the dictator is overthrown.

And so here I am, Presidente Healey, dealing with a bitterly divided
cabinet and army, facing an
irritated U.S. ambassador and a smooth-talking Soviet envoy. I fire
one minister, I accept the advice of
a landless peasant, I am rude to the Church, I make a deal with a
corrupt general, and time passes.
Crises occur. I try to placate one side, then the other. In my first
attempt as president, I didn't last two
years, dying in a bloody battle in the National Palace. The "Verdict
of History," an excerpt from a future
encyclopedia, which can be printed out at the end of each game,
analyzed several weaknesses and flaws
in my presidency, and I studied it carefully with plans for
improvements in my next turn in power.

The new version of Balance of Power (BOP), the "1990 Edition," has
added more complexity and a new
level to the game, moving it further along the continuum away from
Hidden Agenda and toward SIOP.
BOP isn't as personal as Hidden Agenda, though I still felt dismayed
at the outcome in my first attempts
at the gameÑI managed the international rivalry between the United
States and the Soviet Union into
total nuclear war within one or two time periods.

For those (few?) who have not tried any version of BOP, here is a
quick summary. It is a simulation of
"global power politics," in a world of 80 countries divided into two
spheres of influence. The game
focuses on the conflict between the two superpowers; the goal is to
gain prestige (power) at the expense
of the other power by means of

aiding insurgencies, overthrowing governments, or simply forcing the
opposing side to back down. It
has four levels of play, from Beginner to Multipolar, in which the
other 78 countries can react in
simple ways to world events independently of the actions of the United
States or the Soviet Union.

Both Gasperini and Crawford have obviously thought long and hard about
the nature of simulations and
their goals for these simulations. Both have comments in their manuals
about the fit between the games
and "reality," and they express the hope that having played these
games enough, we will question our
own assumptions about the world, our own understandings, and be
challenged to learn more and think
more about the role of the United States in the world.

Each program has different strengths as a simulation. Hidden Agenda is
the greater departure from
realism, and it is a simpler game overall; it can be played without
studying the manual. By casting the
user into a persona in a "typical" country in Central America, and by
making it relatively simple to see
the historic dilemmas these countries face, Hidden Agenda makes us
think about those countries and the
United States in a new way. And we can have fun doing so.

Balance of Power is complicatedÑI never felt confident of what I was
doing beyond the beginner level. So
many countries to worry about, so hard to figure out why the Soviets
would back down in one instance
and not in another.... Of course, all that and more holds in the real
world. A user could get lost in the
details and the gaming itself in BOP, learn a great deal about the
world, have fun and frustrationÑbut he
or she might not reach the level of thinking about assumptions and
about our role in the world quite so
readily.

Both games are well-suited for classroom education, where small groups
could discuss their choices
and the outcome. And I suspect a creative local activist interested in
foreign policy could use these
games as a way of involving and educating new people about foreign
policy. And even if they didn't work
as recruiting tools, everyone would have a good time, which is more
than I can say for much of what we
do.

References

Clark Abt, Serious Games, Viking Press, 1970 John R. Rasor, Simulation
and Society, Allyn and Bacon,
1969

Richard Healey is executive director of Nuclear Times magazine, and
resides in Washington, D. C.


Update on the SDI's National Test Facility

Bill SulzmanÑCPSR/Denver-Boulder

In the spring of 1988, a CPSR study team realeased a report highly
critical of the Strategic Defense
Initiative's National Test Bed, or NTB. The NTB is a proposed $1
billion network of computer resources
for simulating, designing, testing, and perhaps eventually controlling
a space-based ballistic missile
defense. The center of the nationwide network will be the National
Test Facility (NTF), located at Falcon
Air Force Base in Colorado Springs, Colorado. [For a full report on
the structure and mission of the
SDI's National Test Bed, see The CPSR Newsletter, Vol. 6, No. 2,
Spring 1988Ñed.]

The National Test Facility is now operating with approximately 400
personnel from the Martin
Marietta Corporation and 60 others from the Army, Air Force, and the
MITRE Corporation. The
projected date for transfer from their temporary quarters to their new
facility is May 1, 1990. There
has been a 23% cost overrun in the construction of the new structure,
amounting to about $23 million.

According to Colonel Thomas Leib, National Test Bed project director,
things are running smoothly. A
Cray 2, an IBM 3090, and two Elexi computers are already in use.
Eventually the facility will be
receiving a second Cray 2, or a Cray 3 if and when it becomes
available. The NTF has T-1 links to the
SDIO headquarters in Washington, D.C., to General Electric in Blue
Bell, Pennsylvania, and to the Army
Strategic Defense Command in Huntsville, Alabama. There is also a 56
kbs link to Los Alamos National
Weapons Laboratory in New Mexico. The rest of the National Test Bed is
not yet on-line.

Colonel Leib believes the NTF has already completed a number of
accomplishments. Managers there have
conducted two "successful" nuclear war games with various SDI
components, one lasting four days in
November 1988 and another lasting two days in June of this year. The
facility has set up a "data threat
center." They have completed several weapons tradeoff analyses. A
simulation of a Phase 1 SDI system,
including "Brilliant Pebbles," has been completed. There has been one
experiment with the NTF serving
as a command and control center. A software research center has been
started. The war gaming
conducted at the NTF has also started to include real-world data from
Delta rocket and satellite
launches.

Test Facility personnel insist that there have been no applications of
their findings to offensive war
capabilities, but they admit that all their data is being shared with
the Army, Air Force, and various
weapons contractors who develop such applications. There is no longer
any attempt to represent the SDI
as anything other than a significant component of U.S. nuclear war-
fighting capability.

I asked Colonel Leib, "Since you cannot check your simulations against
real-world data derived from a
nuclear war, and since you will be unable to check a completed system
for errors before you have to use
it, isn't the best you can hope for a system which will almost work
and which therefore will be a
colossal failure?" He responded, "I agree with everything but your
conclusion." He then went on to
make the flawed comparison of his current task with that of the
planners of the Apollo manned space
program.

Numerous inquiries of congressional staffers and personnel from the
General Accounting Office reveal
an oversight process greatly overmatched by SDIO's many supporters.
Most government overseers are
scrambling to come up with the right questions to ask. GAO studies in
1987 and 1988 and a major
review of the SDI by the Congressional Office of Technology Assessment
in 1988 were critical of the
SDI in general and of the National Test Bed in particular, but the
money pipeline remains open.

Step by step the ABM Treaty and the arms control process are being
undermined and the groundwork for
an arms race in space is being developed. The normally valuable tool
of computer simulation is being
grossly misapplied in an effort that will make the world an even more
dangerous place than it is now.

Bill Sulzman is executive director of Citizens for Peace in Space, an
organization based in Colorado
Springs, Colorado.

CPSR Chapters September 1989

CPSR/Austin

Ivan M. Milman
4810 Placid Place
Austin, TX 78713
(512) 823-1588 (work)

CPSR/Berkeley

Steve Adams
3026 Shattuck, Apt. C
Berkeley, CA 94705
(415) 845-3540 (home)

CPSR/Boston

Tom Thornton
2 Newland Road
Arlington, MA 02174
(617) 643-7102 (home)

CPSR/Chicago

Don Goldhamer
528 S. Humphrey
Oak Park, IL 60304
(312) 702-7166 (work)

CPSR/Denver-Boulder

Randy Bloomfield
4222 Corriente Place
Boulder, CO 80301
(303) 938-8031 (home)

CPSR/Los Angeles

Rodney Hoffman
845 S. Windsor Blvd., #6
Los Angeles, CA 90005
(213) 932-1913 (home)

CPSR/Madison

Debby Servi
128 S. Hancock St., #2
Madison, WI 53703
(608) 257-9253 (home)

CPSR/Maine

Betty Van Wyck
Adams Street
Peaks Island, ME 04108
(207) 766-2959 (home)

CPSR/Milwaukee

Sean Samis
6719 W. Moltke St.
Milwaukee, WI 53210
(414) 963-2132 (home)

CPSR/Minnesota

David J. Pogoff
6512 Belmore Lane
Edina, MN 55343-2062
(612) 933-6431 (home)

CPSR/New Haven

Larry Wright
1 Brook Hill Road
Hamden, CT 06514
(203) 248-7664 (home)

CPSR/New York

Michael Merritt
294 McMane Avenue
Berkeley Heights, NJ 07922
(201) 582-5334 (work)
(201) 464-8870 (home)

CPSR/Palo Alto

Clifford Johnson
Polya Hall
Stanford University
Stanford, CA 94305
(415) 723-0167 (work)

CPSR/Philadelphia

Lou Paul
314 N. 37th Street
Philadelphia, PA 19104
(215) 898-1592 (work)

CPSR/Pittsburgh

Benjamin Pierce
School of Computer Science
Carnegie Mellon University
Pittsburgh, PA 15213
(412) 268-3062 (work)
(412) 361-3155 (home)

CPSR/Portland

Bob Wilcox
CPSR/Portland
P.O. Box 4332
Portland, OR 97208-4332
(503) 246-1540 (home)

CPSR/San Diego

John Michael McInerny
4053 Tennyson Street
San Diego, CA 92107
(619) 534-1783 (work)
(619) 224-7441 (home)

CPSR/Santa Cruz

Alan Schlenger
419 Rigg Street
Santa Cruz, CA 95060
(408) 425-1305 (home)

CPSR/Seattle

Doug Schuler

CPSR/Seattle
P.O. Box 85481
Seattle, WA 98105
(206) 865-3226 (work)

CPSR/Washington, D.C.

David Girard 2720
Wisconsin Ave., N.W., Apt. 201
Washington, D.C. 20007
(202) 967-6220 (home)

CPSR, P.O. Box 717, Palo Alto, CA 94301 (415) 322-3778

Around the Chapters Summer 1989

Given that summer is often a slow time for CPSR chapters, coupled with
the fact that the last newsletter
came out only six weeks ago, I thought that I would have trouble
pulling this section together for this
issue. What surprised me is not that there are some chapters that
weren't able to report any news, but
that there were several chapters that managed to keep quite busy in
the last two months. For future
issues, we should have more lead time, and I hope to be able to report
from some of the chapters that
were silent this time.

ÑEric Roberts CPSR National Secretary

Northwest Region

The "Computers for Peace" project in Seattle continues to grow, and
the chapter just conducted a one-
day workshop on "Computers for Non- Profits." CPSR/ Seattle has also
taken on, as a chapter project,
local arrangements for the CPSR Workplace Project Conference on
Participatory Design next April
(see the call for participation on page 19). CPSR/Portland sponsored a
talk by Congressional aide Bob
Sherman who spoke on "Technology and Security Issues in Arms
Control"Ña talk which CPSR/Portland
Secretary Nicholas Horton recommends highly to other CPSR chapters,
even though he "disagrees
somewhat with his conclusions." The Portland chapter also sponsored a
talk by Len Tower of the Free
Software Foundation (who also made it to CPSR/Seattle) opposing "look
and feel" software copyrights.
The chapter has also paid attention to outreach, both to its own
members and the larger community,
with a barbeque and tables at the Oregon Peace Fair, respectively.

Western Region

For its summer meetings, CPSR/Berkeley showed the Reliability and Risk
slide show and hosted a talk
by Deborah Brecher, founder of the Women's Computer Literacy Program,
on "Men, Women, and
Computers: Do Women Think Differently?" (Deborah gave a similar talk
at last year's annual meeting).
CPSR/Berkeley has also organized a working group on computers in
entertainment, leisure, and the
arts, which will examine the effect of computers on the ways in which
people express and entertain
themselves. Much of the activity in CPSR/Palo Alto is centered in the
working groups. The Workplace
Project continues to work

toward the conference on participatory design, which will be held in
Seattle next year. In the last
month, there has also been considerable activity directed towards
seeing whether CPSR can offer any
technical assistance to East Palo AltoÑa low-income community just
outside of Silicon Valley affluence
that is struggling to keep its administrative operations running. The
last meeting of CPSR/Los Angeles
consisted of the videotape "Losing Control," a dramatic presentation
of the dangers of accidental nuclear
war. The Los Angeles chapter recommends this highly as a chapter
meeting topic, and the tape is
available from the CPSR National Office.

Midwestern Region

CPSR/Madison has continued its series of presentations on computers in
the workplace with a
presentation by Allen Highman, labor representative at the University
of Wisconsin for the Wisconsin
State Employees Union. The talk focused on the issue of
"telecommuting" in which workers conduct their
work at home via phone line links to central computers. According to
Highman, the workers in this
experimental project are the first and only instance of home-based
employees who are represented by a
labor union.

New England Region

For its summer meeting, CPSR/Boston held a discussion session on the
presentation of computers by
the media. As a starting point for discussion, the group listened to a
report by Michael Skoler of
National Public Radio on the Internet virus and watched the videotape
"Computers in Your Life," which
is available from ACM Press. CPSR/Boston also staffed a booth at the
National Educational Computing
Conference in June, getting CPSR's message out to computer educators
from across the country.

Mid-Atlantic Region

CPSR/Pittsburgh reports that its study group is rolling right along,
currently looking at computers
and privacy and planning units for the fall on literacy, whistle-
blowing, and gender issues in the
computing profession. CPSR/DC will soon play host to the 1989 Annual
Meeting, and we hope that as
many CPSR members as possible can come and learn first-hand about the
activities of the DC chapter.

The 1989 CPSR Annual Meeting October 20-21, 1989 Washington, D.C.

Friday, October 20

Pan American Health Organization Building, 525 23rd Street, N.W.

9:00 a.m. to 10:00 a.m. Keynote speech

"Computers, Privacy, and Civil Liberties in the United States"

Senator Patrick Leahy (D-VT), chairman, Senate Judiciary Subcommittee
on Technology and the Law

10:00 a.m. to 10:30 a.m. Break 10:30 a.m. to 11:45 a.m. Panel
discussion

"Federal Support of Computer Research and Development and the Nation's
Technological Base~

Frederick Weingarten, Congressional Office of Technology Assessment

Kenneth Flamm, Brookings Institution

William Scherlis, Defense Advanced Research Projects Agency

Marilyn Elrod, House Armed Services Subcommittee on Research and
Development

Ann Markusen, professor of urban planning and development, Rutgers
University

Moderator: Lance Hoffman, professor of computer science, The George
Washington University

Friday, October 20

Noon to 1:45 p.m., Luncheon, University Club, The George Washington
University, Marvin Center, 800
21st St., N.W.

Luncheon address. "Computer Monitoring: A Threat to the Right to
Privacy?"

Karen Nussbaum, founder and executive director, National Association
of Working Women, or 9to5

Presentation of the Norbert Wiener Award for Professional and Social
Responsibility to Professor
Daniel D. McCracken, Department of Computer Science, City University
of New York

Friday, October 20

Return to the Pan American Health Organization Building, 525 23rd
Street, N.W.

2:00 p.m. to 3:30 p.m. Panel discussion

"Computers in Education: Mixed Agendas and Uncertain Outcomes"

Carol Edwards, executive director, Southern Coalition for Educational
Equity and Project MICRO

Linda Roberts, Congressional Office of Technology Assessment

Sherry Turkle associate professor of sociology, Massachusetts
Institute of Technology

Chet Bowers, professor of education, University of Oregon

Moderator: Terry Winograd, associate professor of computer science,
Stanford University, and
president of CPSR

3:30 p.m. to 4:00 p.m. Break 4:00 p.m. to 5:30 p.m. Panel discussion

"Patrolling the Programmers: Computer Ethics and Computer
Accountability"

Bryan Pfaffenberger, assistant profesor of humanities, University of
Virginia

John Shore, vice president, Entropic Processing, Inc.

Carol Gould, professor of philosophy, Stevens Institute of Technology

Moderator: Rachelle Hollander, coordinator, Ethics and Values Studies
Program, National Science
Foundation

Saturday, October 21 Marvin Center, 800 21st St., N.W., Fourth Floor
8:30 a.m. to 9:45 a.m. CPSR
reports

Reports will be heard from the national staff of CPSR and from the
regional representatives of the
Board of Directors. The reports will cover the financial status of the
organization, national programs,
and chapter activities.

9:45 a.m. to 10:15 a.m. Break 10:15 a.m. to 12:15 p.m. Workshop

"Local Organizing for Professionals"

Monica Green, National Field Director, SANE/ Freeze

12: 15 p.m. to 2:00 p.m. Lunch (not provided) 2:00 p.m. to 4:00 p.m.
Issue workshops

Participants will be able to break up into small groups to discuss
specific issues within the CPSR
program, such as computers and international security, computers and
privacy, computers and
education, computers and voting, and other subjects.

4:00 p.m to 5:00 p.m. Plenary closing

This session will feature reports from the workshops, and a closing
address by CPSR President Terry
Winograd.

Two-day registration received before October 9:

CPSR members $55

Non-members $75

One-year membership plus registration $105

Two-day registration received after October 9:

CPSR members $85
Non-members $95
One-year membership
plus registration $105
Students/Low Income (lunch not included) $25


From the Secretary's Desk Eric RobertsÑNational Secretary

I suppose that one of the best pieces of news that I can report for
the summer newsletter is that there is
a summer newsletter. Two months ago, in the midst of a serious
financial crisis, we faced the
possibility of significant cutbacks and weren't entirely sure we'd be
able to mail this issue out.
Fortunately, there is good news to report. We've now completely
matched the $15,000 Stern challenge
grant with individual contributions; on top of this, we've received an
additional grant of $20,000 (plus
$10,000 more in matching funds) from the Scherman Foundation and
another $15,000 from the
Baumann Foundation. We've also gotten a good response to our appeal in
the Spring Newsletter, having
raised about $10,000 from generous members as this issue goes to
press.

Thanks in large part to your support, the immediate crisis seems to
have eased. Even so, we need to
continue building for the future. The truth is that we won't be able
to rely as heavily on foundations for
support as we have in the past. Foundations get tired of providing
maintenance support. They want to see
evidence that we are important to our members. The best evidence for
this comes when those members
contribute significantly to the support the organization. We hope
that, when renewal time comes
around, that you will be able to give CPSR a little extra or renew at
a higher level of support. By doing
so, you will be helping us ensure that these times of financial crisis
are indeed in the past.

Those of you who are ACM members should take a look at the August
issue of Communications. This issue
is filled with CPSR material, including an overview of computers and
social responsibility by Doug
Schuler and Jon Jacky, four papers from the last two Directions and
Implications of Advanced
Computing conferences, and a sidebar describing CPSR. Communications
of the ACM has already printed
four other articles that originally appeared in The CPSR Newsletter
and will soon reprint Gary
Chapman's article on the Soviet Union. This is quite a turnaround from
the early days of the Newsletter
when Gary worried that our newsletter might have to rely on ACM
publications for material.

In addition to the material in the Communications, we ourselves have a
new publication. Northwestern
Regional Director and Publications Committee Chair Doug Schuler has
produced and distributed issue
#0 of the CPSR Activists' Newsletter. The Activists' Newsletter will
be a quarterly forum intended to
promote communication among the chapters, which, in Doug's words,
"will help us to exchange 'lessons
learned' on what works and what doesn't" and will encourage "cross-
fertilization on ideas, projects, and
information." If you would like to be on the mailing list, either for
electronic or U.S. Mail distribution,
please contact:

Doug Schuler 2202 N. 41st Street Seattle, WA 98103 Electronic mail
address (Internet): cpsr-
activist-newsletter@csli.stanford.edu

In organizational news, the CPSR Executive Committee has a new member,
with Jeff Johnson replacing
Lucy Suchman, who will continue to serve on the Board of Directors in
her capacity as Western
Regional Director. The Executive Committee meets monthly to oversee
CPSR's activities between the
three-times-a-year meetings of the full Board.

The 1989 Annual Meeting in Washington D.C. is quickly approaching, and
the program (see pages 14-
15) looks extremely exciting. Having this year's Annual Meeting in
Washington, D.C., will focus
attention on the organization and our new presence in the capital, and
we need you to help us show that
computer professionals do care about issues of technology and
societyÑissues for which many of the key
decisions will be made right there in Washington. We would like to see
every chapter represented at
this year's Annual Meeting, as in every year. Last year, CPSR/ Madison
used part of its treasury to
help support travel expenses so that more of its active members could
attend the meeting in Palo Alto.
Other chapters might be interested in following this example. In any
case, I look forward to seeing many
of you there.

CPSR Privacy Notice

The CPSR membership database is never sold, rented, loaned, traded, or
used for any purpose other
than official CPSR business. The information given to CPSR by members
is never released to anyone
except authorized CPSR leaders and staff. The organization may choose
to inform the membership of
information that comes from another group, but the mailing will always
originate with CPSR.

For CPSR members interested in recruiting friends and colleagues to
the organization, this can
sometimes make a difference, so be sure to mention it.


CPSR Presents Norbert Wiener Award to McCracken

The 1989 Norbert Wiener Award for Social and Professional
Responsibility has been awarded to
Professor Daniel D. McCracken of the City College of New York. The
award is given by the CPSR Board of
Directors to a member of the computer profession, or someone dealing
with issues in computing, who
has demonstrated exemplary career standards of social responsibility
and professionalism, and who has
undertaken sacrifices for the prestige of the profession and the
safety of the public.

"The name Dan McCracken is a household word wherever there are
professional or otherwise dedicated
computer programmers," says Norbert Wiener Award recipient Professor
Joseph Weizenbaum of MIT.
Professor McCracken is best known for his many computer science
textbooks, which are considered
classics in the field. He has written or co-authored 25 titles on
computer programming, and his books
have been translated into Spanish, French, German, Italian, Portugese,
Japanese, Polish, Russian,
Dutch, Hungarian, Rumanian, Bulgarian, and Chinese.

Professor McCracken is also well known for his professional activities
in the computer science field.
He has served as president of the Association of Computing Machinery
(ACM), and vice president. He is
a former chairman of the History of Computing Committee of the
American Federation of Information
Processing Societies, a former chairman of the ACM Committee on
Computers and Public Policy, and a
four-time ACM National Lecturer. He was elected a Fellow of the
American Association for the
Advancement of Science in 1985.

Professor McCracken has taken an active role in the development of
public policy involving computer
technology. In 1969 he formed Computer Professionals Against the ABM,
an organization opposed to the
development and deployment of anti-ballistic missile systems in the
United States. He was chairman of
this organization until 1972, when the ABM Treaty was signed which
banned such systems. He spent a
year publicizing the dangers of unreliable software that would be used
in ABM systems, and the
strategic implications of using computers in such a fashion. In this
effort Professor McCracken and his
colleagues were important forerunners of CPSR, which has made similar
arguments about the Strategic
Defense Initiative, or "Star Wars" program. Professor McCracken has
been a CPSR member for many
years as well.

Professor McCracken also anticipated CPSR's concern with privacy and
computers and its work on the
FBI's National Crime Information Center, or NCIC. He testified on the
social impact of the NCIC before
the House Judiciary Committee's Subcommittee on Civil and
Constitutional Rights on September
29,1977. In 1976, he testified before a congressional committee on the
social implications of
electronic funds transfer, or EFT. He was also instrumental in
Congress' enactment of copyright
legislation for software.

Professor McCracken graduated with a bachelor of arts in mathematics
and chemistry from Central
Washington University in 1951. He pursued graduate studies in
mathematics at New York University in
the late 1950s, and worked for General Electric. In 1970 he earned a
Master of Divinity from the
Union Theological Seminary in New York. He currently teaches computer
science at the City College of
the City University of New York.

The Norbert Wiener Award for Professional and Social Responsibility
was named for famed MIT
professor Norbert Wiener, one of the founders of the fields of
cybernetics and information theory.
Wiener was an outspoken proponent of social responsibility in science,
and he worked for decades to
promote the peaceful use of scientific achievement.

The Norbert Wiener Award will be presented to Professor McCracken at a
luncheon to be held in
conjunction with the 1989 CPSR Annual Meeting on Friday, October 20,
at noon, at the University Club
of The George Washington University in Washington, D.C., in the Marvin
Center, 800 21st St., N.W.


Information AccessÑThe Office of Management and Budget, the perennial
bad guys on information
issues, surprised most everyone this summer with a revised ruling on
federal agency dissemination
policies. At issue is whether the government's traditional commitment
to disseminate public
information will be one of the costs of the information revolution.
Moving away from "maximum
feasible reliance" on the private sector, OMB said federal agencies
should distribute government
information in "electronic formats." Now the battle shifts to Congress
and legislation that should clarify
the obligation of the federal agencies to publish computer-based
information.

Morris IndictmentÑMore than six months after the Cornell virus swept
through the Internet, the
Department of Justice has brought an indictment under the 1986
Computer Fraud and Abuse Act. So the
legal case against Robert T. Morris, Jr., begins. One key
consideration in the trial strategy may be the
little discussed overhaul of federal sentencing procedures. Because of
recent changes, Mr. Morris, if
found guilty, will go directly to jail, with no suspended sentence or
probation.

Meanwhile the General Accounting Office released a report saying there
are security flaws in the
Internet. The GAO recommends a coordinating body in the White House
Office of Science and Technology
Policy. But centralizing this authority may lead to the creation of a
security czar and is not likely to
find favor among Internet managers. Still, in the wake of the virus
incident, Internet managers are
likely to post patches more quickly and monitor use more closely.

Consumer Privacy--Interest in Washington in consumer privacy is
building with news that local
supermarkets are planning an electronic coupon system that will create
detailed profiles on consumer
buying patterns. Meanwhile, Equifax announced a "resume verification
service"Ñsubmit a name and
they will do a complete check through their databases for educational,
criminal, government, and
financial information. Privacy advocates aren't pleased about this
computerized dossier serviceÑnor
will be the folks who are denied jobs after this automated background
check.

Military Money- Sorting through the wreckage of the defense
appropriations process, the only thing
that is clear is that very little is clear. There are now three plans
for funding the militaryÑthose of the
President, the House, and the Senate. SDI funding is likely to end up
around the $4 billion, as various
Stars Wars-for-Drug Wars funding transfers fell by the wayside. A
little noticed provision in the DoD
budget will provide DARPA with $20 million for research on "container
monitoring." No doubt the war
on drugs must be fought, but this provision will put DARPA into the
surveillance business.

FBI and AI--Speaking of surveillance, according to recent testimony
from Al Bayse, the head of the
FBI's Technical Services Division, the Bureau is now looking to
artificial intelligence to upgrade law
enforcement activities. But what happens to procedural safeguards and
public accountability when an
expert system tells police to arrest the wrong guy? Both CPSR/Boston
and CPSR/Palo Alto are looking
into this.

Drugs On-Line?ÑThe CPSR/Palo Alto chapter completed a review of the
Health Care Financing
Administration's plan to develop a "claim-capture" system. The system
would provide access to a
national database containing records of prescription drugs currently
being taken by Medicare
beneficiaries. According to the CPSR assessment, key questions of data
accuracy, user accountability,
privacy, system security, and reliability need more attention.

Voting standards under review--The Federal Election Commission
recently issued proposed standards to
improve the integrity of computerized vote tabulation systems. The
standards, minimal at best, will be
voluntary and are unlikely to resolve many of the criticisms voiced by
election officials and those who
have looked at these systems during the last few years. Without
adequate public oversight and the
opportunity for examining source code, voting systems might become the
ultimate "black box." Both the
CPSR/Portland and CPSR/Washington, D.C., chapters are working with
election officials to improve
certification procedures.

ÑMarc Rotenberg

Miscellaneous


The CPSR Newsletter is published quarterly by:

Computer Professionals for Social Responsibility
P.O. Box 717
Palo Alto, CA 94301
(415) 322-3778

Also located at:

1025 Connecticut Ave., N.W., #1015
Washington, D.C. 20036
(202) 775-1588

The purpose of the Newsletter is to keep members informed of thought
and activity in CPSR. We
welcome comments on the content and format of our publication. Most
especially, we welcome
contributions from our members. Deadline for submission to the next
issue is October 31, 1989.

This Newsletter was produced on an Apple Macintosh II, using the
desktop publishing program
Pagemaker 3.0 and Adobe Illustrator 88. The hardware and software used
were donated to CPSR by
Apple Computer, the Aldus Corporation, and Adobe Systems. The
Newsletter was typeset from a
Pagemaker file on a Linotronic 100.


Correction

In the last issue of The CPSR Newsletter (Spring 1989), on page 13,
there was a photograph of a Soviet
personal computer on sale at the Elektronika store in Moscow. The
photograph's caption said the
computer cost 35,280 rubles. In fact, this price was for the computer
shown in the picture plus
twelve student workstations, a standard configuration of networked
computers used in Soviet schools.

A few readers expressed some confusion over the conversion of Soviet
prices in rubles into dollars, as
mentioned in the article on Soviet computing. It is a confusing
subject. There is an official exchange
rate, which typically overvalues nearly everything in the Soviet
Union, especially foreign-made goods.
There is an illegal street market in currency, the rates of which vary
widely depending on where one
goes. Then there is also an illegal exchange rate that applies to
goods themselves, and this rate
fluctuates according to the type of good being traded. All of this is
compounded by the fact that some
goods in the Soviet Union are grossly overvalued, while many are
grossly undervalued energy, for
example, is sold at about 25% of its real market value, while prices
for video cassette recorders and
video tapes are far above their market value. The whole picture is
even further complicated by the fact
that street prices in Moscow are not reproduced throughout the country
as a whole. The prices
estimated in the last Newsletter article were calculated on seven
rubles per dollar in the illegal
currency market in Moscow, for computers sold in Moscow in and around
April, 1989. Other
travellers may have very different experiences.


DIAC-90

Call for Papers

The CPSR research symposium, Directions and Implications of Advanced
Computing (DIAC), will be
held July 28, 1990, in Boston, in conjunction with the 1990 meeting of
the American Association for
Artificial Intelligence.

Papers should address some significant aspect of the social impact of
computing technology. Broad
topics include research directions in computer science, defense
applications, computing in a democratic
society, and computers in the public interest.

Submissions will be read by members of the program committee, with the
assistance of outside
referees. Complete papers should include an abstract and should not
exceed 6000 words. Papers on
ethics and values are especially desirable. Reports on work in
progress or suggested directions for
future work as well as appropriate surveys and applications, will also
be considered. Submissions will
be judged on clarity, insight, significance, and originality. Papers
(4 copies) are due by March 1,
1990. Notices of acceptance or rejection will be mailed by April 15,
1990. Camera ready copy is due
by June 1, 1990. Send papers to Douglas Schuler, Boeing Computer
Services, MS 7L-64, P.O. 24346,
Seattle, WA 98124-0346. For more information contact Doug Schuler at
(206)865-3226.


CPSR Members If You Move...

Please be sure and send your new address, along with your old address,
to the CPSR National Office so
your records can be properly updated. The CPSR Newsletter is mailed at
nonprofit bulk postal rates,
and the Post Office will not forward bulk mail, even if there is a
forwarding order on file. If you have
moved and have not received recent issues of the Newsletter, this may
be the reason.

Preliminary Call For Participation

CONFERENCE ON PARTICIPATORY DESIGN

Sponsored by

The Computers in the Workplace Project of
Computer Professionals for Social Responsibility

Seattle, Washington Saturday and Sunday, March 31-April 1,1990

Despite the proliferation of software design methodologies, we
commonly end up with systems that
appear to have been designed in spite of the user. Participatory
design seeks to avoid this problem by
emphasizing extensive prototyping and user involvement in the entire
design and development process.
Participatory design has been more successful in Europe, especially in
Scandinavia, and less so in the
United States, where the techniques are less well known. This
conference explores and promotes
participatory design methodologies for the workplace by bringing
together diverse groups representing
European and domestic practitioners, designers interested in the
techniques, and, especially, users.

The first day of the conference will emphasize invited speeches on
topics such as the following:

¥ What is participatory design?

¥ International perspectives on participatory design

¥ Experiences in participatory design

¥ Hazards of omitting user involvement (experiences in non-
participatory design)

¥ Role of organized labor

¥ Organizational resources and constraints

¥ Workplace democracy

¥ Strategies and tools

Workshops will dominate the second day. We are soliciting proposals
for half-day workshops on the
above or relevant topics. A wide variety of highly interactive formats
are welcome. Examples might be
interactive tutorials, panel discussions with small audiences that
emphasize audience participation, or
workshops that break into small groups to explore others' views in
detail. One-to-two-page workshop
proposals are due via mail or E-mail by December 15, 1989. Please send
proposals to the address
below. If you need further information, please write to that address
or call Paul Czyzewski at (415)
967-7079 or Jeff Johnson at (415) 4948248.

CPSR Workplace Project
P.O. Box 717
Palo Alto, CA 94302

workplace@csli.stanford.edu