Volume 2 No. 2 The CPSR Newsletter Spring 1984


DARPA Strategic Computing Initiative
A progress report on the CPSR response
Lucy Suchman - CPSR/Palo Alto

Early this year, CPSR formed an ad hoc committee to respond to the Defense
Advanced Research Projects Agency's Strategic Computing Initiative (SCI). The
SCI, made public last October, plans to spend $600 million over the next 5
years. The program has already been funded by Congress, and proposals are
currently solicited from the computer research community. The CPSR Strategic
Computing committee's goal is to produce two position papers on the SCI, one
directed to the professional community, the other to policy-makers and the
public.

The initial focus of the committee s efforts is a statement directed to the
computer research community. The aim is to encourage debate - among those who
might actually work on SCI projects, and within the professional community at
large of the basic questions raised by the SCI. A draft of the statement was
circulated for comment, in March and April, to CPSR chapters across the country.
The recommendations returned by the chapters are currently being compiled and
incorporated into a final draft by members of the committee in Palo Alto. A
final version of the statement will be issued in early May.

Work on the statement has produced a consensus on three basic concerns:

1. The SCI promises specific new weapons systems; autonomous vehicles, such as
robot tanks; a combat pilot's "associate;" and an aircraft carrier group "battle
management system." Our concern is that proposals for computer research will be
assessed by their relevance to these specific applications, rather than by their
general scientific merit.

2. The SCI promotes the use of machine "intelligence," to control the operation
of complex military systems under unpredictable circumstances. Our concern is
that, particularly when the stakes are high, situations d extreme uncertainty
are precisely the wrong environment for the application of artificial
intelligence.

3. The SCI promotes the military application of computer technology as a
solution to perceived problems in defense. Our concern is that, rather then
increasing our security, past attempts to achieve superiority in new weapons
technology have fueled an arms race that has no foreseeable end.

This last concern is the gravest. In the final analysis, we believe neither that
the path to national security lies in military superiority, nor that superiority
can be achieved through the use of computers.

The response intended for policy makers and the public will provide a summary of
the SCI, an analysis of the SCI's objectives in terms of our security, and the
appropriate use of computer technology.

Information or questions regarding the progress of the CPSR response to the SCI
should be addressed to the committee c/o Jorge Mezei, P. O. Box 717, Palo Alto,
CA, 94301.

Some Thoughts on Military Funding
Terry Winograd - CPSR/Palo Alto and
Dept. of Computer Science, Stanford University

When an individual refuses military funding, is it a 'symbolic act' or something
with a recognizable political effect? Over the past years, I have avoided
applying for military funding but, by doing it privately. I have avoided a wider
responsibility to influence others. I share the thoughts that follow in the hope
that they may have a greater effect in the world.

Like many computer professionals. I was supported under large ARPA grants during
my graduate work and first few years of teaching. Since 1975, when 1 got off
Stanford's large Artificial Intelligence Lab grant, I have had a mixture of
partial support from other sources, and have at times come under fire for not
doing my share to bring in faculty salary offset money (which the department
counts on). People have often pointed out to me that DARPA (Defense Advanced
Research Projects Agency) or ONR (Office of Naval Research) would be happy to
get proposals for various aspects of my research. In fact, at one point when I
was seeking foundation funds for basic research on software development, I was
told essentially, "We can't fund that because it is something that would
obviously get DARPA support, and we use our money for things that wouldn't
otherwise get a chance."

The Individual and the Collective

In understanding why I don't want to work with military funding, it is important
to recognize the importance of collective efforts, and the fallacy of taking an
individualistic view. The standard arguments for taking the militaryÕs money are
all of the form 'What difference would it make if I didn't? Somebody else
would....' (in fact there are various opinions as to what someone else would do,
as discussed below). Consider an analogy. Imagine that everyone in my
neighborhood burns their garbage, resulting in noxious pollution. I say that I
am going to quit burning my garbage. In some sense it is a symbolic act. N-1
people burning the garbage will produce essentially the same level of pollution,
and I will end up having to pay for a trash hauler. But of course, if we all
quit burning it we would all breathe better, and that's where the collective
effect comes in. If I take the view that the rest of the world is a Fixed
effect' then indeed anything I do is useless. The issue is to ask what it is
that we are all doing together, and what we all might do differently, then to
look at ways in which individuals can make changes.

Is Military-directed Research Bad?

The first obvious question is whether there is a problem at all. My attitude
makes no sense to someone who sees the scientific support of U. S. military
goals as positive. If you believe that for some reason (to fight global
communism, to keep our country the top military power in the world, etc.) it is
your patriotic duty to help the military, then it of course makes sense to take
their funding and do the best you can. We might argue over whether this policy
actually leads to national security and whether it leaves a future for the
world, but the issue certainly isn't that of funding.

I believe that increased levels of military technology will not help the future
of this country, and in fact will be counterproductive. I do not want to
actively participate in increasing the killing power of the military, and refuse
to work on projects whose applications are primarily weapons related. In the
rest of this I will be addressing those who agree that in some way or another
military goals and programs (as defined by the Pentagon and implemented through
their spending policies) are wrong.

This stance (along with the refusal to do classified work) is fairly widespread
in the university world. It is not at all universal, and there are plenty of
examples of purely military research under university auspices (such as the
University of California management of the Livermore nuclear weapons labs). But
the great majority of academic researchers (in computer science and in other
fields as well) would rather not do weapons work. They go to some length to
characterize their work as Basic research,' which although it may have military
applications will have a wide range of potentially beneficial applications as
well.

What, then, is the problem with taking military money to do research that I
myself have selected, and for which I see positive applications? What is wrong
with getting ONR or DARPA money for a project that I have applied for grants
from other sources to do?

My answer here is that in addition to the world danger created by the weapons
spiral, there is also a strong danger in an increasing rob for the military in
the U. S. in general. As an employer, a purchaser, a lobbying group, a trainer
of people, and many other things, they have a negative influence on attitudes
and practices. Once again, this is not the place to go into detail, and there
may be those who think military governments are better. My view is based on
wanting to decrease the overall militarization of the country.

By allowing the military to be the largest sponsor of university research, we
give them tremendous power over its direction and over the conduct of
individuals within the university. This power is rarely wielded bluntly
(although recent attempts to keep non-classified work from visitors, to require
pre

authorization from the military before the publication of unclassified research,
and to block foreign students from access to university research come close). It
is easy to take a naive individualistic view and say that even though the
military sponsors the research, it is the individual faculty member who chooses
the topic, writes the proposal, and directs the work.

The fallacy is once again the failure to look at the larger collective. DoD does
not need to tell you what to do. They need only announce that some large new
amount of funds is available for work in area X, and to stop funding proposals
in area Y. They do not need to come to campus and squash radical demonstrations.
It is more effective just to let it be known that they find the atmosphere on
campus too hostile, and that they may find it necessary to switch more research
to think tanks, to avoid problems. Once the university has become dependent on
military funds for its survival, it is very hard to take a stand on some Minor'
issue which could jeopardize everything. A few minor issues soon add up to a lot
of control.

Finally, an individual who is being supported by the military will be somewhat
more hesitant to be critical of U.S. military policy or of military spending. It
is doubtful that the DoD would do something as heavy-handed as retaliating by
canceling a contract or grant, but there is no guarantee of what they will be
thinking when the next round of funding is given out. Also, there is the natural
human tendency to want to feel good about what we are doing, and to avoid
thinking about those aspects that might feel unsettling

In addition, there is a pernicious overall effect of using the Pentagon as the
nation's channel for supporting institutions of higher education with federal
funds The balance of research and teaching, of science and humanities, of
graduate and undergraduate education are all shifted to suit the needs and
tastes of the generals. And as mentioned above, every additional dollar that
goes through their budget (regardless of what it ends up being used for) gives
them additional economic influence.

But What About......

Many people who agree with my basic attitude towards the military, and who may
grant the logic in these arguments, still feel they must (or want to) take
military research money. There are a number of common arguments:

1. I don't like militarism, but if we don't do military research the Russians
will anyway.

This gem to the larger issues mentioned above. I simply do not believe that
increasing the military technology of this country is the way to get national or
world security, in spite of whatever the Russians do. There are many articles in
the current anti-nuclear movement supporting this.

2. H I don't take the money, somebody else will anyway.

The answer to this should be obvious from the discussion of collective effects
above. If we can get a consensus not to take it, nobody will. Even if we get
only a partial consensus it will greatly reduce the military influence.

3. If I don't take the money, the military will buy more bombs with it.

This is an obvious application of the fixed effect fallacy. Part of the
Pentagon's argument before the budget committee is that they need a lot of money
to do the scientific research. Without that research there is no reason to
believe they would get the same amount of money. Of course they might, but it
depends on the overall societal vows and pressures that go along with a move
away from military funding. Besides, the total university research budget is a
tiny drop in the military bucket. Cancelling all of the DARPA university
contracts wouldn't buy one more nuclear submarine.

4. It doesn't matter who pays for it, the military can use it.

'So what if the government decides to sponsor everything through NSF. If it has
military uses they will use it anyway.' This is one of those arguments that
rests on taking an all-or-nothing view. Of course there is no part of this
society that is not connected to the military. If I design a computer for Xerox,
Xerox will be willing to sell it to the Army. If I design a better pair of
shoes, who is the largest single buyer of shoes in the country? Nothing I do
that benefits people through technology will be totally 'clean' of military
benefit. Ail that is safe from contamination is isolation in my home-grown
commune in the hills.

But the fact that there is no clear line does not mean that there is no
difference. If I am studying the ways to get the most thorough pattern of
coverage for little metal needles dispersed by an explosive charge, I am clearly
doing highly military research, although a clever enough person might invent a
pacific use. On the other hand, if I develop penicillin (although it was of
great benefit to the military in dealing with wounded soldiers), I can feel that
the overall benefit is to society as a whole. Each project must be seen in this
light, and even within computer science there is a spectrum. LOGO may be useful
for training sailors, but its emphasis has been on
working with children. Distributed sensor systems may have peacetime uses (I was
told by a prominent artificial intelligence researcher that they weren't purely
military since they might be
used somehow in hospital intensive care units), but they are 99% intended for
monitoring the movement of people and machinery in hostile territory. There are
differences, and they are affected by the choices and desires of those who set
up the funding programs.

There are those (such as Weizenbaum) who believe that the net effect of all
technological development (computers being a prime example) will be negative --
if not because of military applications, then because of other effects on the
quality of life. There is no clear-cut answer to this (and indeed the issue
needs to be taken more seriously than it generally is within technological
institutions), but I don't myself believe it. I think there We better and worse
technologies, and that we are better off to explore the differences than to
abandon the field.

5. People doing the research should be those who are sensitive to the political
issues.

It is pure self-deception to believe that the political attitudes of those doing
scientific research (no matter what they are) will be listened to by the agency
that funds it. They don't ask you for your advice on arms limitation, just on
pipelined processors. I do believe, though, that the parallel objection is valid
on the more general scale of who does the scientific and technological work
within the society as a whole. Being 'experts' does (rightly or wrongly) bring
with it a degree of access to public opinion that is politically important, and
that can be used for positive purposes.

6. The military does a better job of research funding.

This is the argument I hear most often from people in high positions in the
computer science academic community. Their experience is that military funding
of computer science has been more efficient, more effective, and has had fewer
Wrings attached than any other source of funding. They may be right. It is clear
from my own experience that getting funding through NSF can be a frustrating and
time-consuming process, and that the choice of scientific goals is influenced by
many more political considerations. This is due to the nature of the peer review
process (both slow and political) and the sensitivity to congressional scrutiny
(which includes special interests and pressures for geographical distribution of
funding).

This is perhaps the most seductive and most dangerous argument. Legend has it
that Mussolini won favor among the Italian people because he made the trains run
on time. The ability to get things done quickly and efficiently is always a
strong point of central authoritarianism, not democracy. The danger is in a
gradual process of legitimization, in which more and more power is placed under
military control because of frustration with the difficulties of democratic
processes.

7. I need the money and can't get it (or can't get it without much greater
difficulty) anywhere else.

This is a legitimate argument which forces each individual to examine his or her
own priorities. It is perfectly logical to grant all of the points made above,
and to take the stand that your own individual good is best served by taking the
money (which I need to do the research, which I need to get tenure, ....) even
though there are negative consequences. It is important not to get moralistic
here and label anyone who takes such a stand as immoral. I can afford to write
this paper because I do have tenure. I didn't do it when I was an assistant
professor. I do not believe that it is realistic now to push for the computer
science department to stop taking all military funding. Given the national
situation, it would be economic suicide for our research and educational
programs. The point is that the way out is to take more collective action, 90
that the larger pattern shifts and it does not require an individual sacrifice
to take a stand. Of course, individuals may choose to make sacrifices either as
a moral position, or as a way of calling attention to the issues.

The patterns can be shifted tin many ways. The government can be pressured to
change the overall distribution of support. We can look for areas of research
that are not as capital-intensive (in fact, I believe that might be better for
the intellectual quality of the department, independent of the military issues).
We can look into ways to develop other sources of funding (as has gone on with
industry, although this raises its own problems).

So... ?

Having laid all this out, I don't have the answers. I don't think a simplistic
approach (stop all military funding, or ignore the problem) will solve things.
It will take creative work to be effective.

The preceding article will be included as part of the forthcoming C.P.S.R.
"packet" describing the Strategic Computing Proposal.

Chairman's Report
Severo Ornstein - CPSR Chairman

During the first quarter of this year, the Executive Committee prepared a
proposal to fund CPSR for a one year period. The total proposed budget
(including membership dues and hoped-for foundation and industrial support) is
$147.000. In addition to operating the national office (rent, phones, supplies,
computer, etc.) this budget includes salaries for three full-time employees: an
Executive Director to manage the office, oversee the organization, seek further
funding, and deal with membership and chapter issues; a Program Director to
coordinate all of the substantive technical work, including writing, speaking,
workshops, symposia, and developing position statements; a secretary to provide
administrative help, principally with the office and organizational chores.

The proposal has been sent to over twenty foundations, and has been accompanied
by individual lepers, phone calls, and in some cases visits to foundation
directors. So far, we have received three grants: $15,000 from the Ploughshares
Fund, $5,000 from the New-Land Foundation, and $1.000 from the Goldman Fund. We
have also had one rejection. We expect positive responses over the next several
months, and are hoping that we may have sufficient funds to permit us to hire an
Executive Director sometime in June.

The day-to-day workload at the CPSR office has become substantial. As a result,
our valiant volunteers have been buckling under the load. The Executive Director
is the person most urgently needed, and we are presently interviewing
candidates. We welcome suggestions and pointers to people who have experience in
building membership organizations. Computer experience is desirable but not
essential.

We are also looking for suitable candidates for the Program Director's job, but
with less urgency for several reasons. First, we believe that once the
organizational workload is passed over to a full time Executive Director and
Secretary, we will be in a much better position to handle the substantive work
ourselves. Second, under even the best of circumstances, we are not likely to
have sufficient funding to cover this position until late summer or fall.
Finally, we have a potential candidate in mind, who we hope will eventually
accept the position but who won't be available until next spring. In the
meantime, we will be considering alternative candidates and would welcome
suggestions. Candidates for this position should have either a Ph.D. in computer
science or equivalent professional credentials. The Program Director will be the
principal spokesperson for the organization and must be able to speak with
authority.

In closing, I would like to say a cautionary word about representing CPSR. The
issues we are addressing are extremely complex, and all of them are open to
question and debate. It is tempting to lay hold of a position, cling to it
firmly, and declaim it loudly. But in the long run I believe that CPSR's
effectiveness will be greater if we make careful, reasoned statements that raise
questions, point to problems and concerns, suggest alternatives, and leave the
ultimate decision to the listener. CPSR is an organization with a diverse
membership; some members hold positions that other, equally concerned members
would consider extreme. Recently, at the West Coast Computer Faire, some
doubtless well-meaning member made a public statement, advocating the
termination of all DoD-funded computer research, that was interpreted as
representing a "CPSR position." The statement brought the following response on
an electronic bulletin board distributed over the ARPANET: "Up until now I
thought CPSR was a reasonable watchdog to prevent work in totally unethical
areas. Now they seem to be a bunch of nubs. I hope this impression is wrong . .
." It is even unlikely that the statement that engendered this response would be
acceptable to the majority of our membership, in mat it is too simplistic and
categorical.

It is difficult to strike a balance between statements that are so bland as to
be useless and so strident as to put off a large fraction of those to whom we
would address ourselves. If we are to move anyone who is not already in
agreement with our arguments, we must strike that balance extremely carefully.
The CPSR response to DARPA's Strategic Computing Initiative, which will be
released shortly, represents an attempt to find such a balance. In the last
analysis such "official statements" represent a melding of attitudes and
opinions from as much of the CPSR membership as we can reasonably contact,
tempered by the tastes and opinions of CPSR's Executive Committee and those who
take it upon themselves to do the final writing.

It is wiser, more defensible, and in the long run more effective to raise
questions in peoples minds, and let them draw their own conclusions, than to
tell them what to believe. CPSR is an educational organization, and the purpose
of education is to make people think about -- not necessarily agree upon -- the
critical social choices that effect us all.

Chapter News

As space allows, we will continue to reprint brief news from our chapters.
Chapter secretaries are encouraged to send contributions to the address on page
1.

New York: About 35 people, mostly new faces, attended our last meeting. We
discussed what CPSR/NY should be doing and what individuals could do to work to
prevent nuclear war. Some people brought up connected issues, such as opposing
AFIPS holding their conference in South Africa. There was a strong sentiment
that this was an issue of professional ethics and letters of protest should be
sent to AFIPS. Our mailing list continues to grow, and we are getting inquiries
from New Jersey and Connecticut as well.

Seattle: The Seattle chapter meets on the 4th Tuesday of each month
(exceptionally, in May on the 5th Tuesday). Recent activities have included
discussion of the CPSR response to the Strategic Computing document; one of our
members spoke on the proposal at a panel on artificial intelligence at the
University of Washington Computer Faire. Our speaker's bureau continues to
receive requests; in the past few months we've given talks at Green River
Community College, the University of British Columbia, and to two Rotary Clubs
in Port Angeles, Washington (on the Olympic Peninsula). Next month, among other
activities we will have a booth at a Career Fair at a local high school.

Palo Alto: Palo Alto continues its program of monthly speakers. Recent speakers
have included Dr. Wolfgang Panofsky, former director of the Stanford Linear
Accelerator Center, discussing the fallacy of a Strategic Defense against
nuclear missiles; Robert Aldridge, author of First Strike (South End Press,
1983) and former Lockheed design engineer for the Polaris and Trident missile
systems; and Thomas Ackerman, co-author of the recent Science report on "Nuclear
Winter" (23 December, 1983). On May 9 Bill Perry, Jr., will discuss the ethics
of weapons-related work, and his decision to resign from the position of
Director of Public Relations at Lawrence Livermore National Laboratories.

More than MAD
Greg Nelson - CPSR/Palo Alto

An earlier piece by Greg Nelson ended with the question, "It seven hundred
nuclear bombs will reduce the Soviet Union to a smoking, radiating ruin, what
are we going to do with the other nineteen thousand three hundredth In this
installment. Greg goes on to examine the arguments for going beyond MAD.

Many people from all sides of the political spectrum are convinced that if
nuclear weapons are useful at all, they are in any case not useful in large
quantities. For example, Nikita Khrushchev said bluntly to us long ago. "You say
you have enough nuclear power to kill the population of the Soviet Union three
times over. Well, we only have enough to kill the population of the United
States once - and that is enough" More recently, Henry Kissinger said in
exasperation to a proponent of some new weapons system. "What in the name of God
is strategic superiority? What is the significance of it, politically,
militarily, operationally, at these levels of numbers? What do you do with it?"

Henry Kissinger's remark is especially notable, because he is reputed to be a
man with a subtle sense of advantage. If by any creative twist of diplomacy we
could derive an advantage out of thirty thousand nuclear bombs but not out of
twenty thousand, I think that Dr. Kissinger would have thought of it.

Robert McNamara, Secretary of Defense under presidents Kennedy and Johnson,
decided that the United States should content itself with enough bombs to assure
the destruction of the Soviet Union He commissioned studies to determine the
exact number of bombs that would give us this assurance Then, to be on the safe
side, he tripled the number, replicating our assured deterrent in the air, in
the sea, and on the land. He tried to get the Joint Chiefs of Staff to abide by
his limits, but he was not very successful The McNamara doctrine is sometimes
called MAD, short for Mutual Assured Destruction.

Although the McNamara doctrine has many eminent supporters, it also has many
critics. The MAD critics give a variety of reasons for building bombs beyond the
McNamara limit. There seem to be three main lines of argument, which I call the
"warfighting," "continuity," and "credibility" arguments.

The warfighting objection disputes the principle that nuclear war would be
mutual national suede The warfighting claim it is possible to tight a nuclear
war and win a meaningful victory It is a mistake, the warfighters say, to think
of nuclear war as an apocalyptic city destroying spasm. Nuclear war, when it
comes, will be fought between our weapons and their weapons, with attempts made
to spare innocent civilians.

Here is Richard Garwin, an American strategic analyst, considering targeting for
minuteman missiles:

"If the United States had infinitely many RVs (Reentry Vehicles; i.e., bombs) it
should attack targets in descending order of ratio of military damage to
civilian casualties; it there are other current or future uses for RVs. however,
it should attach military value per se, as proposed here, until the allotted
number of RVs is used or until the civilian casualties have risen to the number
of U.S. casualties."

In other words Garwin recommends that our policy be to do as much damage to the
Soviet military machine as possible, while killing as few Russian civilians as
possible, unless this strategy would cause us to run out of bombs too soon, in
which case he recommends doing as much military damage as possible, irrespective
of the number of Russian civilians killed, provided that this number does not
rise above the number of U.S. civilians killed

James Fallows writes, ÒThe list of targets for American nuclear weapons is
contained in the Single Integrated Operation Plan, known in the business as the
SIOP (pronounced "sigh-op" ) Despite all the theories about ÔMutually Assured
DestructionÕ neither in the McNamara days nor today has it emphasized
'targeting' Russian cities simply for the sake of killing people. Indeed since
1973 it has been a crucial and clearly stated element of American policy that
the nuclear target list not include population centers Rather, it comprises
industrial concentrations, military bases, missile silos, submarine ports,
government centers and communications networks - many of which of course happen
to be in or near cities.Ó

Modern missiles are very accurate, the warfighters point out, so the war between
our weapons and their weapons could be carried out with a minimum of civilian
casualties. The radioactive fallout would kill many people, but nobody knows
exactly how many. An eminent expert, Dr. Edward Teller, has written in the
ReaderÕs Digest that fallout is an easily-visible white powder that is not
lethal if wiped off promptly.

If you accept the warfighting theory, then the question of the number and type
of nuclear weapons to be built is determined by measuring our military machine
against the Russian military machine. Lately, the warfighters have been
particularly concerned about the vulnerability of our Minuteman missiles to the
Soviet SS-20 missile It is felt that each incoming SS-20 might destroy several
Minutemans. because an SS-20 carries multiple warheads and is highly accurate.
The warfighters do not claim that a Soviet first strike could destroy our
deterrent, they agree that even after a worst case Soviet first strike, we would
retain enough bombers, submarines, cruise missiles, and surviving Minutemans to
utterly destroy the Soviet Union. But they believe that this deterrent is
irrelevant, since the essential nature of nuclear war is a battle between
weapons They believe that if we are left with inferior strategic forces, we will
have to surrender. And in measuring strategic forces, they place a high value on
"fast hard target kill capability,|' which is supplied by our Minuteman missiles
and not by submarines, bombers, or cruise missiles. This is the source of the
debate over the Window of vulnerability" of our Minutemans.

The second line of argument against MAD begins with the observation that even if
our military machine is powerful enough to assure the destruction of the Soviet
Union, it may be useless In deterring undesirable Soviet actions, because of
ÒdiscontinuitiesÓ in our deterrent

A continuous deterrent is one that deters the whole continuum of possible Soviet
military threats, rather than deterring only a Soviet nuclear first strike.
Suppose, for example, that a conflict in Southwest Asia pitted the U.S. Rapid
Deployment Force against Soviet ground troops, and that the Soviets, who have
fewer logistic problems than we do in that theater, were winning We could try to
issue an ultimatum: "Withdraw your troops or we will destroy the Soviet Union!"
But they know that we know that they would have the power to destroy us back, so
they would probably ignore the ultimatum. Our assured capability to destroy the
Soviet Union is counterbalanced by their assured capability to destroy us But
suppose we issued the ultimatum: "Withdraw, or we will destroy your troops with
nuclear weapons!"? This ultimatum is more plausible, since after their troops
were destroyed, the ball would be in their court, and they would be reluctant to
destroy the United States, knowing that we could destroy the Soviet Union in
return.

The consideration of such scenarios led to the doctrine of 'flexible targeting,'
which just means that in a crisis the president should have a choice of where to
shoot our missiles: the all-or nothing choice between firing at Russian cities
and doing nothing would leave us with no response in many conceivable scenarios
The doctrine of flexible targeting is not inconsistent with MAD, and both
McNamara and Kissinger have endorsed it.

If the flexible targeting doctrine IS carried to Its logical extreme, the result
is the doctrine of 'continuous deterrence' we must have a continuity of threats,
ranging from conventional forces through tactical and intermediate range nuclear
weapons to strategic weapons, In the words of the Department of DefenseÕs
Annual Report to the Congress for Fiscal Year 1984, our deterrent must be
continuous "at all points along the spectrum of violence,Ó This report also
identified other dimensions along which our deterrent must be continuous:
geographically and temporally, for example. (Thus we should be able to respond
at any given level of violence, at any given place in the globe, after any given
number of days of war.) All in all the report identifies seven dimensions along
which our deterrent should be continuous. To provide this many military options
requires building more weapons, and more types of weapons; thus, the MAD critics
say, we cannot shut down our bomb factories just because we can assure the
destruction of the Soviet Union.

The third line of argument against MAD concerns the "credibility" of our
deterrent. The concern here is not that the Russians might think our bombs are
duds, or that our missiles won't work. The concern is that the Russians might
think we lack the will to fire them.

Suppose we let the Soviets outpace us In strategic weaponry
There will be some in the Soviet Union who subscribe to the warfighting theory,
and they will imagine that victory over the United States is possible. And the
others in the Soviet Union, who reject the warfighting theory, will reason, "at
any rate, the United States has obviously given up any hope of winning a nuclear
war. Perhaps they have even lost their nerve Since we have all new weapons and
they have obsolete weapons they will not dare to challenge us to a round of
chicken. This is a favorable time for expansion in Southwest Asia," The Soviets
will thus be tempted to try to push us around. We will have to give in, or
challenge them to a round of chicken. Better to be safe than sorry; let's keep
our strategic systems up to-date. So these critics reason

From the Secretary's Desk
Laura Gould - CPSR Secretary

The response to my note in the last Newsletter, requesting that members renew
early and consider sending an extra donation has been quite overwhelming. Many
did indeed renew before receiving a notice, thus saving us considerable effort
and money. We give them our thanks. Many also sent extra donations, some quite
large: we received one of $500, several of $100 and $200, as well as numerous
smaller donations. All of these donors have been thanked individually, but we
wish to thank them publicly as well. Such responses make us feet very well
supported, both financially and psychologically.

Our LANAC lawyer, Paul Valentine, has written some guidelines about what CPSR
should and shouldnÕt do to preserve its recently acquired 501(c)3 tax exempt
status. We will try to prepare a synopsis of his guidelines for the next
Newsletter. In the meantime, it you have questions with regard to these matters,
please call the national CPSR office for advice


Miscellaneous

The CPSR Newsletter is published quarterly by

Computer Professionals for Social Responsibility
Daniel Ingalls and Donna Osgood, editors
PO Box 717 Palo Alto, CA 94301.

The purpose of the Newsletter is to keep members informed of relevant thought
and activity in our organization With this Issue, Mark Hall has at least
temporarily retired as co-editor of the Newsletter . We and fortunate to have
Donna Osgood, an associate editor of BYTE magazine, to help assemble this and
future editions of the Newsletter

We welcome comments on the content and format of our publication Most especially
we welcome contributions from our members. The deadline for our next issue is
July 6, l984. Thank you all for making CPSR what it is.


Random Bits

¥ On April 5, the Union of Concerned Scientists (UCS) put on a two-hour
teleconference on the topic of Reagan's "Star WarsÓ speech, aired just over a
year ago. The program was moderated by Henry Kendall of UCS and featured Carl
Sagan, Richard Garwin, and Noel Gayler (a retired admiral, former head of the
Pacific Fleet and of NSA) as panelists. About eight cities were linked so that
during the second hour questions could be asked of the panel from each center.
The program constituted a call to scientists across the nation to band together
to oppose the development of a ballistic missile defense system on the grounds
that it is destabilizing and fraught with technical difficulties We hope to
acquire a tape of this excellent presentation

¥ At the recently held 7th International Conference on Software Engineering, a
panel discussion entitled "Software Engineering and Social Responsibility" was
conducted by Susan Gerhart of the Wang Institute,. On the panel was Guy Almes
from the University of Washington Computer Science Department and the Seattle
Chapter of CPSR. Guy described CPSR to this technical audience so that those
present would be aware of its aims and its concerns. He was careful to avoid
stereotypes of computer professionals who work in defense, in hopes that those
present would similarly avoid stereotyping CPSR as naive and arrogant A specific
thrust of his talk was that, in spite of the many reasons to be cautious about
the reliability of Command and Control systems, contemporary trends In defense
and foreign policy (e.g., Pershing II deployment) put additional pressure on our
leaders to rely more heavily on these Increasingly sophisticated systems The
talk was well received, particularly in view of the large number of defense
workers present.

¥ We have a tape of an interesting presentation on Nuclear Winter, involving a
TV hookup between US and Soviet scientists discussing the report released in
Science this fall Opening and closing remarks are made by Carl Sagan



Recommended Reading

With this issue we bring to your attention several recent publications of
interest We welcome further recommendations from our readers

"Space-Based Missile Defense," a report issued by the Union of Concerned
Scientists, and excerpted in the New York Review of Books of April 26. The
report was put together by a study panel made up of Kurt Gottfried (Professor of
Physics at Cornell), Henry Kendall (Prof. of Physics at MIT), Hans Bethe. Peter
Clausen (Wilson Fellow, and former government policy analyst), Richard Garwin.
Noel Gayler (Ret. Admiral, USN). Richard Lebow (Professor of Government, Cornell
and former professor of Strategy, National War College). Carl Sagan, and Victor
Weisskopf (Physics Professor Emeritus, MIT).

"Reflections: Weapons and Hope," an excellent series of four articles by Freeman
Dyson in New Yorker Magazine. Questions" (Feb 6) is deliberately easy to read to
lower the guard of the casual reader. "Tools" (Feb 13), reminiscent of Herman
Kahn, faces the unmentionable, and is quantitative, in order to attract the
weapons people. "People" (Feb 20) is discursive, enchanting, very humane, in
order to attract the weapons control people. "Concepts" (Feb 27) is the punch
line. With his audience now attentive and informed, he rejects six positions of
the right and left, and presents his moderate "live and let live" strategy.


CPSR in the News

Our fledgling library attempts to maintain a record of all articles about CPSR
that occur in the public press Please send copies of such articles and
descriptions of TV coverage, to Librarian, CPSR, PO Box 717, Palo Alto, CA 94301
Well try to print a quarterly listing in each issue of the newsletter

February
Staff writer John Verity wrote a long and excellent article on CPSR in the
February Datamation entitled Nuclear War & the Computer (page 50). This story is
tied closely into the preceding one entitled DARPAÕs Big Push in AI (page 48).

February 13
Mark Hall, an original co-editor of this Newsletter, wrote a story published in
the In Depth section of COMPUTER WORLD under the title High-tech Dreams Nuclear
Nightmares (page ID 21) Since he included a box with names and addresses of
relevant organizations (including CPSR), a lot of mail resulted from the
article.


Forthcoming Events

Alan Borning will be representing CPSR at a workshop on Accidental Nuclear War
at the annual meeting of the International Physicians for the Prevention of
Nuclear War in early June in Helsinki. In addition to participation in the
workshop, Alan will address the plenary session which includes representatives
from over forty-five nations. We hope to have AlanÕs report in the next issue of
the CPSR Newsletter

Before going to Helsinki, Alan will spend a day in Bonn, Germany, attending a
gathering of about 100 computer scientists who are interested in forming a group
similar to CPSR in West Germany. They particularly requested the presence of an
American computer scientist so that their efforts will not be viewed as anti-
American, and hope to affiliate in some fashion with CPSR.

This year's National Computer Conference (NCC) will be held in Las Vegas in
early July. We have still received no format written rejection of our request
for a counter, but have been informed verbally that their new policy is to give
counters only to organizations which are members of AFIPS. Incidentally, NCC
plans to honor Lawrence Livermore Labs on their Pioneer Day!