ARMS-D-Request@XX.LCS.MIT.EDU (Moderator) (11/08/86)
Arms-Discussion Digest Friday, November 7, 1986 6:50PM
Volume 7, Issue 50
Today's Topics:
prompt responses to nuclear attack
prompt responses to nuclear attack
prompt response
waiting a while before taking revenge
Yet more on SDI
----------------------------------------------------------------------
Date: Fri, 7 Nov 1986 15:07 EST
From: LIN@XX.LCS.MIT.EDU
Subject: prompt responses to nuclear attack
From: <"NGSTL1::SHERZER%ti-eg.csnet" at RELAY.CS.NET>
Given that we have a place for tankers to land and refuel, we can keep
SOME of the bombers in the air for 24 hours. This means the use of
airfields which will not exist after the Soviet attack so it is not
feasable in this situation. Also, the time needed is more like 36 hours
(they still need to fly to the USSR). After talking with a firend
who is a retired tanker navigator, I was assured that it is not
possible to keep the tanker and bobmer force aloft for the 24 hours
(much less the actual 36 hours).
So, after 24 hours, we have no land based ICBM's, no bombers, some
(exact number unknown) of subs. We also have no way to order the
retaliation (because of no C3).
I agree we can't keep the entire force in the air for 24 hours. But
as you note, we can keep some of it in the air for that long.
Submarines are capable of launching without a go-code. That is still
a formidable force.
------------------------------
Date: Fri, 7 Nov 1986 15:14 EST
From: LIN@XX.LCS.MIT.EDU
Subject: prompt responses to nuclear attack
From: lin
>Please give an example of how waiting 24 hours would significantly
>impede the U.S. response. Without such an example, the only plausible
>conclusion is that you are engaging in hyperbole and ad hominem
>argument, suggesting that "if the Soviets want it, it must be bad for
>us, and if you want the same thing, you must be in their camp."
From: ut-sally!ut-ngp!melpad!reality1!james at seismo.CSS.GOV
Exactly how long would it take the Soviets to completely destroy all
land and
air based systems in an all out attack? My assumption is that
everything but
the submarine based forces would be long gone by the time 24 hours rolled
around..
If it is an assumption, then I can't challenge you. If it is a
prediction, then it depends on the nature of the attack. An attack
heavy enough to kill all the non-submarine forces is enouch to warrant
an all out retaliation on the S.U.
... the president and most other command forces would also be
"incapacitated". I have no idea how fast the military system could be
disabled, but since it is so easy to track the location of the president, I
must assume that the president would be a primary target in an attempt to
confuse the command picture (and there's no way a first strike would miss
the president).
But there are back-ups available to the President.
Rather than say "if the Soviets want it, it must be bad", one might say "If
the Soviets want it, it *might* be bad".
I'd like to keep the Soviet preferences out of any discussion
regarding what is good for the U.S. I don't care why they say what
they say. If the Soviets want X, X might be good or it might be bad.
I don't see how saying they want it should strengthen or weaken an
argument for or against X.
------------------------------
Date: Fri, 7 Nov 1986 15:20 EST
From: LIN@XX.LCS.MIT.EDU
Subject: prompt response
... Do you
really believe that the remaining submarines we would have would have
even a slim chance at destroying the Soviet military after our silos,
bombers, major cities like Washington D.C., and probably all of our
military bases have been destroyed?
After such an attack, we would go after the entire target set of the
S.U. The power of a submarine is not to be trifled with.
Even if they did survive, and
even if they had enough power to do the job, they won't have a common
communication link!
They don't need it to launch a retaliation.
------------------------------
Subject: waiting a while before taking revenge
Date: 07 Nov 86 18:27:16 EST (Fri)
From: dm@bfly-vax.bbn.com
> From: cfccs at HAWAII-EMH
>
> I don't know where your head is on the 24
> hour wait issue. Do you really believe that the remaining submarines
> we would have would have even a slim chance at destroying the Soviet
> military after our silos, bombers, major cities like Washington D.C.,
> and probably all of our military bases have been destroyed? Even if
> they did survive, and even if they had enough power to do the job,
> they won't have a common communication link! It was just destoyed
> along with all the bases and the battle staff! I know the Navy has an
> over-inflated vision of its own power, but this is absolutely
> ridiculous! Even they would say so!
Well, I'm not sure I want to defend the whole 24-hour delay idea, but
it's not as ridiculous as you let on.
Assume the Soviets miss ONE Trident Submarine. I've forgotten the
exact numbers, but I think there are (potentially) 400 independently
targetable warheads on that ONE SUBMARINE (40 Trident missles, each
potentially carrying up to 14 warheads (Herb, are these numbers
correct)?)
If I were the commander of that submarine (and I were willing to
commit mass murder in revenge for mass murder) I would give serious
thought to using a shotgun approach to retaliation, aiming for every
city with population greater than 20000 in the SU. If I were a
Pentagon planner contemplating a potential reactions to a first strike,
there would already be sealed envelopes on board every missle-carrying
sub with just such a contingency plan.
No need for a communication network, although there might be a few
TACAMO planes circling the Pacific and North Atlantic.
Of course if they get all the Tridents and leave only a Poseidon, that
leaves only about 72 warheads (24 missles at 3 warheads each). It is
instructive to look in an almanac at the list of the 100 biggest
cities in the Soviet Union, and think about destroying the biggest 72.
Thinking about it a bit more, I think I might wait even LONGER than 24
hours. It isn't really neccessary to kill the grandmothers and
children in Kiev or the laborers in Krasnoyarsk. Those people are
probably more or less on your side. The people from whom you want to
extract your revenge are going to spend the first couple of weeks
after the war relatively safe in concrete bunkers. I'd wait two
MONTHS, and launch a surprise retaliation in hopes of getting the
Soviet leaders in the open rather than in their bunkers.
------------------------------
3-Nov-86 21:49:24-EST,2913;000000000000
Date: Monday, 3 November 1986 08:04-EST
From: Jane Hesketh <jane%aiva.edinburgh.ac.uk at Cs.Ucl.AC.UK>
To: ARMS-D
Re: Star Wars flawed #1-of-10
[I received a batch of papers on SDI from some folk in the United
Kingdom; they ask me to release it in the Digest. They are about 11 K
chars each, and there are 9 of them. One will appear in each digest.
-- Moderator]
In Edinburgh on 19 June 1986, three representatives of the
Ministry of Defence, including members of the Ministry's
scientific advisory staff and of the SDI Participation
Office, received a day long briefing on the subject of
SDI.
The briefing was organized by the Edinburgh Computing and
Social Responsibility Group, and presented by six experts on
technical and socio-economic issues relevant to SDI. They
argued that SDI makes demands on computational technology
that cannot be met, and that in any case participation in
the SDI research programme is not in Britain's best
interest.
The major points raised during the day were:
o The design of the SDI system and its computing basis
would be necessarily flawed.
o The implementation of that design in computer hardware
and software would introduce further mistakes.
o The system could not be adequately tested to uncover
the resulting problems.
o Artificial Intelligence offers no magic solutions.
o Once deployed, there would be failures that the system
could not detect, which could be catastrophic.
o British computing research would be diverted from more
socially and industrially relevant work by SDI
participation.
o The research community would be divided and weakened by
security requirements.
o The impact on the British economy of SDI participation
would be negative.
The meeting was held as a result of the then Secretary of
State for Defence Michael Heseltine's response to a
parliamentary question by Tam Dalyell (MP) in December
agreeing to such a briefing (Hansard, 9 December, page 631).
The following 9 messages contain the briefing papers as they
were presented at this meeting.
For further information, contact:
Edinburgh CSR
3 Buccleuch Terrace
Edinburgh EH8 9NB
Scotland
or:
Frank van Harmelen, 031-225 7774 x220 (daytime).
Jane Hesketh, 031-225 7774 x248 (daytime).
or:
frankh%eusip@ucl.cs
or:
...!mcvax!ukc!eusip!frankh
==============================
Introduction
Statement of the Problem
On March 23 1983 President Ronald Reagan, in a broadcast
speech, called for the creation of a Strategic Defense
Initiative, to render "nuclear weapons impotent and
obsolete." Three years later no-one, with the possible
exception of the president himself, believes that such a
goal is feasible. No technologies exist or can be
anticipated which would provide the 100% effective wide-area
defense which would be required to achieve it. But billions
of dollars have already been spent in preliminary research
funding, and the budget proposal currently before the US
congress calls for a further sum in excess of four billion
dollars to be spent in the next fiscal year on research,
development and testing of SDI technologies. Towards what
goal then, if not the president's original one? No
unequivocal statement of that goal, in terms as
straightforward as the president's, has appeared. From a
variety of sources and statements one judges that the core
of the revised goal is to "enhance deterrence" by reducing
the degree of certainty with which a power contemplating a
first strike can estimate the effect thereof. Paul Nitze
has established two criteria for any system designed to
achieve this goal - it must be survivable, that is, proof
against a pre-emptive attack, and it must be cost effective
at the margin, that is, it must be more expensive to
increase the attacking force than to increase the defensive
capacity in response. Even this latter criterion has
recently come under pressure from the Pentagon, and has
reportedly been withdrawn from the most recent report from
the SDIO (Scotsman 3 June 1986).
For the sake of today's discussion, we will assume only
the most general technological outline - a layered ballistic
missile defense system, aiming at destroying attacking
missiles by a combination of land- and space-based weapons
and command and control systems. We assume further that the
most critical as well as the most novel aspect of such a BMD
system would be the `earlier' layers - boost-phase and mid-
course.
Goals of the meeting
In our presentations today and in the accompanying
discussions, we hope to demonstrate that some of the demands
placed on computational technology by any such BMD system
could not conceivably be met in any plausible time frame for
research and development, and that the entire programme is
thus fatally flawed and should be abandoned. Leaving aside
the question of participation in a project whose goals are
unachievable, we will also argue that British participation
in SDI research is against the national interest, and that
the MOD should make no effort to solicit or encourage
British institutions, industrial or academic, to participate
in SDI research, and should in fact endeavour to point out
to them the dangers in such participation.
For the record may I state that all those participating
today do so in a private capacity, although in all cases
speaking on the basis of relevant technical expertise.
Overview of the technical arguments
The SDI Battle Management System, if it were to be built,
would be the largest and most complex real-time computer
system ever. It would be composed of land-, air- and
space-based computers networked together, and its task would
be to continuously monitor a vast array of sensors,
determine when a hostile missile attack had occurred,
determine the trajectories of the missiles, allocate first-
layer weapons to missiles, direct their firing, assess the
results and try again at the next layer.
Can we design such a computer system? Given a design,
could we construct it in a way which realised the design
correctly? Having constructed it, could we trust it to do
what we want? The answer to all these questions is no. In
the presentations that follow we will present brief
summaries of some representative technical arguments which
lead to this conclusion. We have left substantial space in
the programme for discussions after each summary, to allow
points raised therein to be elaborated, and/or for related
arguments to be discussed.
The main topics to be covered are:
o+ The nature of the SDI BMS as a fully automatic
decision-making system, and the implication of this
given the active nature of SDI;
o+ The gulf between existing Artificial Intelligence
technologies and the requirements of SDI, particularly
in the area of general-purpose systems;
o+ The inadequacies of existing system-design
methodologies;
o+ The implausibility of using formal program verification
techniques to assess the accuracy of the
implementation;
o+ The inadequacy of built-in test equipment and fault
tolerance, and the insufficiency of testing under
simulation.
Overview of the socio-economic arguments
The total value of UK SDI contracts is likely to be small.
Nevertheless, there are reasons for concern in at least the
following areas: diversion of skilled personnel in key
fields; general distraction and division of the scientific
and technological community; freedom of research and
publication; long-term effects on the public perception of
certain disciplines.
Spin-off beneficial to the UK from SDI research is likely to
be slight.
------------------------------
End of Arms-Discussion Digest
*****************************