Privacy, security and public opinion

[Owen Lewis]

----- Original Message -----
From: "Brian Gladman" <brian.gladman@btinternet.com>
To: <ukcrypto@maillist.ox.ac.uk>
Sent: 01 June 2000 14:19
Subject: Re: Privacy, security and public opinion


> I can now see why you believe that cryptography is neither offensive or
> defensive. You appear to believe that the terms 'offensive' and
'defensive'
> can only be applied to an overall activity (or system) and not to its
> individual sub-components.

That draws things a little larger than I might do and - as I think you will
realise - is a line of thought that must fail if taken to extremes. Rather,
I
observe that artefacts as well as natural objects are not usefully described
in terms such as offensive and defensive or bad and good. Objects simply
are;  to attribute uniquely human values to them does seem to me unhelpful
for most purposes.

> Moreover you appear to be unwilling to accept
> that design intent rather than end use can give a technology an offensive
or
> defensive character.

That is true. Firearms are a very good example. With some exceptions, these
are optimised in design to aid the efficient killing of fur, feather, fin or
even mankind. No doubt such killing may appear offensive to the object
of the killing. However, from any other viewpoint the terms offensive and
defensive may not longer properly apply or if they do properly apply they
are entirely divorced from the design aims. To give a specific example, I
think it would be misplaced to describe a vet's humane killer as either
offensive or defensive. In the natural use of language, similarly I cannot
relate either of those terms to the design principles or to the designer's
aims. However, I accept that it is conceivable that a humane killer might be
used in an offensive or defensive action.

> While a pure science will not generally be offensive or defensive, as soon
> as it is applied for some specific (military) purpose the latter will very
> often give the resulting technology an offensive or defensive character.

I don't think so. The application of science to a (military) requirement
results in the creation of a capability. Many (most?) such capabilites may
be used for both defensive and offensive purposes. Again to illustrate with
an extreme example; the Human Wave attacks mounted by the Chinese in
Northern Korea were offensive in purpose (to drive the UN out of Korea)
whilst the use of aircraft and tanks to defeat them was defensive in
purpose.

That said, yes, certain weapons may be predisposed to offensive or to
defensive operations; tanks and long range hunter killer submarines being
examples of weapon systems lending themselves well to offensive operations
(unlike the
Human Wave). However, in the case of cryptography, the item is as purely
'neutral' as any I can imagine. It's purpose is to conceal meaning and/or to
authenticate origination, no more and no less. I cannot relate the word
offensive or defensive meaningfully to either of those purposes. If you
believe that there is some other purpose served by cryptography and that it
is inherently defensive, I would read any such explanation with care and
interest.

> Turning to cryptography, this is not a pure science but rather an applied
> one in which mathematics (for example) is used for the specific purpose of
> defending information.

Forgive me but that seems to stretch the use of language. It may conceal
information but it does not defend it in any actual way. How could one
defend
information with cryptography? No, one conceals information and, in the
process rendering that information useless for offensive or defensive
purposes until after it has been uncloaked be a process of decryption.

> I believe the design intent here is unambiguously
> defensive whereas you believe that this purpose is neither offensive nor
> defensive.

In relation to cryptography, that is my belief.

>Offence is about being proactive and taking the initiative
> whereas defence is essentially passive and reactive.  Cryptography does
not
> take the fight to the enemy, it sits passively waiting to repel any enemy
> onslaught that might be mounted and this, in my view, gives it the
> characteristscs of defence rather than offence.

As outlined above, I am unhappy with this kind of analogy. However, for the
sake of the argument, let's persist with it for one moment. Cryptography
does 'take the fight to the enemy'. Consider the following.

We might agree that the surprise release of nuclear weapons to gain
political or economic advantage would be an unforgivably offensive act.  We
might further agree that we would expect the release codes for those nuclear
weapons to be thoroughly encrypted in transmission. The sole purpose of that
encryption would be to conceal that a nuclear realease order had been given.
There is no intent to 'defend' that information (from what?). The intent is
to
maintain the element of surprise for as long as possible thus assuring the
success of an offensive action that would otherwise probably fail. A
secondary intent in some scenarios could be also to assure deniability for
the act. The cryptography involved remains, as always neutral. Nevertheless
it plays a vital role on 'carrying the fight to the enemy'. No military
communications officer would advise you otherwise, I think.

>From your past life, you are no doubt aware that, in terms of land force
operations, the British consider communications (including cryptography) to
be one of the five combat (teeth) arms. Most other nations relegate military
communications to a combat support role (together with military engineering,
tactical nuclear weapons, fighter ground attack etc). No one, to my
knowledge, makes any attempt to distinguish in between offensive and
defensive combat in these broad assignment of capabilities. Other land force
operational capabilities, such as logistics, policing and administration
fall into te category of 'service support'.

> You also appear to take the view that if an end system is used for offence
> then every sub-component of this system is, by definition, offensive.

Not at all. May I say again that systems are largely neutral in colouration
and, in the case of a cryptographic sub-system of module, it is always so.
Intentions in the use of such a sub-system may be either offensive or
defensive or even both
at once.

>In
> contrast I take the view that an offensive (defensive) system can contain
> sub-components of the opposite character.  Hence if a tank is being used
> offensively you then characterise its armour as offensive whereas I take
the
> view that its armour is a defensive sub-system designed to protect the
> tank's occupants from attack.

You're approach is rather too humane:-) The purpose of armour on a tank is
to enable heavy firepower both to be moved into a selected area and to used
from there without being it being destroyed, though it will probably be
struck several times during that process. In short it is the system's combat
effectiveness and not the men, per se,  that is the object of the
protection. Were it otherwise, one would choose not put the men in harm's
way in the first place.

> Others can judge for themselves which line of thinking better fits their
own
> perceptions of reality.

I think we can agree that it is perceptions rather than any fundamental
truth that we discuss. However, I do believe that that an understanding both
of govt's perception and, perhaps, the perception of the general public are
of importance in influencing policy in the matter of the general use of
cryptography.

Owen Lewis