WRITTEN ON January 7th, 2009 BY William Heath AND STORED IN Foundation of Trust, What do we want?
From interview today with Jonathan Evans, head of the UK’s secret internal police force MI5. He must be a very clever man, with enough brain capacity to have quite different things going on at the same time:
The public would not want a society in which the security service monitors them all the time, Evans said.
Well spoken sir. Be aware, with your ANPR cameras everywhere and total intercept plans, you’re already way past the point of diminishing returns on this one.
He also robustly defended the government’s plans to allow MI5 and the other security and intelligence agencies to intercept emails and other communications which may have been sent, or posted on websites.
But….but….but…….Isn’t this what is called Doublethink? The ability to hold two completely incompatible views at the same time, without any apparent sense of contradiction? Remember we’re talking here about plans to retain ALL comms data: all the email and web addresses, phone numbers and mobile phone location data that everyone uses. Total monitoring the whole time, in fact. The very thing the public would not want.
5 Responses to “Doublethink is alive and well at MI5”
Let p be the proposition that the public would not want a society in which the security service monitors them all the time.
Let q be the proposition that the public would want the security service to protect it by monitoring all its communications.
q implies r, the proposition that the public would want a society in which the security service monitors them all the time.
But r is equivalent to not-p. So Mr Evans simultaneously believes that p and not-p. That is doublethink, as defined. It is also anathema to the propositional calculus, one proposition can’t have two truth-values. Sympathy doesn’t come into it any more than it comes into the and- and nand-gates of a processor.
We already have a tacit agreement in society that it is a jolly good thing when the police and the security services use communications data to detect and prevent crime and terrorism. If the interception of communications is done in a warranted way on an exception basis if and only if there is an investigation under way, there is no public outcry.
IMP is a different matter. It requires GCHQ or a private company to store all communication details. Why bother? The telcos and ISPs already have those details. What is the benefit of storing a second copy somewhere else?
If a second copy is stored somewhere else, how will it be useful? Only if GCHQ or whoever can emulate the phone companies’ own systems – they will find themselves trying to run a phone company or ISP business. Difficult. And not their job or their speciality.
It fails the Home Secretary’s SOPCom test. The Safeguards count for nothing. It’s not Proportionate. And it’s not Common sensical.
Do you agree?
Evans’ goal, aside from beaming cuddly reason on the 100th birthday of the firm, was elliptically to reinforce the doublethink that comms data collection isn’t really “monitoring”
The answer to “Why bother? The telcos and ISPs already have those details. What is the benefit of storing a second copy somewhere else?”
…is obviously that they want to mine the comms data.
In days of yore, before System X was installed, telephone taps were highly visible yellow wires that ran across the exchange. Anyone working in or visiting the exchange could see them. This provided a certain safeguard, a sort of uninvoked witness, because anyone could see them if they wanted to.
With System X, the process changed so that an unseen hand in a far-off room could set up monitoring using a keyboard. And with this change a subtle aspect of accountability vanished.
Previously we all (in principle) knew a man who knew the chap who could see what was going on without any restriction; or we could go and look for ourselves. Now we have essentially no control over or opportunity to monitor how intercepts are being used and we are forced to take it on trust that the authorities are acting within their proper remit.
A similar sort of problem occurs with electronic voting, as Roger Penrose has pointed out. If we don’t have any opportunity to watch the process (even we usually we don’t bother) and to audit it (which with electronic systems is impossible without losing anonymity) for ourselves if we so choose, then we can’t be assured that the rules are being followed.
As ever, the problem is that of ensuring the accountability of the authorities.
The answer to “Why bother? The telcos and ISPs already have those details. What is the benefit of storing a second copy somewhere else?”
…is obviously that they want to mine the comms data.
One point behind my question is this. If you take a copy of a file in one place (at an ISP, say) and store it in another place (Cheltenham, say), then you have two copies, with a gap between them. Into that gap flood hundreds of difficulties. Are the two copies the same? Not for long. How long? How often will the IMP data be synchronised? In real time? Once a month? Is IMP any use if it is a month behind reality? How often does restoring from backup files work? Not often enough.
All the earnest endeavour and hot air that go into answering these questions is unnecessary if you just leave the data where it is in the first place, in the custody of people who know how to manage it.
A similar issue was discussed at the Crosby forum on public/private identity management. To some extent, people can be identified by their mobile phone. Wouldn’t it be a good idea if IPS had a directory of all mobile phone numbers? No. The directory would be out of date the instant it was compiled. Far more effective, quicker and cheaper to develop a portal which uses Google-style searches to identify the usual user of a phone by interrogating the telcos’ databases. (That’s what I’ve been suggesting for 6 years.)
This IMP idea makes the same clod-hopping mistake. It will have the same result. Implementation will take decades, cost a fortune and won’t work. The will to fail is built into both projects, the NIS and IMP.
Does the database need to be in one place for mining? No. We have had telephone lines for over a century, the internet for 40 years, distributed databases for 30 years, the web for (I should know this, 20?) years and software agents for … some time. Performance might be improved by having the data all in one place. It might be. Or it might be slowed down, who knows, but it is certainly possible to mine data that is distributed all over the place – Google are quite good at it.
And what about this mining exercise? How do GCHQ or whoever know what to look for? What is a suspicious pattern? Given any suspicious pattern, it is always possible to find some transformation which matches any data to it. How do GCHQ or whoever know which is the right transformation?
What Anon finds obvious, I find mystifying. The glib claim that a certain systems design is obvious might pass master when a 12 year-old consultant is briefing a secretary of state but you’ve got to do better than that when you’re dealing with Ideal Government:
Wibbi they maximised the extent to which they worked with user-held, maintained, integrated and updated data
Doublethink? A very unsympathetic interpretation. Evans was acknowledging that there are pressures pulling government both ways – respecting the public’s privacy and allowing the security services to dig up information that can be used to protect the public. Any democratic society has to find the balance between the two, something that is always more difficult when we’ve got a recent history of people trying to blow us up. It’s clear that Evans’ position responds more to the latter, and there’s a lot of scope to argue with him. But he’s not working in some Orwellian realm. Such criticisms could do with an acknowledgement of the pressures that have sparked the proposals.