Purely a figment of your imagination

What amuses, annoys, concerns or otherwise interests me – Noodlemaz

Cogito ergo sum

6 Comments

These are (finally) my recollections of Westminster Skeptics, 7/6/10.

Biometrics and Identification

Belle de Jour (Dr. Brooke Magnanti) came to speak to us and was welcomed warmly.

Brooke is a scientist and supported herself during her doctoral research by becoming a sex worker in London. She documented her experiences anonymously then revealed her true identity in 2009. Obviously people tend to be surprised to hear this, but challenging such attitudes was (I assume) one of her aims.

She came to WSitP to talk not about those adventures but about the science, psychology and challenges of personal identification.

Human Identification in Forensic Science

We have a desire for authenticity, yet want privacy for ourselves – we protest ID cards and CCTV, put up net curtains in our houses – but gossip magazines fly off the shelves (though I wish they didn’t) and rumour mills never fail to be on overdrive.

Why do we want to know about others? That information is often irrelevant; trivial, social things. Here we’re talking the government and personal info.

Identity is how we define ourselves.

‘Unique’ identifiers have been used for a long time. The Bertillon system used a number of them to keep track of who’d been in prison. However, one case destroyed its perceived reliability.

Kansas jailed Will West (a murderer) but in 1903 one William West was sentenced; they looked almost exactly the same and indeed their Bertillon measurements differed only minutely. This discredited the system as it was and led to the use of fingerprinting.

How unique?

So how many people were compared to establish that fingerprints are unique to every individual?

130!

There have been attempts to write image analysis software but computer error makes it unacceptable in court. So it’s down to people, but human fallibility is ever-present.

“Why worry if you have nothing to hide?”

Sometimes innocent people get caught out by imperfect systems (Brooke gives Shirley McKie as an example). Innocent until proven guilty is (or is meant to be) the foundation of the legal system – we ought to be aware of ID measures in place, the possibilities of wrongful accusation and what to do about it.

British citizens don’t need to carry ID cards but everyone else does. Yet no card-readers actually operate in this country. The card contains data from passport, visa, fingerprints. The Government has vowed to scrap ID cards but what about all the information already on file?

Brooke managed to lose everything in her card application (at a bus stop!) – all she had to do to prove her identity was to go to the US embassy with another US citizen vouching for her. So the cards seem fairly pointless, all in all.

What about DNA?

The National DNA Database does not store all 3 billion base pairs (‘letters’) of individual genomes. You can’t store the full sequence; it’s too expensive, time-consuming and generally unacceptable. It’s an issue of privacy.

We don’t want the government knowing more about ourselves than we do.

Instead it uses 20 ‘short tandem repeats’ – relatively small lengths of DNA that are made up of repeat sequences that vary from person to person.

How acceptable is biometrics?

There is no perfect system. Considerations include:

Universality (can it be applied to everyone?)

Uniqueness

Permanence (can it change with time? E.g. retinal scans)

Collectability (how easy is it to access and record?)

Performance (how reproducible?)

Acceptability (likelihood of consent)

Circumvention (ease of avoidance e.g. US embassy incident)

We identify ourselves in broad terms. Race is not an official biological category but we still use it! Still people are assuming that populations don’t and can’t mix. We know there’s more intra -than inter-race variation. It’s scientific fact.

People have moved on to the more ‘PC’ term “ethnicity” but this also suffers from social stereotyping.

ID is currently dependent on what people believe

Me ‘n’ Jorge Cham, creator of PhD comics!

Brooke showed one of her favourite PhD comics – comparing some common perceptions of science to the unfortunate reality! Time to drop in another me-and-someone-cool photo methinks…

privacy.com

Online, people actually tend to be truthful (despite the oft-excessive scaremongering regarding the interwebnet). People still seek trust and authenticity – just the same as irl (in real life, for those who may not know)!

Are our current problems and fears simply ‘growing pains’ like the printing press experienced – like every other technological development?

The web is the first multi-directional medium. We talk back.

I’ve made this point before, regarding Christina Odone’s indignation at people calling her out on her BS.

Personae for Sale

Where there are personal data, there’s business to be had. Tweets/facebook profiles and data for sale – advertising companies have a wealth of information available to them now.

Are we the summation of our entire history? Or do we take each moment as it comes? Is it possible to do that and still be sociable?

On the internet, nobody knows you’re a dog!

Questions

Did anyone realise you were American? Revealing where you learn English

A/ I asked an English friend how to say things. People move; is where you are more important than where you were?

Do we know the maths on DNA ID?  We’ve gone from 16 reference points to 20 now. What’s the chance of a false positive match?

A/ No one’s bothere to check?

Audience: There are estimates! 1:several thousand (that’s really quite high!). Lots of confusion, probably no better than fingerprints.

Martin Robbins: Why come back to science?

A/ Between submitting my thesis and the viva, I needed cash! If I’d wanted to be a writer I’d have wanted to be  like Simon Singh.

What’s the usefulness of DNA fingerprints in court? How are lawyers defending it?

A/ I’m not a criminal lawyer! Statistical assumptions –> invalid model; they’re certifying people to appear as experts on ID in court.

Do you think we should give up? What’s the message?

A/ Confusion! Science, epidemiology, availability of records (histories) – People should think about it more.

Science isn’t necessarily about the results, it’s more about the process

[My boss (and all supervisors, in fact) certainly wouldn’t agree with that!]

Evan Harris: Prosecutors often use fallacious stats

1/1,000,000,000 chance of a match does not equal 999,999:1 guilty odds

Fallibility impacts on the ethics of a database; even if the whole DNA sequence was there, epigenetics brings another level of complexity. [This is the modifications to the basic DNA code that also contain information and can affect our phenotypes – physical manifestations of genetic instructions.]

The example of imprinting disorders was given, specifically Prader-Willi and its maternal equivalent Angelman syndrome (which Brooke could not remember and asked the audience, but when I finally remembered and shouted it out, did not shout loud enough! Oh well. I haven’t forgotten *everything* from uni… it’s in there somewhere).

Brooke asked: how many have at some point fabricated their ID online? Barely any hands went up.

You can’t really stray from what you are

[Here I think of Big Brother and how long people can really keep up some act in front of the cameras before they’re forced to forget it and start being themselves].

How big a problem is DNA contamination?

A/ It’s CSI fiction! I would like to think that people working on a case know what they’re doing (i.e. actually tying their hair up).

What do we have in common with our childhood selves??

A/ Good question; genetics vs. personality, nature/nurture etc. Can we quantify personality (religions, philosophy)?

Is Biometrics related to defense? Is it the fastest-growing industry?

A/ It’s well-funded. Pharma growing fast though. Popularity of crap TV contributes!

——–

Also in attendence were some high-profile ‘bad law’ victims:

Paul Chambers, who was prosecuted for jokingly tweeting that he’d blow up the airport if it didn’t re-open (thus preventing him from visiting his girlfriend). His experience is now infamously known as the Twitter Bomb Hoax Trial.

– Harvey Singh, who has endured a two year libel case brought by a ‘Saint’, who has never even been to the UK.

Dave Osler, who was sued for writing a blogpost.

You can listen to the full talk here on the PodDelusion! Brooke has also made the slides available here.

Advertisements

Author: noodlemaz

I prefer to think of myself as a realist rather than a pessimist, but perhaps that's just optimistic. Honest, atheist, scientist, feminist.

6 thoughts on “Cogito ergo sum

  1. There’s no such thing as computer error (well, except cosmic rays, and there’s a vanishly small probability they would lead to wrongful conviction).

    I think Brooke was talking about classification accuracy, specifically false positives, which is a measurable statistic and a much more sophisticated concept than ‘sometimes computers just get the wrong answer’.

  2. Wow Noodlemaz! No pic of you with Brooke?? Your high standards are slipping!

    I might be in a minority of one, but I’m actually *in favour* of the routine collection of DNA profiles on everyone. I believe that if potential criminals knew that their DNA was already held on a database, it would make them think twice. They would know that if they were to commit a crime, particularly a violent, sexual crime, there would be a much higher chance of them being caught – and quickly.

    Certainly, extra safeguards would have to be introduced to prevent law-officers “lifting” DNA to frame innocent people, but they would have more time to implement these safeguards because they would be concentrating their efforts on fewer serious crimes.

    Imagine a world where many of the potential serial thugs and rapists could be identified within hours of their first victim. Ian Huntley would never have got the caretaker’s job due to his “previous”, Steven Wright might have only struck once, Peter Sutcliffe might never have killed so many…

    Regarding the “rights” of individuals not to have their DNA on the database, what about the rights of people not to be the victims of crime? I’d be happy to be on the DNA database if it meant increased safety for me, my family and friends, for vulnerable children, less violent crime against women, and fewer burglaries.

    And in an age where false passports, driving licences and birth certificates are now easily available on-line, DNA has to be the way forward in establishing the identities and movements of terrorists.

    It won’t prevent all crimes, particularly impulsive crimes due to rage, alcohol or drugs, but even these will certainly have their clear-up rates dramatically improved.

    And on a completely different track, there’s the advantage of quickly assessing potential organ/marrow donors using the database.

    Can someone tell me I’m wrong about this?

    Ok guys – rant over!

    [PS: Excellent article Noodlemaz!]

    • Too kind, Richard, too kind!! 🙂

      I agree with your sentiment in principle. But in practice I really think it’s more difficult and likely sinister.

      Having such a database somewhat undermines the ‘innocent until proven guilty’ premise on which the justice system is based.

      Also depending on what kind of data is stored, there are serious privacy issues. Our DNA says a fair bit about us, and in the wrong hands such information can be misused to devastating effect. Would you really trust someone to look after such details about yourself?

      Also one of the main point s of Brooke’s talk is that ‘unique identifiers’ do not really exist, as far as we can tell. There’s always a chance someone will be matched without having done anything wrong, and that is the worst kind of failing in the justice system. DNA is still fallible, especially at the level we’re using now.

      Preventing crime starts to move into the scary Minority Report-type territory.
      Did you read about Russia’s recent bill that allows screening of personal data to identify individuals who should be arrested based on a predicted *possibility* that they might commit a crime?!

      We don’t want to be going in that direction, I don’t think :/

  3. Thanks for this reply, Noodlemaz. I appreciate you have lots of fingers in lots of pies, so I’ll try to keep this reply short!

    I think the “innocent til proven guilty” idea is admirable, but in many areas of UK law it has already gone. For example, anyone who wants to work in a school, care home or hospital ward now has to prove they are “innocent” by obtaining a clean CRB check. These checks can be handled by several layers of admin, and quite sensitive information about an applicant can (wrongly) enter into the public domain.

    Regarding confidentiality about DNA data, and its appropriate use, some of the millions of pounds saved by scaling back the police force and justice system (remember the huge drop in murders, violence, rapes, burglaries etc.?) can be spent on setting up a double-blind system of DNA data storage & retrieval, so that very few personnel are actually aware of whose DNA details they are accessing or processing.

    Regarding the “unique identifiers”, this doesn’t have to present a problem if a requirement is made that further corroboratory evidence is needed for a prosecution. A murderer arrested hours after a killing will have had little time to destroy evidence or create an alibi, and in the future may even find the police waiting for him when he returns home.

    The odds on an entirely innocent person having the identical DNA to the murderer, and having a plausible motive and opportunity to commit the murder are about the same as the One True God being discovered living on Job-seekers Allowance in a bedsit in Droitwich; i.e., not very likely! 🙂

  4. … and yes, I agree with you that trying to prevent crimes before they happen is a bit scary! But let’s not let that deflect us from catching those who have already perpetrated heinous crimes.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s