Third and final day of ILEC. I’m attending a session on Ethics Culture. First up, Marnie Prasad and Mary-Rose Russell, from Auckland University of Technology Law School, on the ‘Professional and ethical challenges for criminal lawyers in the changing environment of legal representation: a New Zealand perspective’. They gave an engaging review of the structure of the profession in NZ. Regulatory framework: Lawyers and Conveyancers Act 2006, Client Care Rules, other legislation, and the NZ Bill of Rights Act. Criminal Procedure Act 2011 (coming into force July 2014) seems to have introduced remarkable changes in the name of (what else) efficiency, including removal of right of trial by jury. Legal aid is also the subject of major reform, fiscally-driven. LA has been available since 1912 in NZ, granted at the discretion of judges, then the Registrars of Criminal Courts.
From the description the small size of the legal profession and the court-orientation of LA are large factors. Interesting to compare NZ and Scotland in that regard. All changed in Legal Services Act 2000, which created a Legal Services Agency responsible for the granting of legal aid and the legal aid list (ie the lawyers), and a ‘preferred lawyer’ policy. Public Defence Service and the preferred lawyer system created what they called ‘a perfect storm — backlash from the PDS, especially the Criminal Bar. Government declared ballooning LA costs unsustainable, the Report of 2009 ‘Transforming the Legal Aid System’ identifying ‘widespread unethical practices’, lawyers ‘gaming the system’, ‘corrupt behaviour worthy of disbarring and more. New Legal Services Act 2011 with preferred lawyer system removed, PDS expanded, and fixed fees introduced. This has raised ethical/professional challenges for criminal legal aid lawyers and increased gaps in access to justice.
Pro bono – not a ‘strong profile’ in NZ. Only two law schools have clinics. Community law centres (CLCs) began in 1978, +/- 1,200 lawyers provide pro bono services to CLCs. There are pb teams in large law firms, largely offered by motivated individual lawyers. Encouraged by Law Society but service not mandated. Government-sponsored self-help assistance to poor litigants. Speakers were concerned that CLC system and pro bono would be leaned on by government in further cost-cutting in the future. In fact the speakers could see that NZ will lose its independent Criminal Bar as the PDS expands, into what speakers in another session described as the spectrum between pro bono, low bono and voluntary work. Interesting overview of a melancholy subject.
Next, (we seem to have lost a speaker…) John Flood on ‘Are machines ethical’, with a single slide that was provocative enough for me to hold back publishing this post until I’d thought more about it.
He cited road trafficrepresentation.com, as a system that combined information seeking algorithms and pay-for representation by lawyers & barristers. Quotes Calebresi; and Virilio — when you invent cars, you invent car crashes, planes, then plane crashes. High-frequency trading markets, then data crashes with potentially catastrophic effects. Google driverless car? Probably not the case of if but when will an accident happen. The question is, how do its algorithms determine car behaviour in the case of a developing accident?
Which brings us, he said, to the trolley problem — the out of control railway trolley, the choice of killing the 10 working on the line, or the 1 on the spur?[1] The more we abdicate our agency to machines to do things for us, the more we lose agency — existential isolation. Turing test — could there be one for the professions? Eg surgery? Robots, he observed, can lie: self learning and self-reproducing robots are possible. In short, should machines be ethical? What about machine courts and machine judges?
At questions Graham Ferris objected to the ascription of agency and self-awareness to machines. I know what he means, but I think that we are well on our way to creating machines that are so (depending of course on the definition of machine, and agent and what constitute awareness. It’s easy to trivialise or sensationalise the topic but it’s critically important for us as a society.
For what John raised in his paper are topics that have a history not just in philosophy of mind, AI, including robotics, artificial life, embodied cognition and cyber-cognition, technology and culture; so here’s my take on the subject. The subject has always fascinated us — Mary Shelley’s creature is a shocking Romantic emblem of divine power ascribed to human agency — a grotesque inversion of Enlightenment hopes for the future and power of the human to take its destiny in its own hands — an inversion, indeed of Percy Bysshe Shelley’s own Promethean poetry, as he must have realised listening to her reading out loud the story[2]. But in that Genevan chateau with PBS, Byron and others where her story was written and first read out to such chilling effect, Mary was drawing on darker figures from ancient literatures – the golem for example. Golem and Prometheus, the Jekyll and Hyde of the digital domain (to quote another byronic Scot, RLS).
It was my friend the poet and translator Peter McCarey who back in the early 90s pointed out the doppelgänger nature of the dilemma we’re in. His long poem The Devil in Driving Mirror, all 48 pages of it, is a brilliant, funny, prophetic lament, philippic, lyric and much else on the fusion of myth and machine.[3] And he too pointed out the link to the Villa Diodati, on the shores of Lake Geneva, noting ironically how linguistic archaism and obsolescence (‘engine’) is recursive like code itself (or to put it another way, is now resurrecting itself, the frankenstein horror all over again) in describing what is happening to us —
But Byron’s daughter, working in the attic all
night over mathematical
equations didn’t spend too much time puzzling
whether dimity or muslin be this season’s thing or last. No slouch,
no couch potato or cabbage, she
was helping Charles Babbage program, like so many bales of grogram,
Boolean cogs and cogitations, guzzling
man-hours. Boss! Replace yon
clerk with informatic tosh. Gosh! Grouch
the Luddite turns his hand
to lyddite, later dynamite.
The social fabric is undone on the analytic engine, that
bureaucracy in a box, which governs
the trajectory of the unremembering shell.
Engine: a mechanical contrivance,
a locomotive, military machine,
an instrument of torture (obsolete)
anything used to effect a purpose
a device, contrivance, wile, a snare (all obsolete)
a person used as a tool (archaic)
(L: ingenium, skill).
Our washing machines now are computers that wash garments; our cars are computers that carry us to places. What matters is what our fast-developing relationship with the machine, the engine, does to the relationship with ourselves and others.[4]
In the domain of cognition, take Andy Clark, for instance, on the subject. Read his brief response to Jerry Fodor’s rather dismissive review of Clark’s book in the LRB, Supersizing the Mind. What he describes there is the merest beginning of the transformation of cognition. And even at this stage of the process, we need to take heed and watch the space closely.
That was one of the messages of the edited book on Affect and Legal Education that Caroline Maughan and I edited, and neuroscience, among many other themes was a key element of that text. And in our edited collection on Beyond Text, The Arts and the Legal Academy, Zen Bankowski, Maks del Mar and I persuaded Andy to contribute to the volume because his work, little to do with legal education on the surface, actually has profound consequences for the way that we think about law. Moral cognition was Clark’s dominant theme in his chapter and, as he points out, ‘[r]easoning and inference can now be reconstructed as processes of pattern completion and pattern extension’. Much of the chapter is given over to exploring the consequences of this understanding of cognition.
For Clark, connectionist approaches reconceive the ‘role of linguistically formulated moral rules and principles’ precisely because they enable collaboration in the exploration of moral conduct through discussion, argument and dialogue: ‘linguistic expression enables a tuning and orchestration of moral response that is vital to moral expertise’. This is true, Clark holds, for the novice as well as the expert (contra Dreyfus and Dreyfus); and his example of pattern-recognition in the game Tetris bears this out at the level of manipulation. However, he goes further, and argues that it is the collaborative aspects of linguistic formulations of moral cognition that affect its existence. As he puts it, any persuasive account of moral cognition and moral expertise ‘must attend to a variety of thus far neglected, communication-specific, higher-order prototype spaces’.
The rest of the volume is in many ways an exploration of those higher-order prototype spaces, and how the Arts (as I show above) can help us understand what is changing in the Law and how it will affect us. But that’s enough from me. John’s paper was fascinating, and the discussion absorbing. I hope he’s planning to publish. There’s a conference could be held on this topic this alone, and if anyone wants to think about planning it with us at ANU once our centre is up & running (soon…) you know where to find me.
- [1]It’s an enduring narrative device of course. I’m working through Dunnett’s Lymond Scottish historical novel series. At the end of Pawn in Frankincense there’s a dramatic chess match played with living people who are killed when they are captured, which results in the hero Francis Crawford having to make the algorithmic moves that result in a win, but not without tragedy — one of the most compelling and painful passages of game play and historical writing I’ve read in a long time.↩
- [2] And to add point, the story’s subtitle is ‘A Modern Prometheus’↩
- [3] Vennel Press, 1995. Peter’s poem reaches globally, from carpet sweatshops to Provence, Glasgow, Rwanda, and is a powerful comment on the global reach of technology. It’s also one of several attempts to re-habilitate the long poem in Scottish literature; and with a focus on digital technologies.↩
- [4]Advanced Braking Systems gives us an interesting case study of behaviour and risk-regulation. Adams provides some indication of the complexity of the approach to the field, for instance in his comments on the effects of antilock braking systems (ABS) on accident statistics:
When introduced, their superiority persuaded many insurance companies to offer discounts for cars with antilock brakes. Most of these discounts have now been withdrawn. The ABS cars were not having fewer accidents, they were having different accidents. Or perhaps they were having fewer accidents, but no fewer fatal accidents; the evidence from various studies is less than conclusive.
Adams, J. (2010) Managing transport risks: what works? Paper for Risk Theory Handbook. Available at: http://www.john-adams.co.uk/papers-reports/, p.15↩