RSS Feed

“I’m afraid I can’t do that Dave, as a result of subsection 9(b) (iv) (a)”


Another little thought experiment, on Judges this time. Now, clearly Judges at first instance have to very carefully assess the evidence, and the nuances and tone and demeanour of witnesses and attribute weight and balance to a variety of factors. The higher up the Court hierarchy you go, the less important that becomes, to the point where by the time of the Supreme Court, what a witness said or did not say in the box is almost neither here nor there.


In fact, what is very often happening in the Supreme Court is drawing together from a variety of sources – the legislation, the guidance and existing principles derived from authority, applying it to the legal dilemma in the case and achieving a decision that ends up being consistent with all of those decisions and perhaps extending the existing principles in a slightly new (yet consistent) direction.


Now, it occurs to me that as we reach the point of artificial intelligence, it would be theoretically possible to have all the legislation, all of the guidance, all of the previous authorities, held by a computerised mind, who could then just trace a path through them to reach the decision.  If all that one is doing is looking at the precedent decisions and seeing where they would logically take you in deciding the legal dilemma, a sufficiently wonderful computer can solve that logical problem.


Instead of seven law lords, what one would have is a dazzlingly brilliant super computer  S.U.P.R.E.M.E   and the legal dilemma would be inputted and a judgment would come out.

 If you’re like me, then you are probably shifting a little nervously in your seat at this point, and feeling that this is just uncomfortable. There’s more to Supreme Court decisions than just deriving the answer from the principles.

 But that in itself takes us into interesting areas.

 There seems to me to be some sort of qualitative difference between these two questions :-


(a)   What is 53,209 divided by 7.33 




(b)   Is a school’s admission policy to give preference to Orthodox Jewish children, looking for evidence that the mother of the child is Jewish by birth or by Orthodox conversion, discriminatory under the Race Relations Act 1976


It is not simply a matter of complexity – of course the latter question is more complex, since one can solve the first question in a matter of seconds with a calculator, and the second at the very least is going to involve reading the Race Relations Act and the school’s policy, and any decisions that help clarify how either ought to be interpreted.


But there’s still more to it than just complexity, otherwise S.U.P.R.E.M.E could answer both, given the right information (or access to Google to find it for itself)


Isn’t one of the differences between the two questions that the answer to (a) exists already – it is out there to be found, and it is utterly replicable. Anyone who does the calculations will arrive at the same answer, regardless of who they are.   (The same would be true if you swapped (a) for “What is the capital of Guina Bissau?”  – it is a factual question, and the answer is out there to be found)


The answer to (b) – maybe it doesn’t exist   (well, it does now, because the Supreme Court decided it  in R (on the application of E) v JFS : Governing Body 2009 UKSC 15) and it only exists ONCE the question is asked and answered. The answer is CREATED, not found.

 If question (a) is more like maths, or physics or geography – there’s a factual answer that is true for whoever answers it, then maybe (b) is more like history or english literature – there are certain things within it  “In which Shakespeare play does Ophelia appear?” which are definitive, factual answers (like cases which squarely correspond with precedent) and there are others which the person answering the question creates  “What is the nature of the character Hamlet?”


Once we start thinking in those terms though, we inevitably bump into the peculiar wrinkle that the highest legal authorities in the land, which bind future courts and cases and will in turn influence future legal authorities at the highest level are not OBJECTIVE TRUTHS found by the Court, but SUBJECTIVE decisions created by the Court.  And in turn that the Judges who sit on these cases bring something of themselves to that process; that’s why the concept of S.U.P.R.E.M.E deciding it makes us feel a bit edgy.


That must intrinsically be right, because all of the Judges in the Supreme Court hear the same arguments, have the same facts, have access to the same precedent authorities. Yet there are dissenting judgments. So what causes that must be that there’s an element, even in the rarefied air of the Supreme Court, of subjectivity to deciding how the legal dilemma should be resolved.  


[You may recall my previous blog about the impact of extraneous circumstances, such as proximity to lunchtime on judicial decisions   ]


This piece derives from another interesting piece of research, which takes as an example the  UK Supreme Court’s decision in the Jewish Free School’s admission policy.

 My attention was recently drawn to this study, which is available in the Journal of Law and Society  (I’m afraid that it is behind a paywall, and as such I can’t let you peep at it, and I have to be limited in how much I can quote out of it)


The study was written by Rachel J Cahill-O’Callaghan of CardiffUniversityLawSchool  (which frankly is becoming a hot-bed of brainy talent, and one day I must try to visit and have my mind blown. Perhaps during the Six Nations…)

 It opens with a lovely quotation from Lord Reid, in which he manages in four lines to say everything I’ve been fumbling in the dark for, and does so beautifully to boot.

 “Those with a taste for fairy tales seem to have thought that in some Aladdin’s cave there is hidden the Common Law in all its splendour and that on a judge’s appointment there descends on him knowledge of the magic words Open Sesame… But we do not believe in fairy tales anymore”


 There has been a long debate about the make-up of the Supreme Court and whether it reflects the diversity of our society (hint, no, it doesn’t) but this research goes further than that, and analyses how a person’s position on Values is brought to bear on judgments and decisions.


“In reaching a decision, at least one not governed by precedent, a judge will support one or more values above another…. Although the law provides the basis for framing and constraining judicial discretion, in difficult cases at least, it is the personal values of an individual judge that influences how that judicial discretion is exercised and that, in turn, can influence the way in which the law develops”


The way that the research tests this is interesting, and involves  firstly identifying a series of values and defining those so that one knows exactly what to look for in relation to each value, then looking at the judgment in the authority, and analysing each line of it, looking at the emphasis that the individual judge places on particular values, which may compete (for example “flexibility in the law” v “Transparency in the law, Corporate responsibility v individual responsibility, freedom v responsibility).  


On a case such as this, where the Supreme Court was divided in its opinion, that analysis can then be used to see whether the judges who reached a particular conclusion (there was a breach of Race Relations Act) appeared to place similar emphasis on similar values, and do the same exercise with those who reached the opposite conclusion.


On looking at that, there are really stark differences between the values emphasised by the majority judgments and the minority judgments.


The author of the report acknowledges that what was not possible was to go back to the individual Supreme Court judges after the judgment, with that analysis, and ask them whether the analysis of the values that each judge “appeared” to prioritise accords with their own view of where their own values sit, but the research uses some clever techniques to try to fill in this gap.


 There’s an interesting conclusion that if personal values play a role in judicial decision-making and the framing of the law, then in order for the Supreme Court to represent society one doesn’t merely have to look at the very visible aspects of diversity  (gender, ethnicity, socio-economic background) but also diversity of personal values.

About suesspiciousminds

Law geek, local authority care hack, fascinated by words and quirky information; deeply committed to cheesecake and beer.

10 responses

  1. The thing is, that we want people to be better at assessing evidence and making decisions. It’s very hard to persuade people that a computer can do better and we like to think that there is an indefineable quality that humans contribute that computers cannot – and never will – replicate. However, there are examples of computers doing better than people, against the odds or against the preconceptions, which gradually become accepted as being actions that computers do better.

    For example, we hope that going to see a Dr will be better than accessing a computer because Dr Finlay knows us and the wisdom of years of seeing patients will give him intuition that a computer will never have. But computers are better than people:

    Care proceedings are increasingly becoming commodities like soap powder: you have a set process and a set timescale with set orders and, shortly, as set body of evidence. Few cases are highly unusual and the case management orders code system is clearly skewed to provide set outcomes, or at least to allocate blame in a prescribed form.

    I think it would be possible to set out a Re B-S analysis with set weightings for different, set criteria and this would produce an inevitable outcome. In fact there are many cases where it’s clear what the outcome is, because we’ve all done a similar analysis.

    “I am a name, not a number….”?

    Once you start to think about it, there are lots of activities that are now done quickly and more cheaply by computers that once used to be regarded as things that only humans could do. The general thrust of most things seems to be that they start off requiring great skill and expertise and progress towards standardisation and being done by computers. Inevitably one resists this idea.

    • Lovely comment Norma. You may well be right that the problem with S.U.P.R.E.M.E is not that IT is flawed, but that it jibes with our current belief that there is something ineffable about the human process. If all one is doing is analysing the wording of legislation and precedent, and fitting the case into precedent, then why shouldn’t a computer do it? (the answer probably is that we suspect Judges actually have an idea of what they THINK the right answer is, and try as hard as they are able to make the right outcome fit with precedent – Denning was a particularly great example of that approach. The computer would lack that sense.)

      I think “to allocate blame in a prescribed form” is spot on. There’s not really a space in the form for “The case simply needed more time to reach the fair and just conclusion”

      As you become more and more restrictive in the process – standardising and streamlining and checklisting and firm criteria, the process does become more and more susceptible to automation. Perhaps that removes the vagaries of chance, or reduces them, perhaps it reduces the flexibility to respond to unpredictable situations and adjust ones view of the situation. Maybe in twenty years time, there will be robot judges who won’t care about what I am saying to them, but will be analysing my voice-stress to immediately detect whether I am making a valid point or saying something to make the best of a bad job.

      • I agree about Denning but in a sense, and with some regret, I think that proves the point. Because what the law is required to provide is certainty, not reasoning backwards to support an emotionally desired outcome. Denning was a hero when I was taking my solrs’ exams, but someone commented to me that few of his ground breaking points cases remain good law.

        Actually I was wondering the other day what has happened to equitable principles now we have the HRA. Isn’t it interesting that we are told that we all signed up to the ECHR in the 50’s but in truth it took the Act to make the articles an everyday feature of cases. Same with sex discrimination – I remember elderly judges saying “I suppose I’m not allowed to make that sort of comment now” Well, no, Your Honour, and in truth you never should have done.

        And the fuss about female barristers wearing trousers! I seem to have diverted myself…

        I think we have to recognise what really can be standardised and what can’t, and the latter category gets smaller. Perhaps as experiences lawyers, our contribution is in both recognising the difference and in crafting the systems, rather than in administering them. [ note to civil service drafters of the CMO – if the case heading sets out the name of the child you don’t repeat it nx3million in the body of the order etc etc]

        I have come across research that shows that as we get older we use intuition more and analysis less – is that fair? Does that produce certainty and justice? as we age don’t we become more used to confirming our prejudices? Isn’t the point of codes and statutes to reduce litigation?

      • Yes, I too have had those mixed feelings about Denning – on the one hand, we want Judges not to be shackled by the law at the expense of a fair outcome, but on the other the law needs to be predictable and foreseeable if it is to be useful and respected.

        I think your HRA point is interesting – it occurs to me that we have had a stock phrase of judicial window-dressing for 15 years on article 8, and it is only in a single paragraph from Lord Neuberger in Re B that the actual import of the precise meaning of the words in that stock phrase become potentially charged with very significant import.

        (I really wish anyone involved in drafting the standardised orders would actually speak to anyone involved in the process – they always end up being cumbersome and unwieldy and useless when it comes to finding in a moment the answer to “when did so and so happen?”, they have lost the valuable preambles, and frankly, they are “front of house” in the new PLO. The utter uselessness of the last standardised PLO orders poisoned the well of the last PLO, and led to it becoming a “pick and mix” of what people wanted from it, ending up with it being watered down to CMC, IRH, aim for 40 weeks. )

        Your final point about age and intuition is potentially significant given the one absolute constant about Supreme Court Judges is their ‘experience’

  2. Doctor v computer diagnosis. Yes the doctor could know the patient but could also have formed wrong opinions of that patient. i.e. Yes, Mrs Smith you have palpitations and tend to be a little anxious. Mrs Smith returns to the doctors yet again and says ‘But Doctor I have chest pains too’ Dr says, never mind Mrs Smith, Ill give you some tablets to prevent your anxiety. Mrs Smith goes home.
    2 hours later Mrs Smith was in hospital with a life threatening heart attack.
    Behind all this is the answers that a computer could do better on factual,
    But the family court also makes judgements on preceived to be rather than anything factual just as the case of a doctor.
    And the dynamics of a family court and its restrictions and timetable do not allow the court to reach the correct decision in the best interests of the child. It also weighs heavily and often wrongly on the viewpoint of non factual evidence and mere opinions of social workers and pyschological reports brought in by LAs.
    So where in these muddied waters can any decision be reached in the best interests of the child espcially with a 26 week restriction.

    The answer can only be that it is not a computer that is needed or 26wk restrictions. Even a computer could answer that one or blow itself up in the process.
    The answer is as logical as a smart computer. Open family and associated courts to transparancy, remove immunity from prosecution from ALL giving statements into court, make prosecution for contempt and perjury a reality and prevent muddied waters (trolls and bugs).

    Perhaps then the court will and can work to a 26wk schedule without making it a rule of court.

  3. But then there is not so much money to be made. Just as in the following case. Many years ago it was proven that the radiation dose from a computer screen could cause the malformation of a feotus under (under 3months in womb) Doctors knew about it and reported it to the MOD.
    The result was it was swept under the carpet and no-one was prepared to make it public knowledge (Gov.MOD decided it was not in the public interest)
    After all it could cause the big computer companies financial loss.
    Now that is what happens in the family courts.The childs best interests can never be served as they come second to money.

  4. Money over a child again.
    In certain hospitals body parts were removed from living babies born from uumm, lower class familes for the rich. It hit a newspaper but the story was quickly covered up. Again not in public interest by Gov.cover it up

  5. So to remove ‘having to meet the public interest clause in the cps rules of court’ would also be the greatest move in child protection in history.
    In other words where it involves children remove it and open it to public scrunity.

  6. Pingback: “I’m afraid I can’t do that D...

  7. Two well known prayers:

    1. God, grant me the serenity to accept the things I cannot change,

    The courage to change the things I can,

    And wisdom to know the difference.

    2. Lord, You know better than I know myself that I am growing older and will someday be old. Keep me from the fatal habit of thinking I must say something on every subject and on every occassion. Release me from the craving to straighten out everybody’s affairs. Make me thoughtful, but not moody. Helpful, but not bossy. With my vast store of wisdom, it seems a pity not to use it all, but You know, Lord, I want a few friends at the end.

    Keep my mind free from the endless recital of details; give me wings to get to the point. Seal my lips on my aches and pains. They are increasing, and love of rehearsing them is becoming sweeter as the years go by. I dare not ask for grace enough to enjoy the tales of others’ pains, but help me to endure them with patience.

    I dare not ask for improved memory, but for a growing humility and a lessening cocksureness when my memory seems to clash with the memories of others. Teach me the glorious lesson that occassionally, I may be mistaken.

    Keep me reasonably sweet. I do not want to be a saint – some of them are so hard to live with. But a sour old person is one of the crowning works of the devil. Give me the ability to see good things in unexpected places, and talents in unexpected people. And give me, Lord, the grace to tell them so.

%d bloggers like this: