Mainstream
medicine
would simply adopt them.
Richard-Dawkins-The-Devil-s-Chaplain
uk/gmdebate/Story/0,2763,372528,00.
html.
28
? not literally identical to the financial equivalent? Suppose it is sufficiently similar that the main routine can indeed be borrowed, but it still needs tweaking in detail. In that case, it is possible that the rocket could misfire if we naively import the subroutine raw. Switching back to biology, although genes really are watertight subroutines of digital software, they are not watertight in their effects on the development of the organism, for here they interact with their environment, including importantly the environment furnished by other genes. The antifreeze gene might depend, for optimal effect, on an interaction with other genes in the fish. Plonk it down in the foreign genetic climate of a tomato, and it might not work properly unless tweaked (which can be done) to mesh with the existing tomato genes.
What this means is that there is a case to be made on both sides of the argument, and we need to exercise subtle judgement. The genetic engineers are right that we can save time and trouble by climbing on the back of the millions of years of R & D that Darwinian natural selection has put into developing biological antifreeze (or whatever we are seeking). But the doomsayers would also have a point if they softened their stance from emotional gut rejection to a rational plea for rigorous safety testing. No reputable scientist would oppose such a plea. It is rightly routine for all new products, not just genetically engineered ones.
A largely unrecognized danger of the obsessive hysteria surrounding genetically modified foods is crying wolf. I fear that, if the green move- ment's high-amplitude warnings over GMOs turn out to be empty, people will be dangerously disinclined to listen to other and more serious warnings. The evolution of antibiotic resistance among bacteria is a vicious wolf of proven danger. Yet the menacing footfalls of this certain peril are all but drowned out in the caterwauling shrieks over genetically modified foods, whose dangers are speculative at most. To be more precise, genetic modification, like any other kind of modification, is good if you modify in a good direction, bad if you modify in a bad direction. Like domestic breeding, and like natural selection itself, the trick is to introduce the right new DNA software. The realization that software is all it is, written in exactly the same language as the organism's 'own' DNA, should go a long way towards dispelling the gut fears that rule most discussions of GMOs.
I can't leave the subject of gut feelings without a favourite quote from the lamented Carl Sagan. When asked a futurological question, he said that not enough was known to answer it. The questioner pressed him on what he really thought. 'What is your gut feeling? ' Sagan's reply is
SCIENCE, GENETICS A N D ETHICS
29
? SCIENCE AND SENSIBILITY
immortal: 'But I try not to think with my gut. ' Gut thinking is one of the main problems we have to contend with in public attitudes to science. I shall return to the point under Ethics. Meanwhile, some more remarks on the future of genetics in the twenty-first century, especially in the wake of the Human Genome Project (HGP).
The HGP, which will be completed any time now, is really a twentieth-century accomplishment. It is an outstanding success story, but it has limited scope. We have taken the human hard disk and transcribed every jot and tittle of the 11000101000010000111-style bits of information on it, regardless of what they mean in the software as a whole. The HGP needs to be followed up by a twenty-first-century Human Embryology Project (HEP) which, in effect, deciphers all the high-level software routines in which the machine-code instructions are embedded. An easier task will be a series of genome projects for different species (like the Arabidopsis plant genome project, whose completion is announced on the day that I write). These would be quicker and easier than the HGP, not because the other genomes are smaller or simpler than ours, but because the collective expertise of scientists increases cumulatively and rapidly with experience.
There is a frustrating aspect of this cumulative improvement. Given the rate of technological advance, with hindsight, when we started the Human Genome Project it wasn't worth starting. It would have been better to do nothing until the last two years and start then! Indeed, that is pretty much what the rival firm of Dr Craig Venter did. The fallacy in the 'never bother to start' maxim is that later technologies cannot get into a position to 'overtake' without the experience gained in developing the earlier ones. *
The HGP implicitly plays down the differences between individuals. But, with the intriguing exception of identical twins, everybody's genome is unique, and you might wonder whose genome is being sequenced in the HGP. Has some dignitary been singled out for the honour, is it a random person pulled off the street, or even an anonymous clone of cells in a tissue culture lab? It makes a difference. I have brown eyes while you have blue. I can't curl my tongue into a tube, whereas it's 50/50 that you can. Which version of the tongue-curling gene makes it into the published Human Genome? Which is the canonical eye colour? The answer is that, for the few 'letters' of the DNA text that vary, the canonical genome is the majority 'vote' among a sample of
*I have discussed the implications of the rapid growth of our understanding of genetics in more detail in 'Son of Moore's Law' (see pp. 107-15).
30
? people carefully chosen to give a good spread of human diversity. But the diversity itself is expunged from the record.
By contrast the Human Genome Diversity Project (HGDP), now under way, builds on the foundation of the HGP but focuses on those relatively few nucleotide sites that vary from person to person, and from group to group. Incidentally, a surprisingly small proportion of that variance consists of between-race variance, a fact that has sadly failed to reassure spokesmen for various ethnic groups, especially in America. They have dreamed up influential political objections to the project which they see as exploitative and tarred with the brush of eugenics.
The medical benefits of studying human variation could be immense. Hitherto, almost all medical prescribing has assumed that patients are pretty much the same, and that every disease has an optimal recom- mended cure. Doctors of tomorrow will be more like vets in this respect. Doctors have only one species of patient, but in future they will subdivide that species by genotype, as a vet subdivides his patients by species. For the special needs of blood transfusions, doctors already recognize a few genetic typings (OAB, Rh) etc. In the future, every patient's personal record will include the results of numerous genetic tests: not their entire genome (that will be too expensive for the foreseeable future) but, as the century goes on, an increasing sampling of the variable regions of the genome, and far more than the present 'blood group' typings. The point is that for some diseases there may be as many different optimal treatments as there are different genotypes at a locus - more even, because genetic loci may interact to affect susceptibility to disease.
Another important use of the genetics of human diversity is forensic. Precisely because DNA is digital like computer bytes, genetic finger- printing is potentially many many orders of magnitude more accurate and reliable than any other means of individual identification, including direct facial recognition (despite the unshakeable gut feeling of jurors that eyewitness identification trumps everything). Moreover, identity can be established from a tiny trace of blood, sweat or tears (or spit, semen or hairs).
DNA evidence is widely regarded as controversial, and I need to say a little about why. Firstly, human error can obviously vitiate the accuracy of the method. But that is true of all evidence. Courts are already accustomed to taking precautions to avoid the muddling up of specimens, and such precautions now become even more important. DNA finger- printing can establish, almost infinitely far beyond all reasonable doubt, whether a smear of blood came from a particular individual. But
SCIENCE, GENETICS AND ETHICS
31
? SCIENCE AND SENSIBILITY
obviously you must test the right smear.
Secondly, astronomical though the odds against mistaken identity by
DNA fingerprinting theoretically are, it is possible for geneticists and statisticians to come up with what seem like widely different estimates
19
of the precise odds. I quote from Unweaving the Rainbow (Chapter 5,
which is devoted to explain DNA fingerprinting in lay terms).
Lawyers are accustomed to pouncing when expert witnesses seem to disagree. If two geneticists are summoned to the stand and are asked to estimate the probability of a misidentification with DNA evidence, the first may say 1,000,000 to one while the second may say only 100,000 to one. Pounce. 'Aha! AHA! The experts disagree! Ladies and gentlemen of the jury, what confidence can we place in a scientific method, if the experts themselves can't get within a factor of ten of one another? Obviously the only thing to do is throw the entire evidence out, lock, stock and barrel. '
But. . . any disagreement. . . is only over whether the odds against a wrongful identification are hyper-mega-astronomical, or just plain astronomical. The odds cannot normally be lower than thousands to one, and they may well be up in the billions. Even on the most conservative estimate, the odds against wrongful identification are hugely greater than they are in an ordinary identity parade. 'M'lud, an identity parade of only 20 men is grossly unfair on my client. I demand a line-up of at least a million men! '
The idea of a nationwide database, in which all citizens' DNA finger- prints would be held, is now being discussed (only a sample of genes, of course: doing the whole genome would be overkill, far too expensive). I don't see this as a sinister, Big Brotherish idea (and I have written to my doctor volunteering to be a guinea pig in the pilot study of 500,000 now being prepared). But there are potential problems, of a civil liberties character. If your house is burgled, the police will routinely look for (traditional, old-fashioned) fingerprints of the burglar. They need to fingerprint the householder's family too, for elimination purposes, and most people are happy to oblige. Obviously the same principle will apply to DNA fingerprinting, but many people would want to stop well short of a nationwide database. Presumably they would also object to a nationwide database of conventional, old-fashioned fingerprints, but perhaps that is not a practical issue because it would take too long to search through it for a match. DNA fingerprinting doesn't suffer from this difficulty. Computer searches of huge DNA databases could be accomplished swiftly.
What, then, are the civil liberties problems? Surely, those with nothing to hide will have nothing to fear? Perhaps not, but some
32
? people do have legitimate reasons to hide information, not from the law but from each other. A surprisingly large number of people, of all ages, are genetically unrelated to the man they think is their father. To put it mildly, it is not clear that to disillusion them, with conclusive DNA evidence, would increase the sum of human happiness. If a national DNA database were in place, it might be hard to control unauthorized access to it. If a tabloid newspaper were to discover that the official heir to a Dukedom was actually sired by the gamekeeper, the consternation in the College of Heralds might be mildly amusing. But in the population at large it doesn't take much to imagine the family recriminations and sheer private misery that could flow from freely available information of true paternity. Nevertheless, the existence of a national DNA database wouldn't alter the situation much. It is already perfectly feasible for a jealous husband, say, to take a saliva or blood sample from one of his supposed children and compare it with his own, in order to confirm his suspicion that he is not the real father. What the national database could add is a swift computer search to find out who, out of all the males in the entire country, is\
More generally, the study of human diversity is one of very few areas where a good (though in my opinion not overwhelming) case can be made against the pure disinterested search for knowledge: one of very few areas where we might actually be better off ignorant. It is possible that, by the end of the twenty-first century, doctors will be able accurately to predict the manner and time of death of everybody, from the day they are conceived. At present this kind of deterministic prognostica- tion can be achieved only for possessors of genes such as Huntington's Chorea. * For the rest of us, all that is possible is the vague statistical forecast of the life insurance actuary, based on our smoking and drink- ing habits, and a quick listen through a stethoscope. The whole life insurance business depends upon such forecasts being vague and statistical. Those who die old subsidize (the heirs of) those who die young. If the day comes when deterministic forecasting (along Huntington's Chorea lines) becomes universal, life insurance as we know it will collapse. That problem is soluble (presumably by universal compulsory life insurance with no individual medical risk assessment). What will be
*The folk singer Woody Guthrie died of Huntington's Chorea, a horrible disease that waits till early middle age before killing you. It's a dominant gene, so each of Woody's children knows that he has an exactly 50 per cent chance of suffering the same horrible fate. Some people, given these odds, prefer not to be tested. They'd rather not know until they have to. IVF doctors can now push the test back to the newly fertilized zygote, and choose to implant only those that lack the fatal gene. This is obviously a huge boon, but it is attacked by ignorant lobbies fearful of 'scientists playing God'.
SCIENCE, GENETICS AND ETHICS
33
? SCIENCE AND SENSIBILITY
less easy to solve is the angst which will hang over everyone's psychology. As things are now, we all know we are going to die, but most of us don't know when, so it doesn't feel like a death sentence. That may change, and society should be prepared for difficulties as people struggle to adjust their psychologies to it.
Ethics
I have already touched on some ethical issues. Science has no methods for deciding what is ethical. That is a matter for individuals and for society. But science can clarify the questions being asked, and can clear up obfuscating misunderstandings. This usually amounts to the useful 'You cannot have it both ways' style of arguing. I'll give five examples, before turning to a more unusual interpretation of the phrase 'science and ethics'.
Science cannot tell you whether abortion is wrong, but it can point out that the (embryological) continuum that seamlessly joins a non- sentient foetus to a sentient adult is analogous to the (evolutionary) continuum that joins humans to other species. If the embryological continuum appears to be more seamless, this is only because the evolutionary continuum is divided by the accident of extinction. Fundamental principles of ethics should not depend on the accidental contingencies of extinction. * To repeat, science cannot tell you whether abortion is murder, but it can warn you that you may be being inconsistent if you think abortion is murder but killing chimpanzees is not. You cannot have it both ways.
Science cannot tell you whether it is wrong to clone a whole human being. But it can tell you that a Dolly-style clone is just an identical twin, though of a different age. It can tell you that, if you want to object to cloning humans, you must not appeal to arguments such as 'The clone wouldn't be a full person' or 'The clone wouldn't have a soul'. Science cannot tell you whether anybody has a soul, but it can tell you that, if ordinary identical twins have souls, so do Dolly-style clones. f You cannot have it both ways.
Science cannot tell you whether stem cell cloning for 'spare parts' is wrong. But it can challenge you to explain how stem cell cloning differs morally from something that has long been accepted: tissue culture. Tissue culture has been a mainstay of cancer research for decades. The famous HeLa cell line, which originated in the late Henrietta Lacks in
? See 'Gaps in the Mind' (pp. 20-6) for a fuller discussion. tSee 'Dolly and the Cloth Heads' (pp. 152-5).
34
? 1951, is now being grown in labs all over the world. A typical lab, at the University of California, grows 48 litres of HeLa cells per day, as a routine service to researchers in the university. The total daily worldwide production of HeLa cells must be measured in tons - all a gigantic clone of Henrietta Lacks. In the half century since this mass production began, nobody seems to have objected to it. Those who agitate to stop stem cell research today have to explain why they do not object to the mass cultivation of HeLa cells. You cannot have it both ways.
Science cannot tell you whether it is right to kill 'Mary' to save her conjoined twin 'Jodie' (or whether both twins should be allowed to die). * But science can tell you that a placenta is a true clone of the baby it nourishes. You could legitimately 'spin' the story of any placenta as a 'twin' of the baby that it nourishes, to be discarded when its role is completed. Admittedly, nobody is tempted to call their placenta Mary, but one might equally question the emotional wisdom of bestowing such a name on a Siamese twin with no heart or lungs, and only a primitive brain. And if anybody wishes to invoke 'slippery slopes' and 'thin ends of wedges' here, let them think on the following.
In 1998, a television gastronome served on screen a new gourmet dish: human placenta. He
flash-fried strips of the placenta with shallots and blended two thirds into a puree. The rest was flambeed in brandy, and then sage and lime juice were added. The family of the baby concerned ate it, with twenty of their friends. The father thought it so delicious that he had fourteen helpings.
The whole thing was presented in the papers as a bit of a lark. Yet those who worry about slippery slopes need to ask themselves why that television dinner should not be called cannibalism. Cannibalism is one of our oldest and deepest taboos, and a devotee of the 'slippery slope' or 'thin end of the wedge' style of argument might do well to worry at the slightest breach of that taboo. I suspect that, if the television executives had known enough science to understand that a placenta is
? These were widely publicized pseudonyms given to a pair of conjoined 'Siamese' twins who came to Britain for medical treatment around this time. The authorities wanted, against the parents' wishes, to separate the twins, in a mammoth operation which could have given Jodie (some sort of) life but would certainly result in Mary's death. Without the operation both twins would die, because Mary, who lacked most vital organs including a functioning brain, subsisted parasitically on Jodie. Many liberal people thought it right to over-rule the parents' religiously-based reluctance to 'kill' Mary to save Jodie. I thought the parents were right to reject the operation, although for the wrong reasons, and that in any case their wishes should have been respected because it was they whose lives were likely to be profoundly affected by the demands of the severely handicapped surviving twin.
SCIENCE, GENETICS AND ETHICS
35
? SCIENCE AND SENSIBILITY
a true clone of a baby, the dinner would never have gone ahead, especially at the height of the Dolly-inspired cloning controversy. You cannot have it both ways.
I want to conclude with a rather idiosyncratic approach to the matter of science and ethics: ethical treatment of scientific truth itself. I want to suggest that objective truth sometimes needs the same kind of protection as the libel laws now give to individuals. Or at least to suggest that the Trades Descriptions Act might be more imaginatively invoked. I'll say a little about this first, in the light of Prince Charles's recent plea for public money to do research into 'alternative medicine'.
If a pharmaceutical company advertises its pills as curing headaches, it must be able to demonstrate, in double-blind controlled trials, that its pills do indeed cure headaches. Double-blind means, of course, that neither the patients, nor the testers, know until afterwards which patients received the dose, and which the placebo control. If the pills cannot pass this test - if numerous strenuous efforts fail to distinguish them from a neutral placebo - I presume the company might be in danger of prosecution under the Trades Descriptions Act.
Homeopathic remedies are big business, they are advertised as efficacious in various ways, yet they have never been demonstrated to have any effect at all. Personal testimony is ubiquitous, but it is useless evidence because of the notorious power of the placebo effect. This is exactly why 'orthodox' medicines are obliged to prove themselves in double-blind trials. *
I do not want to imply that all so-called 'alternative medicines' are as useless as homeopathy. For all I know, some of them may work. But they must be demonstrated to work, by double-blind placebo-control trials or some equivalent experimental design. And if they pass that test, there is then no longer any reason to call them 'alternative'.
Mainstream medicine would simply adopt them. As the distinguished journalist John Diamond wrote movingly (like many patients dying of cancer, he had false hopes cruelly raised by a succession of plausible quacks) in The Independent recently:
There is really no such thing as alternative medicine, just medicine that works and medicine that doesn't. . . There isn't an 'alternative' physiology or anatomy or nervous system any more than there's an alternative map of London which lets you get to Battersea from Chelsea without crossing the Thames.
? Homeopathy has special problems with double-blind control testing. I discuss this in my Foreword to John Diamond's Snake Oil (see pp. 181-4).
36
? But I began this final section in more radical terms. I wanted to extend the concept of libel to include lies that may not damage particular people but damage truth itself. Some twenty years ago, long before Dolly showed it was plausible, a book was published claiming, in great detail, that a rich man in South America had had himself cloned, by a scientist code-named Darwin. As a work of science fiction it would have been unexceptionable, but it was sold as sober fact. The author and publishers were sued, by Dr Derek Bromhall, who claimed that his reputation as a scientist was damaged by his being quoted in the book. My point is that whatever damage may or may not have been done to Dr Bromhall, far more important was the damage done to scientific truth itself.
That book has faded from memory and I bring it up only as an example. Obviously I want to generalize the principle to all deliberate falsifications, misrepresentations, of scientific truth. Why should a Derek Bromhall have to prove himself personally damaged, before we can prosecute a book which wantonly publishes lies about the universe? As will be obvious I'm no lawyer but, if I was, rather than constantly feel the need to drag things down to the question of whether particular humans have been damaged, I think I would like to stand up and defend truth itself. No doubt I shall be told - and convinced - that a court of law is not the right place for this. But in the wider world, if I am asked for a single phrase to characterize my role as Professor of Public Understanding of Science, I think I would choose Advocate for Disinterested Truth.
SCIENCE, GENETICS A N D ETHICS
37
? /
1. 5 Trial By Jury
Trial by jury must be one of the most conspicuously bad good ideas anyone ever had. Its devisers can hardly be blamed. They lived before the principles of statistical sampling and experimental design had been worked out. They weren't scientists. Let me explain using an analogy. And if, at the end, somebody objects to my argument on the grounds that humans aren't herring gulls, I'll have failed to get my point across.
Adult herring gulls have a bright yellow bill with a conspicuous red spot near the tip. Their babies peck at the red spot, which induces the parents to regurgitate food for them. Niko Tinbergen, Nobel Prize- winning zoologist and my old maestro at Oxford, offered naive young chicks a range of cardboard dummy gull heads varying in bill and spot colour, and shape. For each colour, shape or combination, Tinbergen measured the preferences of the baby chicks by counting their pecks in a standard time. The idea was to discover whether naive gull chicks are born with a built-in preference for long yellow things with red spots. If so, this would suggest that genes equip the young birds with detailed prior knowledge of the world in which they are about to hatch - a world in which food comes out of adult herring gull beaks.
Never mind the reason for the research, and never mind the con- clusions. Consider, instead, the methods you must use, and the pitfalls you must avoid, if you want to get a correct result in any such experiment. These turn out to be general principles which apply to human juries as strongly as to gull chicks.
First, you obviously must test more than one chick. It could be that some chicks are red-biased, others blue-biased, with no tendency for herring gull chicks in general to share the same favourite colour. So, by picking out a single chick, you are measuring nothing more than individual bias.
So, we must test more than one chick. How many? Is two enough? No, nor is three, and now we must start to think statistically. To make
38
? it simple, suppose that in a particular experiment we are comparing only red spots versus blue spots, both on a yellow background, and always presented simultaneously. If we test just two chicks separately, suppose the first chick chooses red. It had a 50 per cent chance of doing so, at random. Now the second chick also happens to choose red. Again, the odds were 50 per cent that it would do so at random, even if it were colourblind. There's a 50 per cent chance that two randomly choosing chicks will agree (half of the four possibilities: red red, red blue, blue red, blue blue). Three chicks aren't enough either. If you write down all the possibilities, you'll find that there's a 25 per cent chance of a unanimous verdict, by luck alone. Twenty-five per cent, as the odds of reaching a conclusion for the wrong reason, is unacceptably large.
How about twelve good chicks and true? Now you're talking. If twelve chicks are independently offered a choice between two alternatives, the odds that they will all reach the same verdict by chance alone are satisfyingly low, only one in 2048.
But now suppose that, instead of testing our twelve chicks indepen- dently, we test them as a group. We take a maelstrom of twelve cheeping chicks and lower into their midst a red spotted dummy and a blue spotted dummy, each fitted with an electrical device for auto- matically tallying pecks. And suppose that the collective of chicks registers 532 pecks at red and zero at blue. Does this massive disparity show that those twelve chicks prefer red? Absolutely not. The pecks are not independent data. Chicks could have a strong tendency to imitate one another (as well as imitate themselves in lock-on effects). If one chick just happened to peck at red first, others might copy him and the whole company of chicks join in a frenzy of imitative pecking. As a matter of fact this is precisely what domestic chicken chicks do, and gull chicks are very likely the same. Even if not, the principle remains that the data are not independent and the experiment is therefore invalid. The twelve chicks are strictly equivalent to a single chick, and their summed pecks, however numerous, might as well be only a single peck: they amount to only a single independent result.
Turning to courts of law, why are twelve jurors preferred to a single judge? Not because they are wiser, more knowledgeable or more practised in the arts of reasoning. Certainly not, and with a vengeance. Think of the astronomical damages awarded by juries in footling libel cases. Think how juries bring out the worst in histrionic, gallery-playing lawyers. Twelve jurors are preferred to one judge only because they are more numerous. Letting a single judge decide a verdict would be like letting a single chick speak for the whole herring gull species. Twelve
TRIAL BY JURY
39
? SCIENCE AND SENSIBILITY
heads are better than one, because they represent twelve assessments of the evidence.
But for this argument to be valid, the twelve assessments really have to be independent. And of course they are not. Twelve men and women locked in a jury room are like our clutch of twelve gull chicks. Whether they actually imitate each other like chicks, they might. That is enough to invalidate the principle by which a jury might be preferred over a single judge.
In practice, as is well documented and as I remember from the three juries that it has been my misfortune to serve on, juries are massively swayed by one or two vocal individuals. There is also strong pressure to conform to a unanimous verdict, which further undermines the principle of independent data. Increasing the number of jurors doesn't help, or not much (and not at all, in strict principle). What you have to increase is the number of independent verdict-reaching units.
Oddly enough, the bizarre American system of televising trials opens up a real possibility of improving the jury system. By the end of trials such as those of Louise Woodward or O. J. Simpson, literally thousands of people around the country have attended to the evidence as assiduously as the official jury. A mass phone-in might produce a fairer verdict than a jury. But, unfortunately, journalistic discussion, radio talk- shows and ordinary gossip would violate the Principle of Independent Data and we'd be back where we started. The broadcasting of trials, in any case, has horrible consequences. In the wake of Louise Woodward's trial, the Internet seethed with ill-spelled and ungrammatical vicious- ness, the cheque-book journalists were queuing up, and the unfortunate judge presiding had to change his telephone number and employ a bodyguard.
So, how can we improve the system? Should twelve jurors be locked in twelve isolation chambers and their opinions separately polled so that they constitute genuinely independent data? If it is objected that some would be too stupid or inarticulate to reach a verdict on their own, we are left wondering why such individuals are allowed on a jury at all. Perhaps there is something to be said for the collective wisdom that emerges when a group of people thrash out a topic together, round a table. But this still leaves the principle of independent data unsatisfied.
Should all cases be tried by two separate juries? Or three? Or twelve? Too expensive, at least if each jury has twelve members. Two juries of six members, or three juries of four members, would probably be an improvement over the present system. But isn't there some way of
40
? testing the relative merits of such alternative options, or of comparing the merits of trial by jury versus trial by judge?
Yes, there is. I'll call it the Two Verdicts Concordance Test. It is based on the principle that, if a decision is valid, two independent shots at making it should yield the same result. Just for purposes of the test, we run to the expense of having two juries, listening to the same case and forbidden to talk to members of the other jury. At the end, we lock the two juries in two separate jury rooms and see if they reach the same verdict. If they don't, neither verdict has been proved beyond reason- able doubt, and this would cast reasonable doubt on the jury system itself.
To make the experimental comparison with Trial by Judge, we need two experienced judges to listen to the same case, and require them too to reach their separate verdicts without talking to each other. Which- ever system, Trial by Jury or Trial by Judge, yields the higher score of agreements over a number of trials is the better system and might, if its Concordance Score is high, even be accredited for future use with some confidence.
Would you bet on two independent juries reaching the same verdict in the Louise Woodward case? Could you imagine even one other jury reaching the same verdict in the O. J. Simpson case? Two judges, on the other hand, seem to me rather likely to score well on the concordance test. And should I be charged with a serious crime, here's how I want to be tried. If I know myself to be guilty, I'll go with the loose cannon of a jury, the more ignorant, prejudiced and capricious the better. But if I am innocent, and the ideal of multiple independent decision-takers is unavailable, please give me a judge.
TRIAL BY JURY
41
? 1. 6 Crystalline Truth and
Crystal Balls21
A celebrated film star 'places four quartz crystal clusters in the four corners of her bathtub every time she takes a bath'. This doubtless has some mystic connection with the following recipe for meditation.
Each of the four quartz crystals in the meditation room should be 'programmed' to project gentle, loving, relaxing, crystalline energy towards all those present within the Meditation group. The quartz crystals will then generate a field of positive crystalline energy surrounding everyone in the room.
Language like this is a con-trick. It sounds 'scientific' enough to bamboozle the innocent. 'Programming' is what you do to computers. The word means nothing when applied to crystals. 'Energy' and 'field' are carefully defined notions in physics. There is no such thing as 'loving' or 'crystalline' energy, whether positive or no. *
New Age lore also advises placing a quartz crystal in your water jug. 'You will soon appreciate the sparkling purity of your crystal water. ' See how the trick works. Somebody with no understanding of the real world could make a kind of 'poetic' association with 'crystal clear' water. But that is no more sensible than trying to read by the light of a ('bright as a') button. Or putting ('hard as') nails under your pillow to assist an erection.
Try the following experiment when you next suffer from 'flu': hold your personal quartz crystal and visualize yellow light radiating through it. Then place your crystal in a jug of water and drink this water the next day; one cup of water at two-hourly intervals. You will be amazed at the result!
Drinking water at two-hourly intervals is a good idea anyway, when you have flu. Putting a quartz crystal in it will have no additional effect. In
*And, by the way, the next time you visit an 'alternative' therapist who claims to be 'balancing your energy fields', challenge them to say what they mean. The answer will be absolutely nothing.
42
? CRYSTALLINE TRUTH AND CRYSTAL BALLS
particular, no amount of 'visualizing' of coloured light will change the composition of either the crystal or the water.
Pseudoscientific drivel like this is a disturbingly prominent part of the culture of our age. I have limited my examples to crystals because I had to draw a line somewhere. But 'star signs' would have done just as well. Or 'angels', 'channelling', 'telepathy', 'quantum healing', 'homeopathy', 'map-dowsing'. There is no obvious limit to human gullibility. We are docile credulity-cows, eager victims of quacks and charlatans who milk us and grow fat. There is a rich living to be made by anyone prepared to prostitute the language - and the wonder - of science.
But isn't it all - crystal ball gazing, star signs, birth stones, ley-lines and the rest - just a bit of harmless fun? If people want to believe in garbage like astrology, or crystal healing, why not let them? But it's so sad to think about all that they are missing. There is so much wonder in real science. The universe is mysterious enough to need no help from warlocks, shamans and 'psychic' tricksters. These are at best a soul- sapping distraction. At worst they are dangerous profiteers.
The real world, properly understood in the scientific way, is deeply beautiful and unfailingly interesting. It's worth putting in some honest effort to understand it properly, undistracted by false wonder and prostituted pseudoscience. For illustration, we need look no farther than crystals themselves.
In a crystal such as quartz or diamond the atoms are arranged in a precisely repeating pattern. The atoms in a diamond - all identical carbon atoms - are arrayed like soldiers on parade except that the precision of their dressing far outsmarts the best-drilled guards regiment, and the atomic soldiers outnumber all the people that have ever lived or ever will. Imagine yourself shrunk to become one of the carbon atoms in the heart of a diamond crystal. You are one of the soldiers in a gigantic parade, but it'll seem a little odd because the files are arrayed in three dimensions. Perhaps a prodigious school of fish is a better image.
Each fish in the school is one carbon atom. Think of them hovering in space, keeping their distance from each other and holding their precise angles, by means of forces that you can't see but which scientists fully understand. But if this is a fish school, it is one that - to scale - would fill the Pacific Ocean. In any decent-sized diamond, you are likely to be looking along arrays of atoms numbering hundreds of millions in any one straight line.
Carbon atoms can take up other crystal lattice formations. To revert to the military analogy, they can adopt alternative drill conventions.
43
? SCIENCE AND SENSIBILITY
Graphite (the 'lead' in pencils) is also carbon, but it's obviously nothing like diamond. In graphite, the atoms form sheets of hexagons, like chicken wire. Each sheet is loosely bonded to those above and below it, and when impurities are present the sheets slide easily against each other, which is why graphite is a good lubricant. Diamond is very much not a lubricant. Its legendary hardness abrades the toughest materials. The atoms in soft graphite and hard diamond are identical. If you could persuade the atoms in graphite crystals to adopt the drill rules of diamond crystals, you'd be rich. It can be done, but you need colossal pressures and high temperatures, presumably the conditions that naturally manufacture diamonds, deep in the earth.
If hexagons make a sheet of flat graphite, you can imagine that interspersing some pentagons among the hexagons could make the sheet buckle into a curve. Place exactly 12 pentagons strategically among 20 hexagons and the curve bends round into a complete sphere. Geometers call it a truncated icosahedron. This is exactly the pattern of the sewing seams on a football. The football is, therefore, theoretically a pattern into which carbon atoms might spontaneously fall.
Mirabile dictu, exactly this pattern has been discovered among carbon atoms. The team responsible, including Sir Harry Kroto of Sussex University, won the 1996 Nobel Prize for Chemistry. Called Buckminsterfullerene, it is an elegant sphere of 60 carbon atoms, linked up as 20 hexagons interspersed with 12 pentagons. The name honours the visionary American architect Buckminster Fuller (whom I was privileged to meet when he was a very old man*) and the spheres are affectionately known as buckyballs. They can combine together to make larger crystals. Like graphite sheets, buckyballs make good lubricants, probably because of their spherical shape: they presumably work like tiny ball bearings.
Since the buckyball's discovery, chemists have realized that it is just a special case of a large family of 'buckytubes' and other 'fullerenes'. Carbon atoms can theoretically join up to form an Aladdin's cave of fascinating crystalline forms - another aspect of the unique property that qualifies carbon to be the fundamental element of life.
28
? not literally identical to the financial equivalent? Suppose it is sufficiently similar that the main routine can indeed be borrowed, but it still needs tweaking in detail. In that case, it is possible that the rocket could misfire if we naively import the subroutine raw. Switching back to biology, although genes really are watertight subroutines of digital software, they are not watertight in their effects on the development of the organism, for here they interact with their environment, including importantly the environment furnished by other genes. The antifreeze gene might depend, for optimal effect, on an interaction with other genes in the fish. Plonk it down in the foreign genetic climate of a tomato, and it might not work properly unless tweaked (which can be done) to mesh with the existing tomato genes.
What this means is that there is a case to be made on both sides of the argument, and we need to exercise subtle judgement. The genetic engineers are right that we can save time and trouble by climbing on the back of the millions of years of R & D that Darwinian natural selection has put into developing biological antifreeze (or whatever we are seeking). But the doomsayers would also have a point if they softened their stance from emotional gut rejection to a rational plea for rigorous safety testing. No reputable scientist would oppose such a plea. It is rightly routine for all new products, not just genetically engineered ones.
A largely unrecognized danger of the obsessive hysteria surrounding genetically modified foods is crying wolf. I fear that, if the green move- ment's high-amplitude warnings over GMOs turn out to be empty, people will be dangerously disinclined to listen to other and more serious warnings. The evolution of antibiotic resistance among bacteria is a vicious wolf of proven danger. Yet the menacing footfalls of this certain peril are all but drowned out in the caterwauling shrieks over genetically modified foods, whose dangers are speculative at most. To be more precise, genetic modification, like any other kind of modification, is good if you modify in a good direction, bad if you modify in a bad direction. Like domestic breeding, and like natural selection itself, the trick is to introduce the right new DNA software. The realization that software is all it is, written in exactly the same language as the organism's 'own' DNA, should go a long way towards dispelling the gut fears that rule most discussions of GMOs.
I can't leave the subject of gut feelings without a favourite quote from the lamented Carl Sagan. When asked a futurological question, he said that not enough was known to answer it. The questioner pressed him on what he really thought. 'What is your gut feeling? ' Sagan's reply is
SCIENCE, GENETICS A N D ETHICS
29
? SCIENCE AND SENSIBILITY
immortal: 'But I try not to think with my gut. ' Gut thinking is one of the main problems we have to contend with in public attitudes to science. I shall return to the point under Ethics. Meanwhile, some more remarks on the future of genetics in the twenty-first century, especially in the wake of the Human Genome Project (HGP).
The HGP, which will be completed any time now, is really a twentieth-century accomplishment. It is an outstanding success story, but it has limited scope. We have taken the human hard disk and transcribed every jot and tittle of the 11000101000010000111-style bits of information on it, regardless of what they mean in the software as a whole. The HGP needs to be followed up by a twenty-first-century Human Embryology Project (HEP) which, in effect, deciphers all the high-level software routines in which the machine-code instructions are embedded. An easier task will be a series of genome projects for different species (like the Arabidopsis plant genome project, whose completion is announced on the day that I write). These would be quicker and easier than the HGP, not because the other genomes are smaller or simpler than ours, but because the collective expertise of scientists increases cumulatively and rapidly with experience.
There is a frustrating aspect of this cumulative improvement. Given the rate of technological advance, with hindsight, when we started the Human Genome Project it wasn't worth starting. It would have been better to do nothing until the last two years and start then! Indeed, that is pretty much what the rival firm of Dr Craig Venter did. The fallacy in the 'never bother to start' maxim is that later technologies cannot get into a position to 'overtake' without the experience gained in developing the earlier ones. *
The HGP implicitly plays down the differences between individuals. But, with the intriguing exception of identical twins, everybody's genome is unique, and you might wonder whose genome is being sequenced in the HGP. Has some dignitary been singled out for the honour, is it a random person pulled off the street, or even an anonymous clone of cells in a tissue culture lab? It makes a difference. I have brown eyes while you have blue. I can't curl my tongue into a tube, whereas it's 50/50 that you can. Which version of the tongue-curling gene makes it into the published Human Genome? Which is the canonical eye colour? The answer is that, for the few 'letters' of the DNA text that vary, the canonical genome is the majority 'vote' among a sample of
*I have discussed the implications of the rapid growth of our understanding of genetics in more detail in 'Son of Moore's Law' (see pp. 107-15).
30
? people carefully chosen to give a good spread of human diversity. But the diversity itself is expunged from the record.
By contrast the Human Genome Diversity Project (HGDP), now under way, builds on the foundation of the HGP but focuses on those relatively few nucleotide sites that vary from person to person, and from group to group. Incidentally, a surprisingly small proportion of that variance consists of between-race variance, a fact that has sadly failed to reassure spokesmen for various ethnic groups, especially in America. They have dreamed up influential political objections to the project which they see as exploitative and tarred with the brush of eugenics.
The medical benefits of studying human variation could be immense. Hitherto, almost all medical prescribing has assumed that patients are pretty much the same, and that every disease has an optimal recom- mended cure. Doctors of tomorrow will be more like vets in this respect. Doctors have only one species of patient, but in future they will subdivide that species by genotype, as a vet subdivides his patients by species. For the special needs of blood transfusions, doctors already recognize a few genetic typings (OAB, Rh) etc. In the future, every patient's personal record will include the results of numerous genetic tests: not their entire genome (that will be too expensive for the foreseeable future) but, as the century goes on, an increasing sampling of the variable regions of the genome, and far more than the present 'blood group' typings. The point is that for some diseases there may be as many different optimal treatments as there are different genotypes at a locus - more even, because genetic loci may interact to affect susceptibility to disease.
Another important use of the genetics of human diversity is forensic. Precisely because DNA is digital like computer bytes, genetic finger- printing is potentially many many orders of magnitude more accurate and reliable than any other means of individual identification, including direct facial recognition (despite the unshakeable gut feeling of jurors that eyewitness identification trumps everything). Moreover, identity can be established from a tiny trace of blood, sweat or tears (or spit, semen or hairs).
DNA evidence is widely regarded as controversial, and I need to say a little about why. Firstly, human error can obviously vitiate the accuracy of the method. But that is true of all evidence. Courts are already accustomed to taking precautions to avoid the muddling up of specimens, and such precautions now become even more important. DNA finger- printing can establish, almost infinitely far beyond all reasonable doubt, whether a smear of blood came from a particular individual. But
SCIENCE, GENETICS AND ETHICS
31
? SCIENCE AND SENSIBILITY
obviously you must test the right smear.
Secondly, astronomical though the odds against mistaken identity by
DNA fingerprinting theoretically are, it is possible for geneticists and statisticians to come up with what seem like widely different estimates
19
of the precise odds. I quote from Unweaving the Rainbow (Chapter 5,
which is devoted to explain DNA fingerprinting in lay terms).
Lawyers are accustomed to pouncing when expert witnesses seem to disagree. If two geneticists are summoned to the stand and are asked to estimate the probability of a misidentification with DNA evidence, the first may say 1,000,000 to one while the second may say only 100,000 to one. Pounce. 'Aha! AHA! The experts disagree! Ladies and gentlemen of the jury, what confidence can we place in a scientific method, if the experts themselves can't get within a factor of ten of one another? Obviously the only thing to do is throw the entire evidence out, lock, stock and barrel. '
But. . . any disagreement. . . is only over whether the odds against a wrongful identification are hyper-mega-astronomical, or just plain astronomical. The odds cannot normally be lower than thousands to one, and they may well be up in the billions. Even on the most conservative estimate, the odds against wrongful identification are hugely greater than they are in an ordinary identity parade. 'M'lud, an identity parade of only 20 men is grossly unfair on my client. I demand a line-up of at least a million men! '
The idea of a nationwide database, in which all citizens' DNA finger- prints would be held, is now being discussed (only a sample of genes, of course: doing the whole genome would be overkill, far too expensive). I don't see this as a sinister, Big Brotherish idea (and I have written to my doctor volunteering to be a guinea pig in the pilot study of 500,000 now being prepared). But there are potential problems, of a civil liberties character. If your house is burgled, the police will routinely look for (traditional, old-fashioned) fingerprints of the burglar. They need to fingerprint the householder's family too, for elimination purposes, and most people are happy to oblige. Obviously the same principle will apply to DNA fingerprinting, but many people would want to stop well short of a nationwide database. Presumably they would also object to a nationwide database of conventional, old-fashioned fingerprints, but perhaps that is not a practical issue because it would take too long to search through it for a match. DNA fingerprinting doesn't suffer from this difficulty. Computer searches of huge DNA databases could be accomplished swiftly.
What, then, are the civil liberties problems? Surely, those with nothing to hide will have nothing to fear? Perhaps not, but some
32
? people do have legitimate reasons to hide information, not from the law but from each other. A surprisingly large number of people, of all ages, are genetically unrelated to the man they think is their father. To put it mildly, it is not clear that to disillusion them, with conclusive DNA evidence, would increase the sum of human happiness. If a national DNA database were in place, it might be hard to control unauthorized access to it. If a tabloid newspaper were to discover that the official heir to a Dukedom was actually sired by the gamekeeper, the consternation in the College of Heralds might be mildly amusing. But in the population at large it doesn't take much to imagine the family recriminations and sheer private misery that could flow from freely available information of true paternity. Nevertheless, the existence of a national DNA database wouldn't alter the situation much. It is already perfectly feasible for a jealous husband, say, to take a saliva or blood sample from one of his supposed children and compare it with his own, in order to confirm his suspicion that he is not the real father. What the national database could add is a swift computer search to find out who, out of all the males in the entire country, is\
More generally, the study of human diversity is one of very few areas where a good (though in my opinion not overwhelming) case can be made against the pure disinterested search for knowledge: one of very few areas where we might actually be better off ignorant. It is possible that, by the end of the twenty-first century, doctors will be able accurately to predict the manner and time of death of everybody, from the day they are conceived. At present this kind of deterministic prognostica- tion can be achieved only for possessors of genes such as Huntington's Chorea. * For the rest of us, all that is possible is the vague statistical forecast of the life insurance actuary, based on our smoking and drink- ing habits, and a quick listen through a stethoscope. The whole life insurance business depends upon such forecasts being vague and statistical. Those who die old subsidize (the heirs of) those who die young. If the day comes when deterministic forecasting (along Huntington's Chorea lines) becomes universal, life insurance as we know it will collapse. That problem is soluble (presumably by universal compulsory life insurance with no individual medical risk assessment). What will be
*The folk singer Woody Guthrie died of Huntington's Chorea, a horrible disease that waits till early middle age before killing you. It's a dominant gene, so each of Woody's children knows that he has an exactly 50 per cent chance of suffering the same horrible fate. Some people, given these odds, prefer not to be tested. They'd rather not know until they have to. IVF doctors can now push the test back to the newly fertilized zygote, and choose to implant only those that lack the fatal gene. This is obviously a huge boon, but it is attacked by ignorant lobbies fearful of 'scientists playing God'.
SCIENCE, GENETICS AND ETHICS
33
? SCIENCE AND SENSIBILITY
less easy to solve is the angst which will hang over everyone's psychology. As things are now, we all know we are going to die, but most of us don't know when, so it doesn't feel like a death sentence. That may change, and society should be prepared for difficulties as people struggle to adjust their psychologies to it.
Ethics
I have already touched on some ethical issues. Science has no methods for deciding what is ethical. That is a matter for individuals and for society. But science can clarify the questions being asked, and can clear up obfuscating misunderstandings. This usually amounts to the useful 'You cannot have it both ways' style of arguing. I'll give five examples, before turning to a more unusual interpretation of the phrase 'science and ethics'.
Science cannot tell you whether abortion is wrong, but it can point out that the (embryological) continuum that seamlessly joins a non- sentient foetus to a sentient adult is analogous to the (evolutionary) continuum that joins humans to other species. If the embryological continuum appears to be more seamless, this is only because the evolutionary continuum is divided by the accident of extinction. Fundamental principles of ethics should not depend on the accidental contingencies of extinction. * To repeat, science cannot tell you whether abortion is murder, but it can warn you that you may be being inconsistent if you think abortion is murder but killing chimpanzees is not. You cannot have it both ways.
Science cannot tell you whether it is wrong to clone a whole human being. But it can tell you that a Dolly-style clone is just an identical twin, though of a different age. It can tell you that, if you want to object to cloning humans, you must not appeal to arguments such as 'The clone wouldn't be a full person' or 'The clone wouldn't have a soul'. Science cannot tell you whether anybody has a soul, but it can tell you that, if ordinary identical twins have souls, so do Dolly-style clones. f You cannot have it both ways.
Science cannot tell you whether stem cell cloning for 'spare parts' is wrong. But it can challenge you to explain how stem cell cloning differs morally from something that has long been accepted: tissue culture. Tissue culture has been a mainstay of cancer research for decades. The famous HeLa cell line, which originated in the late Henrietta Lacks in
? See 'Gaps in the Mind' (pp. 20-6) for a fuller discussion. tSee 'Dolly and the Cloth Heads' (pp. 152-5).
34
? 1951, is now being grown in labs all over the world. A typical lab, at the University of California, grows 48 litres of HeLa cells per day, as a routine service to researchers in the university. The total daily worldwide production of HeLa cells must be measured in tons - all a gigantic clone of Henrietta Lacks. In the half century since this mass production began, nobody seems to have objected to it. Those who agitate to stop stem cell research today have to explain why they do not object to the mass cultivation of HeLa cells. You cannot have it both ways.
Science cannot tell you whether it is right to kill 'Mary' to save her conjoined twin 'Jodie' (or whether both twins should be allowed to die). * But science can tell you that a placenta is a true clone of the baby it nourishes. You could legitimately 'spin' the story of any placenta as a 'twin' of the baby that it nourishes, to be discarded when its role is completed. Admittedly, nobody is tempted to call their placenta Mary, but one might equally question the emotional wisdom of bestowing such a name on a Siamese twin with no heart or lungs, and only a primitive brain. And if anybody wishes to invoke 'slippery slopes' and 'thin ends of wedges' here, let them think on the following.
In 1998, a television gastronome served on screen a new gourmet dish: human placenta. He
flash-fried strips of the placenta with shallots and blended two thirds into a puree. The rest was flambeed in brandy, and then sage and lime juice were added. The family of the baby concerned ate it, with twenty of their friends. The father thought it so delicious that he had fourteen helpings.
The whole thing was presented in the papers as a bit of a lark. Yet those who worry about slippery slopes need to ask themselves why that television dinner should not be called cannibalism. Cannibalism is one of our oldest and deepest taboos, and a devotee of the 'slippery slope' or 'thin end of the wedge' style of argument might do well to worry at the slightest breach of that taboo. I suspect that, if the television executives had known enough science to understand that a placenta is
? These were widely publicized pseudonyms given to a pair of conjoined 'Siamese' twins who came to Britain for medical treatment around this time. The authorities wanted, against the parents' wishes, to separate the twins, in a mammoth operation which could have given Jodie (some sort of) life but would certainly result in Mary's death. Without the operation both twins would die, because Mary, who lacked most vital organs including a functioning brain, subsisted parasitically on Jodie. Many liberal people thought it right to over-rule the parents' religiously-based reluctance to 'kill' Mary to save Jodie. I thought the parents were right to reject the operation, although for the wrong reasons, and that in any case their wishes should have been respected because it was they whose lives were likely to be profoundly affected by the demands of the severely handicapped surviving twin.
SCIENCE, GENETICS AND ETHICS
35
? SCIENCE AND SENSIBILITY
a true clone of a baby, the dinner would never have gone ahead, especially at the height of the Dolly-inspired cloning controversy. You cannot have it both ways.
I want to conclude with a rather idiosyncratic approach to the matter of science and ethics: ethical treatment of scientific truth itself. I want to suggest that objective truth sometimes needs the same kind of protection as the libel laws now give to individuals. Or at least to suggest that the Trades Descriptions Act might be more imaginatively invoked. I'll say a little about this first, in the light of Prince Charles's recent plea for public money to do research into 'alternative medicine'.
If a pharmaceutical company advertises its pills as curing headaches, it must be able to demonstrate, in double-blind controlled trials, that its pills do indeed cure headaches. Double-blind means, of course, that neither the patients, nor the testers, know until afterwards which patients received the dose, and which the placebo control. If the pills cannot pass this test - if numerous strenuous efforts fail to distinguish them from a neutral placebo - I presume the company might be in danger of prosecution under the Trades Descriptions Act.
Homeopathic remedies are big business, they are advertised as efficacious in various ways, yet they have never been demonstrated to have any effect at all. Personal testimony is ubiquitous, but it is useless evidence because of the notorious power of the placebo effect. This is exactly why 'orthodox' medicines are obliged to prove themselves in double-blind trials. *
I do not want to imply that all so-called 'alternative medicines' are as useless as homeopathy. For all I know, some of them may work. But they must be demonstrated to work, by double-blind placebo-control trials or some equivalent experimental design. And if they pass that test, there is then no longer any reason to call them 'alternative'.
Mainstream medicine would simply adopt them. As the distinguished journalist John Diamond wrote movingly (like many patients dying of cancer, he had false hopes cruelly raised by a succession of plausible quacks) in The Independent recently:
There is really no such thing as alternative medicine, just medicine that works and medicine that doesn't. . . There isn't an 'alternative' physiology or anatomy or nervous system any more than there's an alternative map of London which lets you get to Battersea from Chelsea without crossing the Thames.
? Homeopathy has special problems with double-blind control testing. I discuss this in my Foreword to John Diamond's Snake Oil (see pp. 181-4).
36
? But I began this final section in more radical terms. I wanted to extend the concept of libel to include lies that may not damage particular people but damage truth itself. Some twenty years ago, long before Dolly showed it was plausible, a book was published claiming, in great detail, that a rich man in South America had had himself cloned, by a scientist code-named Darwin. As a work of science fiction it would have been unexceptionable, but it was sold as sober fact. The author and publishers were sued, by Dr Derek Bromhall, who claimed that his reputation as a scientist was damaged by his being quoted in the book. My point is that whatever damage may or may not have been done to Dr Bromhall, far more important was the damage done to scientific truth itself.
That book has faded from memory and I bring it up only as an example. Obviously I want to generalize the principle to all deliberate falsifications, misrepresentations, of scientific truth. Why should a Derek Bromhall have to prove himself personally damaged, before we can prosecute a book which wantonly publishes lies about the universe? As will be obvious I'm no lawyer but, if I was, rather than constantly feel the need to drag things down to the question of whether particular humans have been damaged, I think I would like to stand up and defend truth itself. No doubt I shall be told - and convinced - that a court of law is not the right place for this. But in the wider world, if I am asked for a single phrase to characterize my role as Professor of Public Understanding of Science, I think I would choose Advocate for Disinterested Truth.
SCIENCE, GENETICS A N D ETHICS
37
? /
1. 5 Trial By Jury
Trial by jury must be one of the most conspicuously bad good ideas anyone ever had. Its devisers can hardly be blamed. They lived before the principles of statistical sampling and experimental design had been worked out. They weren't scientists. Let me explain using an analogy. And if, at the end, somebody objects to my argument on the grounds that humans aren't herring gulls, I'll have failed to get my point across.
Adult herring gulls have a bright yellow bill with a conspicuous red spot near the tip. Their babies peck at the red spot, which induces the parents to regurgitate food for them. Niko Tinbergen, Nobel Prize- winning zoologist and my old maestro at Oxford, offered naive young chicks a range of cardboard dummy gull heads varying in bill and spot colour, and shape. For each colour, shape or combination, Tinbergen measured the preferences of the baby chicks by counting their pecks in a standard time. The idea was to discover whether naive gull chicks are born with a built-in preference for long yellow things with red spots. If so, this would suggest that genes equip the young birds with detailed prior knowledge of the world in which they are about to hatch - a world in which food comes out of adult herring gull beaks.
Never mind the reason for the research, and never mind the con- clusions. Consider, instead, the methods you must use, and the pitfalls you must avoid, if you want to get a correct result in any such experiment. These turn out to be general principles which apply to human juries as strongly as to gull chicks.
First, you obviously must test more than one chick. It could be that some chicks are red-biased, others blue-biased, with no tendency for herring gull chicks in general to share the same favourite colour. So, by picking out a single chick, you are measuring nothing more than individual bias.
So, we must test more than one chick. How many? Is two enough? No, nor is three, and now we must start to think statistically. To make
38
? it simple, suppose that in a particular experiment we are comparing only red spots versus blue spots, both on a yellow background, and always presented simultaneously. If we test just two chicks separately, suppose the first chick chooses red. It had a 50 per cent chance of doing so, at random. Now the second chick also happens to choose red. Again, the odds were 50 per cent that it would do so at random, even if it were colourblind. There's a 50 per cent chance that two randomly choosing chicks will agree (half of the four possibilities: red red, red blue, blue red, blue blue). Three chicks aren't enough either. If you write down all the possibilities, you'll find that there's a 25 per cent chance of a unanimous verdict, by luck alone. Twenty-five per cent, as the odds of reaching a conclusion for the wrong reason, is unacceptably large.
How about twelve good chicks and true? Now you're talking. If twelve chicks are independently offered a choice between two alternatives, the odds that they will all reach the same verdict by chance alone are satisfyingly low, only one in 2048.
But now suppose that, instead of testing our twelve chicks indepen- dently, we test them as a group. We take a maelstrom of twelve cheeping chicks and lower into their midst a red spotted dummy and a blue spotted dummy, each fitted with an electrical device for auto- matically tallying pecks. And suppose that the collective of chicks registers 532 pecks at red and zero at blue. Does this massive disparity show that those twelve chicks prefer red? Absolutely not. The pecks are not independent data. Chicks could have a strong tendency to imitate one another (as well as imitate themselves in lock-on effects). If one chick just happened to peck at red first, others might copy him and the whole company of chicks join in a frenzy of imitative pecking. As a matter of fact this is precisely what domestic chicken chicks do, and gull chicks are very likely the same. Even if not, the principle remains that the data are not independent and the experiment is therefore invalid. The twelve chicks are strictly equivalent to a single chick, and their summed pecks, however numerous, might as well be only a single peck: they amount to only a single independent result.
Turning to courts of law, why are twelve jurors preferred to a single judge? Not because they are wiser, more knowledgeable or more practised in the arts of reasoning. Certainly not, and with a vengeance. Think of the astronomical damages awarded by juries in footling libel cases. Think how juries bring out the worst in histrionic, gallery-playing lawyers. Twelve jurors are preferred to one judge only because they are more numerous. Letting a single judge decide a verdict would be like letting a single chick speak for the whole herring gull species. Twelve
TRIAL BY JURY
39
? SCIENCE AND SENSIBILITY
heads are better than one, because they represent twelve assessments of the evidence.
But for this argument to be valid, the twelve assessments really have to be independent. And of course they are not. Twelve men and women locked in a jury room are like our clutch of twelve gull chicks. Whether they actually imitate each other like chicks, they might. That is enough to invalidate the principle by which a jury might be preferred over a single judge.
In practice, as is well documented and as I remember from the three juries that it has been my misfortune to serve on, juries are massively swayed by one or two vocal individuals. There is also strong pressure to conform to a unanimous verdict, which further undermines the principle of independent data. Increasing the number of jurors doesn't help, or not much (and not at all, in strict principle). What you have to increase is the number of independent verdict-reaching units.
Oddly enough, the bizarre American system of televising trials opens up a real possibility of improving the jury system. By the end of trials such as those of Louise Woodward or O. J. Simpson, literally thousands of people around the country have attended to the evidence as assiduously as the official jury. A mass phone-in might produce a fairer verdict than a jury. But, unfortunately, journalistic discussion, radio talk- shows and ordinary gossip would violate the Principle of Independent Data and we'd be back where we started. The broadcasting of trials, in any case, has horrible consequences. In the wake of Louise Woodward's trial, the Internet seethed with ill-spelled and ungrammatical vicious- ness, the cheque-book journalists were queuing up, and the unfortunate judge presiding had to change his telephone number and employ a bodyguard.
So, how can we improve the system? Should twelve jurors be locked in twelve isolation chambers and their opinions separately polled so that they constitute genuinely independent data? If it is objected that some would be too stupid or inarticulate to reach a verdict on their own, we are left wondering why such individuals are allowed on a jury at all. Perhaps there is something to be said for the collective wisdom that emerges when a group of people thrash out a topic together, round a table. But this still leaves the principle of independent data unsatisfied.
Should all cases be tried by two separate juries? Or three? Or twelve? Too expensive, at least if each jury has twelve members. Two juries of six members, or three juries of four members, would probably be an improvement over the present system. But isn't there some way of
40
? testing the relative merits of such alternative options, or of comparing the merits of trial by jury versus trial by judge?
Yes, there is. I'll call it the Two Verdicts Concordance Test. It is based on the principle that, if a decision is valid, two independent shots at making it should yield the same result. Just for purposes of the test, we run to the expense of having two juries, listening to the same case and forbidden to talk to members of the other jury. At the end, we lock the two juries in two separate jury rooms and see if they reach the same verdict. If they don't, neither verdict has been proved beyond reason- able doubt, and this would cast reasonable doubt on the jury system itself.
To make the experimental comparison with Trial by Judge, we need two experienced judges to listen to the same case, and require them too to reach their separate verdicts without talking to each other. Which- ever system, Trial by Jury or Trial by Judge, yields the higher score of agreements over a number of trials is the better system and might, if its Concordance Score is high, even be accredited for future use with some confidence.
Would you bet on two independent juries reaching the same verdict in the Louise Woodward case? Could you imagine even one other jury reaching the same verdict in the O. J. Simpson case? Two judges, on the other hand, seem to me rather likely to score well on the concordance test. And should I be charged with a serious crime, here's how I want to be tried. If I know myself to be guilty, I'll go with the loose cannon of a jury, the more ignorant, prejudiced and capricious the better. But if I am innocent, and the ideal of multiple independent decision-takers is unavailable, please give me a judge.
TRIAL BY JURY
41
? 1. 6 Crystalline Truth and
Crystal Balls21
A celebrated film star 'places four quartz crystal clusters in the four corners of her bathtub every time she takes a bath'. This doubtless has some mystic connection with the following recipe for meditation.
Each of the four quartz crystals in the meditation room should be 'programmed' to project gentle, loving, relaxing, crystalline energy towards all those present within the Meditation group. The quartz crystals will then generate a field of positive crystalline energy surrounding everyone in the room.
Language like this is a con-trick. It sounds 'scientific' enough to bamboozle the innocent. 'Programming' is what you do to computers. The word means nothing when applied to crystals. 'Energy' and 'field' are carefully defined notions in physics. There is no such thing as 'loving' or 'crystalline' energy, whether positive or no. *
New Age lore also advises placing a quartz crystal in your water jug. 'You will soon appreciate the sparkling purity of your crystal water. ' See how the trick works. Somebody with no understanding of the real world could make a kind of 'poetic' association with 'crystal clear' water. But that is no more sensible than trying to read by the light of a ('bright as a') button. Or putting ('hard as') nails under your pillow to assist an erection.
Try the following experiment when you next suffer from 'flu': hold your personal quartz crystal and visualize yellow light radiating through it. Then place your crystal in a jug of water and drink this water the next day; one cup of water at two-hourly intervals. You will be amazed at the result!
Drinking water at two-hourly intervals is a good idea anyway, when you have flu. Putting a quartz crystal in it will have no additional effect. In
*And, by the way, the next time you visit an 'alternative' therapist who claims to be 'balancing your energy fields', challenge them to say what they mean. The answer will be absolutely nothing.
42
? CRYSTALLINE TRUTH AND CRYSTAL BALLS
particular, no amount of 'visualizing' of coloured light will change the composition of either the crystal or the water.
Pseudoscientific drivel like this is a disturbingly prominent part of the culture of our age. I have limited my examples to crystals because I had to draw a line somewhere. But 'star signs' would have done just as well. Or 'angels', 'channelling', 'telepathy', 'quantum healing', 'homeopathy', 'map-dowsing'. There is no obvious limit to human gullibility. We are docile credulity-cows, eager victims of quacks and charlatans who milk us and grow fat. There is a rich living to be made by anyone prepared to prostitute the language - and the wonder - of science.
But isn't it all - crystal ball gazing, star signs, birth stones, ley-lines and the rest - just a bit of harmless fun? If people want to believe in garbage like astrology, or crystal healing, why not let them? But it's so sad to think about all that they are missing. There is so much wonder in real science. The universe is mysterious enough to need no help from warlocks, shamans and 'psychic' tricksters. These are at best a soul- sapping distraction. At worst they are dangerous profiteers.
The real world, properly understood in the scientific way, is deeply beautiful and unfailingly interesting. It's worth putting in some honest effort to understand it properly, undistracted by false wonder and prostituted pseudoscience. For illustration, we need look no farther than crystals themselves.
In a crystal such as quartz or diamond the atoms are arranged in a precisely repeating pattern. The atoms in a diamond - all identical carbon atoms - are arrayed like soldiers on parade except that the precision of their dressing far outsmarts the best-drilled guards regiment, and the atomic soldiers outnumber all the people that have ever lived or ever will. Imagine yourself shrunk to become one of the carbon atoms in the heart of a diamond crystal. You are one of the soldiers in a gigantic parade, but it'll seem a little odd because the files are arrayed in three dimensions. Perhaps a prodigious school of fish is a better image.
Each fish in the school is one carbon atom. Think of them hovering in space, keeping their distance from each other and holding their precise angles, by means of forces that you can't see but which scientists fully understand. But if this is a fish school, it is one that - to scale - would fill the Pacific Ocean. In any decent-sized diamond, you are likely to be looking along arrays of atoms numbering hundreds of millions in any one straight line.
Carbon atoms can take up other crystal lattice formations. To revert to the military analogy, they can adopt alternative drill conventions.
43
? SCIENCE AND SENSIBILITY
Graphite (the 'lead' in pencils) is also carbon, but it's obviously nothing like diamond. In graphite, the atoms form sheets of hexagons, like chicken wire. Each sheet is loosely bonded to those above and below it, and when impurities are present the sheets slide easily against each other, which is why graphite is a good lubricant. Diamond is very much not a lubricant. Its legendary hardness abrades the toughest materials. The atoms in soft graphite and hard diamond are identical. If you could persuade the atoms in graphite crystals to adopt the drill rules of diamond crystals, you'd be rich. It can be done, but you need colossal pressures and high temperatures, presumably the conditions that naturally manufacture diamonds, deep in the earth.
If hexagons make a sheet of flat graphite, you can imagine that interspersing some pentagons among the hexagons could make the sheet buckle into a curve. Place exactly 12 pentagons strategically among 20 hexagons and the curve bends round into a complete sphere. Geometers call it a truncated icosahedron. This is exactly the pattern of the sewing seams on a football. The football is, therefore, theoretically a pattern into which carbon atoms might spontaneously fall.
Mirabile dictu, exactly this pattern has been discovered among carbon atoms. The team responsible, including Sir Harry Kroto of Sussex University, won the 1996 Nobel Prize for Chemistry. Called Buckminsterfullerene, it is an elegant sphere of 60 carbon atoms, linked up as 20 hexagons interspersed with 12 pentagons. The name honours the visionary American architect Buckminster Fuller (whom I was privileged to meet when he was a very old man*) and the spheres are affectionately known as buckyballs. They can combine together to make larger crystals. Like graphite sheets, buckyballs make good lubricants, probably because of their spherical shape: they presumably work like tiny ball bearings.
Since the buckyball's discovery, chemists have realized that it is just a special case of a large family of 'buckytubes' and other 'fullerenes'. Carbon atoms can theoretically join up to form an Aladdin's cave of fascinating crystalline forms - another aspect of the unique property that qualifies carbon to be the fundamental element of life.
