'And he gave unto Moses two tables of testimony, tables of stone, written with the finger of God'

Exodus 31:18

The governing principle of forensic science, as laid down by

Edmond Locard at the beginning of the last century, is that 'every contact leaves a trace'. But unless we know how to analyse, categorise and understand those traces, they're not much use when it comes to catching criminals. As scientists have made new discoveries, so the art of detection has advanced. And the technique of identification from fingerprints was a headline-grabbing trailblazer in terms of bringing criminals to justice

Forensic science did not begin with fingerprinting, but it caught the public imagination in a way that no other development had. And because it was so easy to comprehend, the courts also took to it readily. In the early 1900s the law-abiding citizenry fell in love with the idea that a silent burglar who touched what wasn't his could be identified just as silently; that the murderer who took a life with a blunt instrument could swing from the gallows thanks to the pattern on the tip of his pinkie; that a moment's carelessness would lead inevitably to conviction thanks to the unique arrangement of a clutch of ridges and loops.

One of the first Europeans to grasp the idea of the individuality of fingerprints was a young man named William Herschel. In 1853 he set sail from England to work for the East India Company, which effectively ruled large parts of India. Four years later a dispute over the type of grease used in gun cartridges led to a group of the company's Indian soldiers mutinying against their British commanders. The subsequent rebellion – known as the Indian Mutiny – spread across the country, leading to widespread violence met with vicious reprisals from the British forces. When the dust settled, the East India Company was forced to turn over its territories to the British Crown and many of the company's employees were transferred to the Indian Civil Service. Herschel was put in charge of a rural region in Bengal.

The brutality of the rebellion had left feelings running high and many Indian citizens were determined to make life as difficult as possible for their British overlords. They stopped turning up for work, paying taxes and cultivating British farms.

Herschel, an ambitious 25-year-old, was determined not to let civil unrest stand in the way of his making a mark. One of the first decisions he took in his new role was to build a road. He drew up a contract with Konai, a local man, to supply equipment for the project. Then he did something very odd.

'I dabbed Konai's palm and fingers over with home-made oil-ink used for myofficial seal, and pressed the whole hand on the back of the contract, and westudied it together, with a good deal of chaff about palmistry, comparing hispalm with mine on another impression.' When Herschel printed Konai's hand, hewasn't thinking of it in terms of identification, but as a kind of insurance –'to frighten him out of all thought of repudiating his signature'.

Although rare, and by 1861 illegal, Herschel may have come up with the idea forthe handprint from the Hindu practice of suttee, whereby a widow was burnedalive on the funeral pyre of her dead husband. As she passed through the'Suttee Gate' on her way to death, the doomed woman would dip her hand in reddye and place it on the Gate. The stonework around the print would then becarved away to make it stand out in basrelief.

Twenty years later, Herschel was appointed magistrate at Hooghly, nearCalcutta, where he was responsible for the courts, the prison and pensions. Wethink of benefit fraud as something modern, but Herschel was aware of it 140years ago. He set up a system for taking the fingerprints of pensioners so thatwhen they died other people couldn't fraudulently collect their pensions. Healso took the prints of people when they were given jail sentences, to stopconvicted criminals paying others to serve their prison sentences for them.

The idea of being able formally and categorically to identify criminals wasgathering traction in various jurisdictions. Around the same time that Herschelwas developing his system, a police clerk in Paris called Alphonse Bertillonwas buckling under the volume of incoming prisoners. He decided to try toidentify them systematically using anthropometry, the science of measuringhumans. He chose eleven body measurements, including width of head anddistances from elbow to end of middle finger. Bertillon put the odds of twopeople sharing all eleven of his measurements at one in 286 million. Herecorded the individual measurements on file cards, and in the middle of thecards he stuck two photographs – full face and profile – and thus the mug shotwas born.

Meanwhile, near Tokyo, a Scottish medical missionary started experimenting withfingerprints. Henry Faulds had noticed that ancient potters marked their potswith finger and thumb marks. He also discovered that subtle prints could bemade visible when dusted with powder, and he used the technique to exonerate aman accused of burglary. When Faulds showed the real thief the similaritiesbetween his print and the one on a windowpane in the burgled house, he brokedown and confessed. As a result of his observations, Faulds devised a method offingerprint classification based on prints from all ten fingers. He tried topersuade Scotland Yard to set up a fingerprint department using his system, buthe was rejected.

The Bertillonage record of twenty-one-year-old George Girolami, arrested forfraud

Undaunted, Henry Faulds wrote to Charles Darwin detailing his fingerprintingmethod. Darwin was intrigued by the idea but felt that it was work more suitedto a younger man, so he passed it on to his cousin, Francis Galton. Galtonspent ten years studying fingerprints and wrote the first book on the subject,Finger Prints (1892), in which he distinguished eight basic fingerprintpatterns of arches, loops and whorls. He also demonstrated that every humanfinger fits into one of these categories in its own unique way.

After reading an article by Galton, Croatian-born police official Juan Vucetichbegan collecting the fingerprints of arrested men in Buenos Aires, Argentina.He devised his own ten-finger system of classification which he called'dactyloscopy' – and which many Spanish-speaking countries still use today. Aswell as being used in criminal cases, it was quickly adopted by the Argentiniangovernment as a form of verification on internal identity passes.

Vucetich's system soon faced a demanding and disturbing test case. On 29 June1892, in a village near Buenos Aires, 4-year-old Teresa Rojas and her6-year-old brother Ponciano were found brutally murdered in their home. Theirmother, Francisca, was alive but her throat had been cut.

Francisca told the police that her neighbour, Pedro Velázquez, had stormed intoher home, murdered her children and tried to slit her throat.

The police tortured Velázquez for a week, but he stuck to his alibi: he hadbeen out with a group of friends at the time of the murders.

Frustrated at the lack of a confession, Inspector Commissioner Alvarez returnedto the house. This time he noticed a brown patch on a doorframe, which hethought might be a bloody fingerprint. He removed the bloodstained section ofwood from the frame and took it, along with fingerprints taken from Velázquez,to Juan Vucetich, who had just opened a fingerprint bureau in Buenos Aires.

Vucetich confidently declared that the prints did not match those on thedoorpost. Then he took the prints of Francisca Rojas. They were identical.Faced with the damning bloody fingerprint evidence, the mother confessed tomurdering her two children, cutting her own throat for effect and accusing aninnocent man. She had wanted to improve her chances of marrying her boyfriend,who didn't like children. Instead she became the first person to be convictedof a crime based on fingerprint evidence. She was sentenced to lifeimprisonment.

Following the Rojas case, Argentina abandoned Bertillon's anthropometric systemand began organising its criminal records on fingerprints alone. Soon, othercountries began to follow suit. The following year, Edward Henry, the chief ofpolice in Bengal, added thumbprints to his anthropometric criminal records.Although fingerprinting had been used officially in civil matters in Bengalsince William Herschel had introduced the system forty years previously, thepolice had never taken advantage of it. Working with an Indian police officer,Azizul Haque, Henry improved on Galton's system to produce one which allowed aninvestigator to use the physical characteristics of a fingerprint to create aunique reference number. These numbers were then used to file the fingerprintsin one of 1024 pigeonholes in the police headquarters; when a new fingerprintwas taken, its characteristics were coded, and the appropriate pigeonhole checkedto see if it had been filed before. In 1897, the 'Henry Classification System'was adopted throughout British India.

In 1901 Henry was recalled to London to head the Criminal InvestigationDepartment (CID) at Scotland Yard. He immediately set up the Fingerprint Bureauto record the identity of criminals in the belief that it would foil repeatoffenders. Before there was a reliable system of recording their identities, itwas common practice among career criminals to dodge a harsher sentence bymaking up an alias and pretending to be a first-time offender rather than arecidivist. In its first year alone the bureau cracked the pseudonyms of 632habitual criminals.

As is often the case with new developments, it takes a sensational case toestablish a new forensic technique in the public consciousness. Forfingerprints, that leap into the forensic limelight came four years later, in1905. On a Monday morning at the end of March, William Jones turned up for workat Chapman's Oil and Colour Shop in Deptford High Street, London. The16-year-old was surprised that, although it was half-past eight, the paint shopwas locked and shuttered. The manager and his wife lived above the shop andthey usually opened up for their early customers at halfpast seven. William wasconcerned that they might be ill; at seventy-one and sixty-five, it seemedplausible. When knocking produced no response, William barged the door with hisshoulder. It held fast. He stood on tiptoe and squinted through a gap in theshutters. At the back of the shop, by the fireplace, he could make out a loungechair tipped on its side.

Anxious now, William ran to fetch a friend. Together they hurried back andbroke the door down. William discovered his boss, Thomas Farrow, lyingunderneath the overturned chair, his bald head smashed open, blood seeping intothe ashes of the fireplace. During the subsequent autopsy, the pathologistexpressed the view that Thomas had been hit six times around his head and face,probably with a crowbar.

Sergeant Albert Atkinson was the first policeman on the scene and it was he whofound Ann Farrow in bed upstairs, grievously battered and unconscious, butstill clinging on to life. Atkinson noticed a moneybox lying open and empty onthe floor by the Farrows' bed. William explained that Mr Farrow normally tookthe box to the bank on Mondays to deposit the shop's weekly takings of around£10.

Melville Macnaghten, Edward Henry's successor as head of the CID, took chargeof the case. On his first day at Scotland Yard, in 1889, Macnaghten's new bosshad briefed him on the unsolved Jack the Ripper murders of the previous year.For the rest of his career Macnaghten kept photographs on his desk of Jack theRipper's mutilated victims to remind him to try harder. Still, like any experienceddetective, he had unsolved murders of his own. Just three days after he joinedthe CID, Macnaghten had found himself picking up pieces of a dismembered womanfrom along the riverbank. Her murderer was never found and the case becameknown as the 'Thames Mystery.'

Macnaghten was determined to solve Thomas Farrow's brutal murder. The crime hadshocked local residents. Deptford was a polluted and crowded district. Diseaseand crime were facts of life. But cold-blooded murder was rare.

Because the elderly couple had been found in their night-clothes, and becausethe pathologist estimated Thomas Farrow had died not long before Williamdiscovered his body, the police believed that Thomas had been tricked intoopening his front door early in the morning. They speculated that the assailanthad attacked Thomas immediately, then run upstairs to find his moneybox.Detectives found a pool of blood at the top of the stairs, so they deduced thatThomas had somehow managed to chase his attacker upstairs, where his wife layunprotected. They concluded that the assailant had callously finished him off,silenced his wife with complete ruthlessness, taken the money and run.

Macnaghten examined the moneybox carefully and noticed a greasy finger mark onthe underside of its inner tray. He picked the box up with his handkerchief,wrapped it in paper and took it to the Fingerprint Bureau. Macnaghten knew thathe was risking public ridicule because, although fingerprint evidence hadnailed a burglar, Harry Jackson, in 1902, it still had the taint of palmistryabout it. Not everyone had been convinced by the efficacy of fingerprinting atJackson's trial. After hearing the guilty verdict, someone signing off as 'ADisgusted Magistrate' wrote a letter to The Times: 'Scotland Yard, once knownas the world's finest police organisation, will be the laughing stock of Europeif it insists on trying to trace criminals by odd ridges on their skins.'

A CID assistant checks a new set of prints against Scotland Yard'sfingerprinting records in 1946

Charles Collins, head of the Fingerprint Bureau, looked at the tray with amagnifying glass and, from the size of the print and the slope of its ridges,recognised it as the right thumb of a sweating hand. He was also pleased to seesignificant differences when he compared it to the prints of Sergeant Atkinsand the Farrows, which he had also taken. Those differences would strengthenthe case against a suspect whose right thumb matched the print.

Although the bureau was only four years old, it had already amassed around90,000 finger and thumbprints, stored in its enormous wooden cabinet ofpigeonholes. Collins looked in the appropriate pigeonhole, but found no match.

The next blow to the investigation came five days later, when Ann Farrow died fromher injuries. Macnaghten had hoped she would return to consciousness anddescribe her attacker.

Then there was one of those breaks that sometimes come detectives' way thanksto the media. Having read about the murders in the press, a milkman came forward,saying that he had spotted two men leaving the Oil and Colour Shop at 7.15a.m., and had shouted out to them that they had left the front door ajar. Oneof them had turned round and said, 'Oh! It don't matter' before they had bothwalked off. The milkman described their appearance. One had a dark moustacheand was wearing a blue suit and a bowler hat; the other was wearing a brownsuit and a cap.

Then another eyewitness, a painter, came forward to explain why young WilliamJones had found the front door locked. The painter had glimpsed an old man withblood on his face shutting the door, at 7.30 a.m. Macnaghten reasoned thatThomas Farrow had somehow survived a second battering, this time at the top ofthe stairs, and, in a state of deliriousness, had staggered downstairs andclosed the door, before moving to the back of the shop and finally succumbingto his wounds.

A third eyewitness came forward: a woman who had seen two men matching themilkman's description running down Deptford High Street at 7.20 a.m. Evenbetter for the police, she recognised one of them. The man in the brown suit,she said, was 22-year-old Alfred Stratton. The descriptions of his companionmatched that of Alfred's 20-year-old brother, Albert. When the police went toquestion Alfred's girlfriend, she admitted that the day before the murdersAlfred hadn't had enough money to buy food; but that the day after, he had comeback with bread, bacon, wood and coal. This was enough for Macnaghten. TheStratton brothers were arrested a week after Thomas Farrow's murder.

But the ill luck that had dogged the investigation continued. Neither themilkman nor his assistant was able to pick the Strattons out in an identityparade. The brothers were filled with bravado; they joked that, when CharlesCollins took their fingerprints, it tickled them.

Collins had the last laugh, however. When he examined the fingerprints, hefound the mark on the moneybox matched Alfred Stratton's right thumb.

Still the prosecution knew they had their work cut out. Could a one-inch sweatmark persuade the jury? So much hung on this case: the conviction of a pair ofcold-blooded killers; the restoration of the reputation of Scotland Yard, sodamaged by Jack the Ripper's murders; and the acceptance of fingerprints as keyevidence. Both Macnaghten and the Metropolitan Police Chief, Edward Henry,understood how much was at stake.

Ironically, Henry Faulds, back from Japan, was ready to testify for thedefence. He had his own axes to grind. First of all, Scotland Yard had rejectedhis calls to set up a fingerprint bureau. Then it had opened one up based onthe Henry system, and refused to acknowledge Faulds' part in the development offingerprinting. Faulds was intent on claiming that not enough research had beendone to prove that a print from one finger could identify an individual beyonddoubt.

Charles Collins took the stand, several blown-up photographs under his arm. Heshowed the jury the smudged print taken from the moneybox, then the perfectlycrisp prints of the Farrows and Sergeant Atkinson. The jury didn't need muchexplanation to see that the prints were different from each other. Then Collinsproduced Alfred Stratton's thumbprint. The similarity was immediately obvious.Collins pointed out eleven separate points of similarity. The jurors weremesmerised.

When the defence cross-examined Collins he argued cogently that no two printsfrom the same finger are ever exactly the same, because of differences in thepressure and angle of contact. That was just as well, for, when the firstfingerprint expert for the defence, John Garson, took to the stand, heproceeded to discredit Collins' eleven points of similarity. The distancesbetween the points were slightly shorter in some and longer in others, he said.And the lines curved a tad differently as they reached the points. 

Buck Ruxton's fingerprints, taken in Liverpool Prison, 1936

The eminent counsel for the prosecution Richard Muir started hiscrossexamination of Garson by placing two letters down in front of him. Theyhad both been written by Garson on the same date. Garson had written one to theStratton lawyers, offering to testify for them. The other was to the Directorof Public Prosecutions, making the same offer. Muir's implication was thatGarson was a hired gun, willing to sell his services to the highest bidder. Tothis charge, Garson replied, 'I am an independent witness.' To which the judgeadded sternly, 'An absolutely untrustworthy one.' Garson got down from thestand with his credibility in tatters.

Henry Faulds was due to give evidence next. He was ready to deliver hisdevastating view that, after comparing many thousands of fingerprints, he couldnot definitively say that a single fingerprint could only belong to one personon earth. But the defence panicked lest Muir would undermine him assuccessfully as he had Garson. Faulds never got to speak.

After a two-hour deliberation, the jury came back with a guilty verdict. On 23May 1905, nineteen days after their trial had begun, the Stratton brothers bothwent to the gallows. The British judicial system had stepped into a whole newrealm of scientific evidence.

By 1905, fingerprint bureaus had been established in India, the UK, Hungary,Austria, Germany, Switzerland, Denmark, Spain, Argentina, the US and Canada,but their evidence had so far only been used as proof of guilt in Buenos Airesand London. The Stratton case demonstrated how powerful this evidence could be.In 1906, the year after that keynote trial, four other British men wereprosecuted based on crime scene prints. In the same year, the New York CityPolice Department (NYPD) introduced fingerprinting to police departments acrossthe US.

Edward Henry's system of classifying and finding matching prints remainedessentially unchanged until computers led to the automatic fingerprintidentification system in the 1980s. And so has the work of the fingerprintexaminer.

First of all an examiner needs to understand what they are looking at. The padsof our fingers are home to intricate patterns of ridges and valleys. If we coatone of those pads with ink and press it on to a piece of paper, the resultingridge pattern is instantly recognisable as the iconic image of a fingerprint.Our fingerprints are part of us from before birth; they first appear in thetenth week of pregnancy, when the foetus measures only 8 cm. As one of thethree layers of tissue that make up the foetus's skin – the basal layer –starts to grow at faster rate than the other two, ridges form to relieve theresulting stresses, 'like the buckling of land masses under compression'. Ifyour finger pads were flat, the pressure on the skin would be equal and theridges would be parallel. But because finger pads slope, ridges form alonglines of equal stress, most usually in concentric circles.

Ridge patterns also appear across the palms of our hands and on the bottoms ofour feet. Other primates have them, too, and evolutionary biologists believethere are good reasons for them. They help our skin to stretch and deform,protecting it from damage; they create valleys down which sweat can escape,reducing slipperiness when we hold things; and they give us more contact (andhence grip) with rough surfaces like tree bark.

When we touch a surface with a finger, the ridges leave their unique pattern onit. Even the prints of identical twins differ. In all the years thatfingerprinting has been practised, no one has yet found two identical clear andcomplete prints belonging to two different fingers.

To identify people from the marks they leave behind is simple enough in afamily setting. These small muddy prints belong to the toddler who forgot totake his shoes off. Those even smaller ones are the dog's. But deductions likethese are made from a small number of possible culprits, and the prints inquestion are patent – visible to the naked eye. Invisible, latent prints aremuch trickier. Substances like sweat, mud, blood and dust can produce bothpatent and latent prints. The more absorbent or irregular the surface, the moredifficult it is for a CSI to get a good print. Though it was once impossible totake prints from plastic bags or human skin, the technologies have improved andthere are now ways to do it.

British CSIs use various methods for recovering prints in a logical sequence,starting with the least destructive. The order of actions is set out in theHome Office's Manual of Fingerprint Development. First the CSI examinessurfaces for patent prints, like the bloody one on the doorframe at the Rojas'house; if necessary, she photographs them. Then she shines lasers andultraviolet light on to surfaces to illuminate latent marks, and make themphotographable. If special lighting doesn't work, she delicately brushes a darkpowder over the mark, takes a photo, and then presses an adhesive strip overit. Once the strip has been peeled off and pressed on to a white card, it isknown as a 'last'. This is Henry Faulds' classic way of recovering fingerprintsfrom a crime scene, and it is still the most commonly used method today. If aprint remains stubbornly invisible, as it might well do on more poroussurfaces, the CSI can use a range of chemicals which react with the salt andamino acids in human sweat, to make it visible.

The photographs and lasts are then sent to a fingerprint officer, who decideswhether they contain enough ridge detail to make identification possible. Ifthe print isn't too smudged or incomplete, the officer compares it tofingerprints of neutrals – that is, people who had every right to be there andwho are not suspects – including victims and police officers, before looking atprints from potential suspects. This is inevitably a subjective process. If theofficer judges that none matches, she scans the print and encodes it into ageometric pattern. Then she runs an automatic search using a national database,such as the UK's IDENT1, which holds the prints of around 8 million people.

IDENT1 is the modern equivalent of Edward Henry's pigeonholes. Both IDENT1 andthe FBI's database use a slightly modified version of Henry's classificationand identification system. A computer program asks the fingerprint a series ofquestions, such as, 'How many whorls have you got?' Each answer is given anumerical value – 'two whorls' is two points. The values are strung together togive the print an overall code. IDENT1 then compares this code to the 8 millionothers it contains, and presents the officer with the closest ten or so prints.

She must now judge if any of them are a positive match. This, again, is asubjective process. Once she has found a similarity in the overall pattern ofthe ridges, she concentrates on tiny distinguishing points known as 'minutiae',which include points where ridges start, end and join together; where they sitindependently; and where they form little bridges between two other ridges.

In 1901, when Scotland Yard established its Fingerprint Bureau, officers likeCharles Collins needed to find at lease twelve minutiae in agreement beforethey could testify to a match in an English court. In 1924 this was increasedto sixteen points, which was higher than in most other countries. At the timemost fingerprint experts thought that eight was enough. If an officer foundbetween eight and fifteen points, they usually reported it to the police,because it might give them a valuable lead. But by 1953 all the UK policeforces had adopted the 16-point standard.

Since the Stratton brothers case, belief in fingerprinting has gone fromstrength to strength amongst the civilians, judiciary and police forces of theworld. And for many, including large numbers of experts, it has an aura ofabsolute infallibility. As Jim Fraser writes in Forensic Science (2010): 'Inthe view of most marks examiners, identification of an individual byfingerprints can be done unequivocally, that is, with 100% certainty.'

When a print is clear, the chances of an officer making a wrong call are nextto nothing. But when a print is smeared, overlaid with other marks, made inblood, one officer may see points of agreement that another does not. A case in1997 tested the subjective nature of fingerprinting to breaking point. On 6January, the body of Marion Ross was discovered in her house in Kilmarnock,Scotland. She had been the victim of a horrific attack: her injuries includedmultiple stab wounds and crushed ribs, and a pair of scissors was foundembedded in her throat. CSIs set to work recovering evidence, and found morethan 200 latent finger marks in Marion's house, which they sent to the ScottishCriminal Records Office to be checked off against the neutrals – paramedics,doctors, police officers.

The finger mark that became the eye of a perfect storm was a left thumbprint onthe bathroom doorframe. Despite it being quite badly smudged, a fingerprintofficer confidently identified it as belonging to 35year-old DetectiveConstable Shirley McKie, who was supposed to have remained outside the house topreserve the scene while investigators gathered evidence inside. She would havehad to leave her post in order to touch the door; that would have been grossmisconduct.

Officers are thoroughly trained how to treat crime scenes; CSIs always wearprotective gloves so they don't damage the fragile traces left by criminals.Because of the seriousness of the case, a further three experts at the ScottishCriminal Records Office examined the thumbprint, and confirmed it to beMcKie's. It looked as if the detective had indeed abandoned her post.

Meanwhile, the prime suspect for the murder had been identified as DavidAsbury, a 20-year-old handyman. Investigators had found his finger marks inMarion's house, and her prints on a tin box in his. Asbury explained that hehad recently done a job in Marion's house, hence the marks. But detectivesthought there was enough evidence to arrest him.

At Asbury's trial, McKie testified that she had not been inside Marion Ross'shouse at any time, so the thumbprint could not have been hers. All the otherfifty-four officers who had worked the crime scene corroborated her assertion.Nevertheless she was suspended from Strathclyde Police, and eventuallydismissed.

But that wasn't the end of her nightmare. Early one morning in 1998 ShirleyMcKie was arrested. She had to get dressed under the watchful gaze of apolicewoman. The officers took her to the police station where her own father,Iain McKie, had been commanding officer. She was strip-searched and shut in acell. She learned she was to be charged with perjury, which carries the threatof an 8-year jail sentence. Her father's long and distinguished police careerhad left him convinced of the integrity of fingerprint evidence. It was easierto believe that his own beloved daughter was lying than to doubt the experts.'People have been hung on a fingerprint,' he reminded her.

In May 1999 Shirley McKie was brought to trial at the High Court of theJusticiary. Two American experts who had examined the thumbprint maintainedthat the print was not hers. One said the 'obvious' differences took 'onlyseconds' to see. On this evidence the jury found McKie not guilty of perjury. InAugust 2002, David Asbury's conviction for murder was also quashed by the Courtof Criminal Appeal in Edinburgh – because of faulty fingerprint evidence. Hehad spent three and a half years in jail.

After Shirley McKie's innocence had been established, the Scottish CriminalRecords Office and four police officers of the Strathclyde Police were accusedof misconduct. McKie subsequently launched a damages case and in 2006 receiveda £750,000 settlement.

But by then she had lost the job that she loved, spent years working in a giftshop and suffered from severe depression. Iain McKie now travels the worldcampaigning for better expert evidence to be produced in courts of law, andwarning people about the entrenched attitudes of fingerprint experts.

In 2001, the 16-point standard was scrapped in England and Wales, partlybecause of the McKie–Asbury fiasco, and partly because it wasn't really astandard. If fingerprint officers had found fourteen points they wouldsometimes search for two more, to get a 'match'. They were looking forsimilarities rather than differences, and that was dangerous. Since the sixteenpoints were scrapped no numerical standard has existed. But other experts veryseldom challenge the individual decisions of fingerprint officers.

Catherine Tweedy is one of the very few people alive today whose job it is tochallenge fingerprint officers. On first impressions, she seems like the kindof teacher kids love because she brings out the best in them – interested,encouraging, knowledgeable. But five minutes in her company reveals somethingelse: a steely intelligence devoted to ferociously logical argument and apassion for getting things right. She has completed a range of fingerprintcourses in the UK and abroad, including 'Advanced Latent Fingerprinting' withthe Miami Police in Florida. She currently works as a fingerprint expert for aforensic consultancy based in Durham, mostly working as a defence expert, whereit is her job to double-check a proportion of the fingerprint identificationsmade in the UK – albeit a smaller proportion than she'd like. 'I've been doingthis from the mid1990s,' she says, 'and I've come to it as a scientist. I tearmy hair out at the assumption people have made right the way through that it isan absolute rock solid science. It's not a science at all. It's a comparison.'The rhetoric used to support forensic fingerprinting has always been scientificin tone. But Catherine Tweedy has spent twenty years trying to remind peoplethat moving forward along a road to certainty is not the same thing as arrivingat it; and that going in reverse is also possible.

In 2006, the year of McKie's settlement, Scotland followed England and Wales inscrapping the 16-point system. In 2011 the results of a public enquiry into theMcKie–Asbury fiasco were published. It put the misidentifications down to'human error' and not to misconduct on the part of the Strathclyde Police. Itrecommended that fingerprint evidence should from now on be regarded as'opinion evidence' not fact, and thus should be treated by courts 'on itsmerits'.

But this message has not trickled down to all fingerprint officers, saysCatherine Tweedy: 'They are not being trained to think that an opinion is anopinion. Once you are trained to see things as facts it is extremely difficultto be pulled back to understand that there are shades of grey. You can't be 100per cent certain in a lot of cases because you only get a fraction of afingerprint.'

Even when the fingerprint is correctly matched with an individual, mistakes cansometimes be made when the investigator tries to work out what this means. Inone of her first cases, Catherine dealt with Jamie, a 14year-old boy chargedwith burgling a house in Northern Ireland. His handprint had been recovered ona windowsill in the bathroom. When she met him, he said he'd never entered thehouse in his life. Catherine went inside and could see why that might be true.The house was such a stinking mess that it was tough to carry out any kind ofexhaustive examination. When she looked at the handprint, she saw that it was aclear match for Jamie's. But if someone had climbed in or out of the bathroomwindow, they would have left footprints in the bath or sink as well asdisturbing all the junk underneath the windowsill. And there was no evidence ofthat.

The SOCO hadn't been into the other rooms or looked at the two external doors.Catherine carried out her own examination and she couldn't find any otherevidence to link him with the inside of the house.

Fired up by Catherine's work, Jamie's defence team found out that the owners ofthe burgled house had heartlessly kicked their daughter out of the house and onto the streets on her sixteenth birthday. She had gone to stay with friends fora couple of weeks. Then, when she knew her parents were out shopping, she hadreturned home with her key, walked through the front door and collected herghetto blaster, moneybox, some of her clothes and a few of her videos.

When her parents returned home they spotted the missing items and called thepolice to report a burglary. The investigation had begun and ended with thepalm print in the bathroom. No further questions had been asked. When CatherineTweedy had Jamie's friends questioned they told her that they used to play agame called Pirates round the back of the house. Pirates is a variation on thegame of tag where players avoid being tagged 'it' by getting both feet off theground. Jamie, it transpired, was a good climber. His best trick was to shimmyup the drainpipe and hang onehanded off the bathroom windowsill. WithoutCatherine's tenacity, his agility might have landed him in jail.

Some fingerprints are lifted in far more horrific contexts. On 11 March 2004,during the peak of the rush hour, ten bombs exploded simultaneously on fourcommuter trains in Madrid, killing 191 people and wounding 1800. The FBIsuspected Al-Qaeda involvement.

The Spanish police discovered an abandoned set of detonator caps inside aplastic bag that bore a single, incomplete finger mark. This was run throughthe FBI database, which showed up twenty possible matches.

Spanish forensic experts search for clues in a blasted train carriage outsideAtocha train station. The

2004 terrorist attacks were among the worst in Spain's history

One of those possibles was Brandon Mayfield, an American-born lawyer living andpractising in Oregon. He was on the FBI fingerprint database because he hadserved in the US army. But, more significantly in counter-terrorist terms, hehad married an Egyptian and converted to Islam. He had defended one of thePortland Seven, a group of men who had tried to go to Afghanistan to fight forthe Taliban, albeit in a child custody case. But he also worshipped at the samemosque as them.

The FBI decided that Brandon Mayfield was implicated in the bombing, eventhough his fingerprints were not an exact match, his passport had expired andthey could find no evidence that he had travelled abroad for years. They beganwatching him and his family.

Despite the Spanish police insisting that the fingerprint evidence should berejected, the FBI agents tapped Mayfield's phone, broke into his house andoffice, went though his desk and financial records, examined his computers andtailed him. When Mayfield realised he was under surveillance he panicked, sothe FBI detained him to prevent him making a run for it. Two agonisingly longweeks passed before the Spanish police matched the fingerprints to the realculprit, an Algerian man named Ouhane Daoud.

Mayfield sued the US government for wrongful detention, and in 2006 received aformal apology and a $2 million settlement.

The FBI later acknowledged one of the problems in the handling of the Mayfieldcase was that their fingerprint experts had failed to separate the analysis andcomparison stages of their examination. First of all the expert should analysethe mark in detail, describing as many minutiae as she can. Only afterwardsshould she examine possible matches and carry out a comparison. When analysisand comparison happen simultaneously, experts run the risk of finding matchingminutiae because they are looking for them. In the view of Itiel Dror, acognitive psychologist at University College London, 'The vast majority offingerprints are not a problem, but even if only 1 per cent are, that'sthousands of potential errors each year.'

An American experiment in 2006 showed that even experienced fingerprint expertscan be swayed by contextual information. Six experts were shown marks that eachone had analysed before. But this time they were given certain details aboutthe case – that the suspect was in police custody at the time the crime wascommitted, for example, or that the suspect had confessed to the crime. In 17per cent of these secondary examinations, the experts changed their decision inthe direction suggested by the information. In other words, they couldn'tdivorce themselves from the context and judge objectively. This kind of bias isless likely to occur in the UK, where forensic divisions are separated fromothers divisions in most police forces.

In spite of the question marks raised by experts such as Catherine Tweedy,courts around the world continue to treat fingerprint as infallible andsolitary finger marks still send people to jail. In her popular book TheForensic Casebook (2004), N. E. Genge states that 'Examiners don't think in anypercentages except 100 and 0.' But Christophe Champod, a Swiss expert inforensic identification, calls for fingerprint evidence to be treated in termsof probabilities – bringing it in line with other forensic disciplines – andthat examiners should be free to talk about probable or possible matches. Hehas also called for the overall importance of fingerprinting to be downgraded:'Fingerprint evidence should be expressed by fingerprint examiners only ascorroborative evidence.'

If forensic science were a family, fingerprinting would be the greedygrandfather, hanging on to the best armchair, trying to exercise the sole rightto pass judgement, unaware that the times they are a-changin'. Only when therest of the family understands that he sometimes gets people and places andanecdotes mixed up can his wisdom be treated with appropriate circumspection;then his contribution to the family can be regarded as a healthy and balancedone.