Life in crime

'Give every man thine ear,' Polonius advised his son in Shakepeare's Hamlet. He might have added, 'but be careful not to eavesdrop behind arrases or press your ear against any windows.'

In the case of R v Kempster [2008] EWCA Crim 975, the Court of Appeal gave guidance on the new area of forensic science '“ that of ear-print comparison.

In 2001, Mr Kempster was convicted of three counts of burglary and one count of attempted burglary. In relation to count 1 (the subject of the successful appeal), the appellant was alleged to have burgled the home of an 89-year-old woman who was asleep in her home at the time. She awoke to find a man in her room. His head was covered with a hood. He took £45 from her purse and then fled. The old lady was later unable to identify the defendant at a video identification parade. The mainstay of the prosecution's case, therefore, was an ear-print found on the outside of a rear window to the premises '“ the suggestion being that the burglar had pressed his ear to the window to check whether anyone was at home.

At trial, the prosecution presented evidence by a fingerprint expert, who had also been trained in ear-print analysis. The expert's opinion was that ear-prints were as reliable as fingerprints, in that no two ears left the same mark and that in her view the ear-print found on the window pane was an exact match with that of Mr Kempster. The appellant was ultimately convicted and sentenced to a total of 10 years' imprisonment.

In 2003, following the case of Dallagher [2002] EWCA Crim 1903, Mr Kempster appealed. He called fresh scientific evidence which concluded that in the light of the fact that this area of science was in its infancy, and still developing, ear-print comparison was a valuable investigative tool and could properly be used to exclude a person as a suspect, but that it could not provide a positive identification of a suspect. The Court of Appeal rejected that submission and said that the evidence was admissible and could be used by the jury to conclude that it was indeed the appellant who was the maker of the mark. However, the court certified a question of general public importance for the House of Lords as to the admissibility of this new science. Their Lordships declined to grant leave to appeal.

The appellant then applied to the Criminal Cases Review Commission (CCRC). The application was based on an expert report by a mathematician and statistical analyst, Dr Ingleby, who had been closely involved with a European research project called Fear ID. The project had been established to evaluate the use of ear-print evidence which is widely used in continental Europe, especially in Holland. Fear ID is the most detailed examination to date of this infant science. On the basis of his report, the CCRC concluded that there was a real possibility that the conviction on count 1 might be overturned and so referred the matter back to the Court of Appeal.

The Court of Appeal heard evidence from both Dr Ingleby and the Crown's original expert at trial. It appears from the judgment that the evidence was hotly contested, with the Crown maintaining that the ear-print comparison provided a reliable match.

The experts agreed that ear-prints present a different and more difficult problem that fingerprints. Ears are cartilaginous structures which are flexible and deform when subjected to pressure. Further, ear-prints are left by those who are listening for something or someone by pressing their ear against a surface, and will not necessarily remain motionless but may adjust their position, thereby further distorting the shape of the ear and the mark it leaves. Dr Ingleby argued that Crown's case relied only on gross detail (shape, size and folds) which was of insufficient quality to provide a match. One had to go on to examine the anatomical minutiae, such as notches, nodules and creases in the ear structure.

Lord Justice Latham, giving the judgment of the court, was only just persuaded that the scientific evidence in this case lacked sufficient certainty and so quashed the conviction. He said that: 'Ear-print comparison is capable of providing information which could identify the person who has left an ear-print on a surface. That is certainly the case where the minutiae can be identified and matched. Where the only information come from the gross features, [it is not that case] that no match can ever be made, but there is likely to be less confidence in such a match because of the flexibility of the ear and the uncertainty of the pressure. . . applied. [The prosecution's expert] still remains of the view that gross features are capable of providing a reliable match . . . we are of the view that [this] can only be the case where the gross features truly provide a precise match.'

In Kempster's case, the court concluded that the gross features did not provide a precise match. 'The differences may well be explicable by differences in pressure or movement, but the extent of the mismatch is such as to lead us to the conclusion that it could not be relied on by itself as justifying a verdict of guilty.'

No magic bullet

This is a cautionary case. Every generation throws up new forensic techniques which, though in their infancy, are presented as the magic bullet. First it was fingerprints and blood groups, then DNA and now ear-print analysis. No doubt in the future, with the advent of ID cards, iris scans will be presented as infallible. And yet as those scientific techniques progress from infancy through adolescence to maturity, their fallibility is exposed. Meanwhile, those convicted on the basis of such techniques pay the price of progress. Kempster's case took seven years and two appeals to put right.

So caution is required to guard against forensic fashions. Perhaps Polonius was right when he said: 'Give every man thine ear, but few thy voice / Take each man's censure, but reserve thy judgment.'