2
Away With the Fuzz The introduction of numerical evidence in court proceedings has been with us for a long time. Statutory rules on the amount of meat in sausages and fat in ice cream co-exist with rules on the distance of car headlamps from the road surface and the time yo3 can leave your car parked in certain places. Similarly, evidence on the level of alcohol in the blood is now commonplace and there are few judges or practised advocates in the criminal courts who have not heard the forensic biologist testify to the bloodgroups present in bloodstains and their relative commonness in the population at large. This trend towards more numerical assistance for the courts is one in which this journal has been amongst the pioneers. An early editorial probably represents the first published argument for a statutory blood alcohol level being used as an offence per se as distinct from the analytical figure being considered as evidence that an accused person is "under the influence of alcohol", what- ever that may mean (Kind, 1963). Similarly our pages have carried enormous amounts of data on blood group frequencies in selected populations, mercury levels in fingerprint officers and iron and calcium levels in the glass that showers the housebreaker as he smashes a window to gain entry. Inevitably the growth of this useful (and intellectually fascinating) trend has attracted the attention of that guardian of numerical significance, the statistician. Whilst many of his contributions have been useful some of his more extreme pronouncements have been less so. An example of the latter variety is one where it is proposed that the presumption of innocence should be abolished and, thereafter, a probability of guilt, calculated from a statistical assessment of the worth of all the evidence, should be combined with a factor which assesses the harm done to Society by convicting the innocent, relative to that caused by absolving the guilty. The final numerical result would then be used to determine the sentence of the courts (Lindley, 1977). Such a suggestion stimulates the imagination. One can imagine certain politicians attracting such adverse social utility factors that the merest whiff of a suspicion of kicking the neighboxr's cat would be sufficient to warrant life imprisonment (not altogether a bad thing in some cases). But apart from these bizarre diversions there remains a hard kernel of opinion which seeks not only to quantify evidence at the scientific stage, but also to assess it at the legal stage, in the absence of statutorily defined figures. Thus no longer would the qualified expert state the scientific facts and give his opinion on their worth under the alleged circumstances. He would be reduced to stating his findings in probabilistic terms surrounded by a haze of statistical fuzz. Human nature and professional groups being what they are, we would soon see statisticians add the job of legal assessors to that of scientific assessors. Thus one more function of commonsense will be superseded. The monitoring of discrete scientific procedures, by numerical criteria in the laboratory, is invaluable. In this way not only do we identify the specific and accurate method, the able experimentalist and the unsuspected phenomenon but we have a strong weapon to expose the charlatan. But to take these methods out of the laboratory and into the court of law with the hope that they will be of equal effectiveness in assessing the evidence of lay and expert witnesses alike ignores the fact that procedures in a court of law are several orders of magnitude more complex than in the laboratory. Perhaps one day it will come to pass that sociological phenomena will be sufficiently well understood to apply these procedures in the legal process

Away With the Fuzz

  • Upload
    ld

  • View
    235

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Away With the Fuzz

Away With the Fuzz

The introduction of numerical evidence in court proceedings has been with us for a long time. Statutory rules on the amount of meat in sausages and fat in ice cream co-exist with rules on the distance of car headlamps from the road surface and the time yo3 can leave your car parked in certain places. Similarly, evidence on the level of alcohol in the blood is now commonplace and there are few judges or practised advocates in the criminal courts who have not heard the forensic biologist testify to the bloodgroups present in bloodstains and their relative commonness in the population at large.

This trend towards more numerical assistance for the courts is one in which this journal has been amongst the pioneers. An early editorial probably represents the first published argument for a statutory blood alcohol level being used as an offence per se as distinct from the analytical figure being considered as evidence that an accused person is "under the influence of alcohol", what- ever that may mean (Kind, 1963). Similarly our pages have carried enormous amounts of data on blood group frequencies in selected populations, mercury levels in fingerprint officers and iron and calcium levels in the glass that showers the housebreaker as he smashes a window to gain entry.

Inevitably the growth of this useful (and intellectually fascinating) trend has attracted the attention of that guardian of numerical significance, the statistician. Whilst many of his contributions have been useful some of his more extreme pronouncements have been less so. An example of the latter variety is one where it is proposed that the presumption of innocence should be abolished and, thereafter, a probability of guilt, calculated from a statistical assessment of the worth of all the evidence, should be combined with a factor which assesses the harm done to Society by convicting the innocent, relative to that caused by absolving the guilty. The final numerical result would then be used to determine the sentence of the courts (Lindley, 1977). Such a suggestion stimulates the imagination. One can imagine certain politicians attracting such adverse social utility factors that the merest whiff of a suspicion of kicking the neighboxr's cat would be sufficient to warrant life imprisonment (not altogether a bad thing in some cases).

But apart from these bizarre diversions there remains a hard kernel of opinion which seeks not only to quantify evidence at the scientific stage, but also to assess it at the legal stage, in the absence of statutorily defined figures. Thus no longer would the qualified expert state the scientific facts and give his opinion on their worth under the alleged circumstances. He would be reduced to stating his findings in probabilistic terms surrounded by a haze of statistical fuzz. Human nature and professional groups being what they are, we would soon see statisticians add the job of legal assessors to that of scientific assessors. Thus one more function of commonsense will be superseded.

The monitoring of discrete scientific procedures, by numerical criteria in the laboratory, is invaluable. In this way not only do we identify the specific and accurate method, the able experimentalist and the unsuspected phenomenon but we have a strong weapon to expose the charlatan. But to take these methods out of the laboratory and into the court of law with the hope that they will be of equal effectiveness in assessing the evidence of lay and expert witnesses alike ignores the fact that procedures in a court of law are several orders of magnitude more complex than in the laboratory.

Perhaps one day it will come to pass that sociological phenomena will be sufficiently well understood to apply these procedures in the legal process

Page 2: Away With the Fuzz

(although many may believe that such an exposure of human beings may be more than human beings can stand). Then shall we see the expert witness as part of a chimera with his own instruments, stripped of all integrity (for who requires integrity now?) and the whole assessed by the statistician on the bench, ephemerally assisted by a judge.

Indeed it may come to pass but, until that day arrives-away with the statistical fuzz !

L.D. References KIND, S. S., 1963, Driving over the Level, J. Forens. Sci. Soc., 4, 1. LINDLEY, D. V., 1977, Probability and the Law, The Statistician, 26, 203.