Read an Excerpt
Legal Alchemy
The Use and Misuse of Science in the Law
By David L. Faigman W. H. Freeman and Company
Copyright © 2000 David L. Faigman
All rights reserved.
ISBN: 978-1-4299-2642-3
CHAPTER 1
FROM THE DARK AGES TO THE NEW AGE
The Strange History of Science in the Law
In the beginning God created the heaven and the earth.
And the earth was without form, and void; and darkness was upon the face of the deep. And the Spirit of God moved upon the face of the waters.
And God said, Let there be light: and there was light.
— Genesis
At about one-hundredth of a second after the beginning, ... the temperature of the universe is 100,000 million degrees Kelvin (10 K)....
It is filled with an undifferentiated soup of matter and radiation, each particle of which collides very rapidly with the other particles. Thus despite its rapid expansion, the universe is in a state of nearly perfect thermal equilibrium. The contents of the universe are therefore dictated by the rules of statistical mechanics, and do not depend at all on what went before.
— STEVEN WEINBERG, The First Three Minutes
On February 11, 1900, a jury returned a verdict of guilty and a sentence of death in the case of the People of New York v. Roland Molineux. Molineux was charged with the murder of Katherine Adams, who had died as a result of ingesting cyanide, which had been added to a popular headache medication and sent to her nephew through the mails. Her nephew, Harry Cornish, worked at Molineux's athletic club and had recently had a heated dispute with him. The poison, the police surmised, was intended for Cornish. Adams's headache had made his fate hers.
According to the police, this form of mail order poison was Molineux's modus operandi; he had allegedly employed the scheme successfully six weeks before. Early in 1898, Molineux had competed with Henry C. Barnet for the affections of Blanche Cheeseborough. Later reports would indicate that the woman was partial to Barnet. In October 1898, Barnet received a package of Kutnow powders through the mail from an anonymous sender. Shortly after taking some of the powders, he fell desperately ill. Although his physician attributed the illness to diphtheria, Barnet maintained that the cause was "those damned Kutnow powders." It was later determined that the powders contained cyanide of mercury. Barnet died in November. Nineteen days later, Molineux married Blanche Cheeseborough.
Molineux was never charged with the death of Henry Barnet. Nonetheless, the details of Barnet's death and the innuendo surrounding it became a major part of the state's case against Molineux in the trial involving Katherine Adams's death. More damning, however, was the discovery of the private letter boxes rented in the names of H. Cornish and H. C. Barnet. At trial, the person who had let the Barnet box, Nicholas Heckman, would testify — fortified by a promised reward from The World (a leader in the yellow journalism of the day) — that Molineux had rented the box from him. Several newspapers soon discovered that these boxes were used to receive sundry pharmaceutical products, especially medications for sexual debility. Molineux denied any connection with these postal boxes.
The core of the state's case, however, came in the form of the "science" of forensic document examination. At trial, the state introduced eighteen witnesses who specialized in handwriting identification to testify that the Cornish and Barnet letters sent to the pharmaceutical companies were written by Molineux. These experts testified further that the writing in the anonymous note that accompanied the package that contained Katherine Adams's death warrant was also in the hand of Roland Molineux. The experts thus were able to provide the empirical link between Molineux and Adams's and Barnet's deaths.
The prosecution's handwriting experts were a motley crew. None had any formal training in handwriting identification. Apparently, American universities did not offer such programs in the late nineteenth century. Most still do not. In addition, none of the prosecution's experts had conducted any research on handwriting patterns in the population or offered data indicating their proficiency in the delegated task. Several were employed as bank tellers and cited their experience and responsibility for evaluating the veracity of signatures for their respective employers. Other than that, these experts had conducted no systematic and rigorous study of handwriting. Fourteen of the handwriting analysts were professional experts, however, with considerable experience testifying in court, though apparently none in checking the accuracy of the conclusions to which they testified. They researched handwriting much as someone might study Milton or Shakespeare. And, as with many "experts" on literature, such musings gave them settled opinions they sought to share with the world. All the state's experts testified confidently to the conclusion that Molineux had penned the questioned documents.
The handwriting experts employed various methodologies to establish the identity of the author of the letters. The bank tellers tended to rely on gestaltlike subjective judgments from "close study" of the questioned writings. As Gilbert B. Sayres, a bank teller for thirteen years, explained, "After I began to make it a study I studied it conscientiously and for a long time, and the more I studied it the more convinced I became of the similarity ..., the characteristics being the same in both."
The professional experts, on the other hand, employed a more technical and seemingly more sophisticated method of comparison. Persifor Fraser, for instance, a geologist and chemist and a "student of handwriting for about twenty-one years," testified in somewhat greater detail about the comparison he made between Molineux's handwriting and the note that accompanied the "poison package" received by Cornish. It is worth quoting at some length from Persifor Fraser's testimony. (The numbered exhibits he refers to were known handwriting exemplars provided by the defendant; Exhibit A was the note found with the poison package.)
Roland Molineux wrote the address on the wrapper, Exhibit A, because, firstly, there are twenty-one characteristics in the conceded writings which are visibly on Exhibit A, secondly, the patterns of a great many letters on the unnumbered [sic] exhibits accord very closely with the patterns of the letters on Exhibit A, and these differences, which exist, are those which would be naturally adapted for disguise; thirdly, the microscopic structure of the ink lines in the numbered exhibits, the conceded writings, agree with the microscopic structure in the ink lines of the Exhibit A in two respects; firstly, in the swelling and tremors, deviations through tremor of the line when highly magnified; secondly, in the characters of the margins of the lines, the edges of the ink lines ... [and] in searching for some characteristic which was not simply a question of comparison of model or a similarity, it occurred to me that these differences were more than could be accounted for by accident.
There are a number of aspects of the nineteenth-century methods of handwriting identification that seem to be problematic. Most striking, perhaps, is how unscientific the process appears. The experts all knew what results would confirm the hypotheses they were testing. Experimenter bias, usually avoided at all costs in empirical research, was palpable here. The experts also approached the samples looking for confirming instances and were quick to discount or dismiss differences as "adapted for disguise."
The practice of searching a multitude of exemplars for similarities actually turns the scientific method on its head. These experts seemed to take the view that if you have a hundred points of comparison and five constitute "matches," this observation supports the conclusion that the two samples came from the same hand. A less biased method would ask what percentage of matches would be expected if the person did not write the disputed document. A comparison of this number to the number discovered would provide a more accurate statement concerning likely authorship.
The subjective element manifest in the handwriting experts' examinations is tantamount to fraud. Finally, the proficiency of the handwriting experts was never rigorously tested, nor did the courts require tests. In the late nineteenth century, then, although it aspired to scientific status, forensic document examination embraced none of the rigorous methods that would allow us today to label it as a science.
Ultimately, however, what should be most disturbing for the modern reader is the fact that handwriting experts today employ virtually the same methods they used at the turn of the nineteenth century. The lack of empirical validation and failure to conduct proficiency testing, the failure to "blind" testers to expected results and the overwhelming subjective component in the conclusion that the samples "match," are as much a part of handwriting identification analysis today as they were one hundred years ago. In fact, today's handwriting analysts rely on essentially the same authorities as did their nineteenth-century brethren. In the hundred years since Molineux was convicted, largely on the basis of the state's handwriting experts, we have moved from a Newtonian universe to a universe of Einstein, Bohr, and Hawking. We have moved from the biology of Darwin to the DNA helix of Watson and Crick. Cars have replaced the horse and carriage, and planes, spaceships, and satellites fill the sky. And through it all, handwriting experts continue to count similarities and note "swelling and tremors," just as they did a hundred years ago. And courts continue to qualify them as experts.
I do not mean to suggest that handwriting identification cannot be done or that it is not done successfully at times. Common experience suggests that people's handwriting does vary, and there is nothing implausible about developing techniques to identify individual differences. But common experience and lack of implausibility do not make a science. If it did, we would continue to believe that the earth is the center of the universe and that bleeding with leeches is an effective medical therapy. Most aspects of handwriting analysis are subject to tests. But there has been virtually no systematic attempt to study intrawriter variation or interwriter variation, and there have only been sporadic and insufficient attempts to study the reliability of the practitioners of this craft.
An assortment of reasons explain this inertia in the science of handwriting identification analysis. All of these reasons, however, stem from a single cause: market failure. Unlike many other sciences, the primary market for handwriting experts is the law. Neither do they compete among themselves to discover new insights about handwriting comparison, nor do their discoveries have value to other fields. They are a discrete and insular sect of self-validating specialists. They are not trained in the scientific method and they have little clue how to test their claims of expertise. So long as their customers, the courts, keep buying the old model there is no need to come up with anything new. It is as if they began making the Edsel and over the years nondiscriminating car buyers just kept plunking down money for the same old car.
On appeal, Roland Molineux's conviction was overturned largely because the trial court had permitted proof of the defendant's alleged involvement in the Barnet death, a killing with which he was not charged. The appellate court did not object to the surfeit of handwriting testimony. Upon retrial, however, the new judge had little patience for this long queue of handwriting specialists. At one point he exclaimed in exasperation, "What! Another expert? — Well, I suppose we must hear him. Make it quick." In addition, the defense, which had presented no evidence at all in the first trial, responded in the second with its own group of handwriting experts. Not surprisingly, these experts, bought by the defense, testified that the similarities identified by the prosecution were what might be expected in the handwriting of a number of strangers and that the poison package note was not penned by Molineux. This they were confident was true, since the poison package note was written in a better hand than Molineux's. The defendant's expert concluded that one cannot disguise one's handwriting in a better hand than was natural. While this explanation "smells of the lamp" as much as the prosecution's experts' testimony, it apparently had the desired effect on the second Molineux jury. It took the jury just twelve minutes to return a verdict of "not guilty."
Bedrock Principles and Changing Times
As the case of Roland Molineux illustrates, the law does not always stay current with changing science and technology. And this is true at much deeper levels than the courts' failure to insist on validation of handwriting expertise.
Consider a simple thought experiment. Suppose we were able to pluck a Greek citizen from the Athens of Aristotle's time, around 350 B.C. and place him in modern Washington, D.C. That person would probably be able to understand the basic proceedings and decisions made in an assortment of modern legal contexts, from the courtroom to Congress. Certainly, particular mechanisms or procedures have changed, and particular forms of legal relationships have appeared or disappeared, as they have come and gone throughout history. Slavery is gone and democracy is more inclusive than ever. But the rhetoric of politics, the daily business of government, and the resolution of legal disputes have remained much the same through the ages. Our Greek time-traveler, however, would almost surely be completely overwhelmed by our science and technology. Just the ride in the Metro from his hotel to Congress would probably kill him.
This difference in the rates of change between law and science says much about how law and science approach their respective tasks and thus how they inevitably relate to one another. The law's prestige depends largely on adhering to the traditions of the past, while science's prestige turns on how swiftly it advances into the future. But their incompatibility is even more fundamental. Science and law approach the world in profoundly different ways. Even brief reflection reveals stark differences in perspective between the two. Science explores what is; the law dictates what ought to be. Science builds on experience; the law rests on it. Science welcomes innovation, creativity, and challenges to the status quo; the law cherishes the status quo. Science assumes behavior is largely determined by biology and experience; the law typically assumes man has free will.
We might be tempted to conclude that, institutionally, these two professions are so alien to one another that there is little prospect of their ever finding accommodation. In some ways, the personalities of law and science are like those of the tortoise and hare in the familiar Aesop fable. The law's tortoise moves forward deliberately, almost reluctantly, while science's hare bounds forward with great enthusiasm. But in other ways law and science are nothing like the characters in the fable. They are not racing against one another — at least there is no reason for them to be — and it is not even clear that they are going in the same direction. It appears certain, however, that wherever law and science are going, they will not arrive together.
The past can tell us much about these disciplines' respective characters. And more to the point, it can give us insights into their relationship with each other. It turns out that they have more in common than might at first appear to be the case. In fact, in a twist that is reminiscent of a Dickens novel, they are blood relations. The question, though never in doubt in a Dickens story, is whether law and science can live together happily ever after.
Understanding Nature — Including the Nature of Man
Human understanding of the world we inhabit has not always been specifically scientific as we understand the term. To begin with, the word science had a different historical meaning. In ancient times, the word science referred to any body of knowledge that resulted from systematic and rigorous study; it was not limited to the world of experience. Even as late as the nineteenth century, legal scholars presumed that the law itself could be studied as a science. This allowed medieval theologians to claim the scientific mantle.
(Continues...)
Excerpted from Legal Alchemy by David L. Faigman. Copyright © 2000 David L. Faigman. Excerpted by permission of W. H. Freeman and Company.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
