Yes, I know the title is provocative. I hope it has earned your undivided attention because we all count on you to do what is right.
Several of your own — other attorneys — wrote this statement: “Forensic science expert evidence that is erroneous (that is, honest mistakes) and fraudulent (deliberate misrepresentation) has been found to be one of the major causes, and perhaps the leading cause, of erroneous convictions of innocent persons”1.
Unlike the natural and physical sciences that have been able to self-correct over many centuries, leading to marvelous discoveries and technological breakthroughs, forensic science or science applied to the courtroom setting has not been able to self-correct. Year after year, we hope that scientists will provide reliably meaningful answers to important questions raised in the courtroom, but year after year, we continue to have injustices perpetrated and error promulgated.
There is a reason for this, and that reason is understood when one understands the basics of logical inference.
A Primer in Logical Inference
An inference is “a conclusion reached on the basis of evidence and reasoning”2. Scientists, as courtroom experts, are expected to make inferences, but only certain forms of inference are reliable for courtroom testimony.
There are three forms of logical inference: induction, deduction, and abduction.
Induction is “the inference of a general law from particular instances”2. Logician, mathematician, philosopher, and scientist Charles Sanders Peirce illustrates induction3:
Case: These beans are [randomly selected] from this bag.
Result: These beans are white.
Therefore Rule: All the beans from this bag are white.
In the courtroom, this is the inference made from “experience.” “I have performed over 5000 autopsies,” the forensic pathologist may say, “and I have always seen…” or “I have never seen…” (Fill in the blank). Although this form of inference has its limitations and problems, it is acceptable to offer as evidence to a jury. After all, experts are approved or denied by the court as experts on the basis of their experience; nevertheless, it is important to note what the recent president of the American Academy of Forensic Sciences properly recognizes:
“The six most questionable words used to formulate the justification for a conclusion by any forensic analyst are ‘BASED ON MY TRAINING AND EXPERIENCE…’ Training and experience in the absence of demonstrative evidence mean little to me. A reputable examiner should be able to show the decision makers — the prosecutor, the defense attorney, the judge and the jury — the basis for a conclusion which is understandable and can be justified by data or images. If the examiner resorts [only] to the ‘trust me, I know what I am doing logic,’ a red flag should immediately go up: DON’T TRUST HIM!”4
Deduction and abduction differ from induction in that the reasoning is from general principle to particular instance rather than the opposite. Note Peirce’s illustration for deduction:
Rule: All the beans from this bag are white.
Case: These beans are from this bag.
Therefore Result: These beans are white.
Imagine watching someone reach into a bag filled only with white beans. If he were to enclose several beans in his hand and to ask you, “What color are the beans in my closed hand?” you would answer — without hesitation — “white.”
As a logically valid (properly constructed) deductive inference, the conclusion that “these beans are white” is predictable as long as the premises (“rule”, “case”) are sound (truthful). With a deductive inference, there is one and only one possible outcome — that the “beans are white” — and not multiple outcomes. Since there is only one outcome in a deductive inference and not multiple outcomes, this would allow a person the possibility that he or she can be reasonably certain of the outcome. Without seeing the beans in the hand yet knowing that the beans are from the bag and that all beans in the bag are white, he or she can be certain that the beans are white.
Now consider another illustration:
Rule: All the beans from this bag are white.
Case: The witness says, “I pulled beans from this bag.”
Therefore Result: Those beans are black.
Something is wrong here! If all the beans from the bag are indeed white, then the witness did not pull the beans from that bag! By the use of a variant of deductive inference — known in classical logic as modus tollens — the witness account can be shown to be not truthful.
Consequently, a scientist in the courtroom can state with reasonable certainty if the physical evidence in a case is consistent or not consistent with witness accounts or with hypotheticals offered by attorneys. This is valid deductive inference.
Other names for deductive inference include “forward reasoning” and “cause-to-effect reasoning”5.
Now consider abduction:
Rule: All the beans from this bag are white.
Result: These beans are white.
Therefore Case: These beans are from this bag.
Abductive inference (hypothesis, retroduction, effect-to-cause reasoning, “backward reasoning”5) involves surmising past events from physical evidence, surmising the cause of an effect from the effect. If that same “someone” from the previous illustration, opened his hand and showed you white beans, would you be able to surmise that he pulled those beans out of the bag? He certainly could have, but he also could have obtained beans from numerous other places besides the bag, i.e. from the cupboard, from his neighbor’s cupboard, from the grocery store, from the farm, etc. Because of the numerous possibilities, abductive inference is not reliable and does not allow one to be reasonably certain. To put it plainly, abductive inference is guesswork or speculation.
Abductive Inference Masquerades as Science in the Courtroom
Scientists are accustomed to abductive inference. If a scientist observes something in his laboratory that arouses his curiosity, he may hypothesize (abductively infer) to explain the cause of that observation. The scientist, however, does not stop with the hypothesis. He or she then makes a prediction of what would happen if the hypothesis were true, then designs an experiment to test the hypothesis. Ideally, testing involves attempting to falsify the hypothesis (demonstrate that it is not true). In science, most hypotheses are falsified and replaced by better hypotheses in trial-and-error fashion. This allows self-correction and explains the progress that modern science has enjoyed.
As accustomed to forming hypotheses as scientists are, most do not realize that such hypothesis formation does not work for past events. The past no longer exists in material form and is not subject to manipulation and measurement for the purpose of scientific inquiry and testing6. An experiment can be designed to see if something is possible, but such an experiment will never uncover the unique and complex succession of past events. Consequently, hypotheses offered by scientists about events in the past are not eliminated; rather, they are asserted strongly and repeatedly in courtrooms — even to a reasonable degree of medical or scientific certainty!
This occurs frequently and constantly in every courtroom where scientific testimony is offered. Although the courts are the “gatekeepers” of what should be appropriate scientific testimony, they have been unable — until now — to spot the illogical and the unscientific elements in the testimonies of learned men and women. This has allowed scientists to use abductive inference in courtroom testimony freely without repercussion, all the while expressing that they are certain of their abductively inferred opinions!
Consider this illustration. A woman is found dead in her home. The forensic pathologist, while performing an autopsy, discovers hemorrhages in the muscles of the neck, pinpoint hemorrhages in the face and eyes, and a dilated anus. At a later date, the pathologist testifies in court with reasonable certainty that the woman was strangled and sodomized.
Rule: Women who are strangled and sodomized can have neck muscle hemorrhages, petechial (pinpoint) hemorrhages in the face and eyes, and injured anuses.
Result: The dead woman has neck muscle hemorrhages, petechial hemorrhages in the face and eyes, and a dilated anus.
Therefore Case: The dead woman was strangled and sodomized.
What the pathologist and the court fail to realize is that a woman found dead in the face-down or prone position for several hours can develop changes that look like hemorrhages in the neck muscles, petechiae in the face and eyes, and a dilated anus78. What is even more important to realize is that the testimony is abductive inference expressed to a reasonable degree of medical certainty. This testimony based on speculation dressed up as science should never have been allowed as evidence in the first place.
But it gets worse!
Abductive Inference and “Child Abuse.”
Over a period of about 50 years, a “science” of “child abuse” has developed. The testimonies of “child abuse pediatricians” and other physicians have led to the accusation, conviction and incarceration of numerous people, all on the basis of abductive inference.
These doctors believe that they can discover certain lesions (physical abnormalities due to disease or injury) in a child and infer from them the commission of “child abuse.” Injuries include — but are not limited to — retinal hemorrhages, subdural hemorrhage, brain swelling, posterior rib fractures, multiple fractures of varying age, long-bone spiral fractures, metaphyseal fractures (“classic methaphyseal lesion”), etc.
Rule: Children who are intentionally and abusively assaulted show lesion X (See above list and substitute for X).
Result: A child has lesion X.
Therefore Case: The child was intentionally and abusively assaulted.
This abductive inference is made constantly, every day, throughout the world, in hospitals and particularly in children’s hospitals. Courtroom testimony in the form above is offered constantly, every day, throughout the world. Such testimony should never be allowed as evidence, and hopefully it will cease to be allowed once you decide to stop the madness.
There may be numerous explanations for lesion X. Learned men and women discount such explanations offered by the accused and other witnesses in favor of their own abductive inferences. Law enforcement officers readily accept such inferences and on the basis of such use psychological techniques to evoke confessions out of people who understand very little of what is going on. These confessions are presented before juries and accepted in peer-reviewed journal articles as sound evidence. The accumulations of abductive inference are compiled into studies published in peer-reviewed journals and offered as proof of the “experience” of multiple reputable doctors. None of the abductive inferences are falsified or even capable of falsification. Instead, they form the conventional thinking of a wide variety of medical practitioners and are taught to each new generation of practitioners.
From generation to generation, there is no self-correction. There is no attempt to falsify. Instead, there are loud denunciations of anyone critical of the conventional thinking. The reputations of those who forcefully articulate abductive inference with confidence and conviction are enhanced, and they occupy editorial positions in peer-reviewed journals. Large sums of money, including government funds, are spent in public health efforts and on strategies to incarcerate people based on abductive inference. As a result, nothing new is learned. The complexity of the human organism and its myriad reactions to disease and injury are simply reduced to a few well-worn comparisons to accidental injury in court, involving “2 to 3 story falls” and “30 mile-per-hour car crashes.”
What Would Happen If … ?
Attorneys and judges, I still hope I have your attention. I invite you for the next few moments to dream a dream with me.
What would happen if the judges in every courtroom throughout the land stated, “All expert opinions offered to a reasonable degree of medical certainty or probability will be from deductive inference only; abductive inference will not be allowed”? What would happen if attorneys were allowed to object by stating, “Your honor, the witness is speculating”? What would happen if the courts applied the Inferential Test for Expert Testimony published on this website9 to all experts?
Everything would change.
Forensic science would take a more limited but appropriate role in the courtroom. Doctors would no longer be able to state diagnoses with any hint of “abuse” in them; they would be restricted to stating if witness accounts are consistent or not consistent with the physical evidence they viewed. Consequently, medical doctors and others would have to “unlearn” what they have been trained to do and learn how to reason deductively. Terms like shaken baby syndrome, battered child syndrome, abusive head injury and non-accidental injury would disappear into the dustbin of history. That is because eyewitnesses do not use these terms; only backwardly reasoning, abductively inferring doctors do.
Research into childhood trauma and the mimics of childhood trauma would employ deductive inference. There would be a renaissance in our understanding because we would have dispensed with hackneyed hypotheses not based on witness accounts. Law enforcement officers would concentrate on obtaining accurate witness accounts for comparison with physical evidence rather than trying to evoke confessions from witnesses.
The Department of Justice would withhold funding for abductively inferred but not deductively supported research. Science would change. It would have to. And for the first time, science in the courtroom would become self-correcting. I am not saying there would be no human error because to err is human. I believe, however, that the errors would markedly decrease and would be caught for the most part by the time trial deliberation begins. Truthful evidence properly inferred has the marvelous ability to correct erroneous notions, in spite of our mistakes.
Well, what are you waiting for?
If you are waiting for scientists to change, you wait in vain. There are strong incentives for scientists to maintain the status quo and strong disincentives to change. For many doctors, the consequences and professional humiliation are too horrible to contemplate.
Certainly, if it were up to me, everything would change immediately, but I don’t possess the power to enact change. I can only point out the faulty logic and make suggestions, but I have no control over what some expert says in the courtroom. Only you do.
It is your choice to make. You can stop the madness now, if you so decide.
1 Saks MJ, et al. Model Prevention and Remedy of Erroneous Convictions Act. Arizona Law Journal 2001;33:665.
2 McKean E, ed. New Oxford American Dictionary, 2nd ed, 2005.
3 Peirce CS. Illustrations of the logic of science. Sixth paper — deduction, induction, and hypothesis. The Popular Science Monthly 1878;13:470-82.
4 Bono JP. President’s Message. In: Academy News, American Academy of Forensic Sciences, 2010;40(5):3.
5 Young TW. Is Sherlock Holmes’ “reasoning backwards” a reliable method for discovering truth? Analyses of four medicolegal cases. September 14, 2010.
6 Young TW. Forensic Science and the Scientific Method. February 13, 2008.
7 Pollanen MS, Channa Perera SD, Clutterbuck DJ. Hemorrhagic lividity of the neck. Controlled induction of postmortem hypostatic hemorrhages. Am J Forens Med Pathol 2009 Dec;30(4):322-6.
8 Perper JA. Time of death and changes after death. Part 1: Anatomical considerations. In: Spitz WU, editor. Spitz and Fisher’s Medicolegal Investigation of Death, 4th edition. Springfield, IL: Charles C. Thomas; 2006; p.120.
9 Young TW. An Inferential Test for Expert Testimony. April 5, 2009.