In 1923, a federal appeals court in Frye v. United States handed down a landmark ruling on the use of scientific evidence in the courtroom. The trial court in Frye had refused to allow testimony from an expert saying that the defendant had passed a blood pressure lie detection test—a forerunner of the modern lie detector test. The appeals court upheld the trial court's ruling. It stated that, to prevent experts from misleading juries, forensic witnesses could only present evidence based on scientific techniques "sufficiently established to have gained general acceptance in the particular field in which it belongs." The court found the lie detector did not pass this "general acceptance" test. A vast majority of other courts in the country adopted this "general acceptance" test for scientific evidence. Using this test, courts have excluded most controversial scientific evidence.
Some commentators, however, felt the Frye test went too far. They believed that modern juries would not be fooled by questionable scientific evidence because the opposition could cross-examine, place other experts on the stand, and ask for the judge to instruct the jury to discredit testimony it finds unconvincing. They argued that juries are already allowed to evaluate conflicting testimony from doctors, engineers, and accountants.
In 1974, Congress passed the Federal Rules of Evidence. These rules seemed to allow all evidence, including scientific, as long as it was relevant to the case. But most courts still clung to the Frye test for scientific evidence.
In 1993, the U.S. Supreme court in Daubert v. Merrell Dow Pharmaceuticals, Inc. overruled the 70-year-old Frye test saying it didn't conform to the Federal Rules of Evidence. But the court stated:
That the Frye test was displaced...does not mean, however, that the Rules themselves place no limits on the admissibility of purportedly scientific evidence. Nor is the trial judge disabled from screening such evidence. To the contrary, under the Rules the trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable.
The court in Daubert pointed to a number of factors that might indicate the scientific evidence was reliable. Among them were whether the scientific technique can be tested, its known rate of error, whether standards exist to control its operation, its general acceptance in the field, and the possibility the evidence would overwhelm, confuse, or mislead the jury. The general acceptance test became just one of several factors.
Since Daubert merely interpreted the Federal Rules, the case only applied to federal courts. Most criminal trials take place in state courts, which are governed by state evidence codes. Many state legislatures have left the decision over admissibility of scientific evidence to the courts. Some have adopted codes resembling the Federal Rules. Several legislatures have enacted laws banning specific controversial techniques, such as lie detector tests, from being presented in evidence. In short, many state courts still follow the Frye test, others follow Daubert, and others follow legislative bans on specific evidence.
The lie detector, or polygraph, is based on the idea that, when people lie, their body reacts in ways that they cannot control. The polygraph machine continuously measures changes in blood pressure and in rates of breathing, pulse, and perspiration. A needle records the responses on graph paper. During a test, the examiner asks the subject a series of yes-no questions. One question must be an anxiety-provoking question that the subject must lie about. The examiner compares the response to this question to the other responses and interprets whether the subject is lying or telling the truth.
For years, government agencies have used lie detector tests to screen possible security risks. Law enforcement considers them a useful investigative tool. But Congress banned their use in private industry in 1988.
Critics of polygraph tests say they don't work. Defense attorney William G. Hundley watched as his client, implicated in the Koreagate scandal of the 1970s, took a lie detector test. "The needle never moved," said Hundley. Yet his client later confessed that he had lied throughout the test. Hundley concluded: "I don't think there's any medical or scientific evidence which ever tends to establish that your blood pressure elevates, that you perspire more freely or that your pulse quickens when you tell a lie."
Polygraph examiners admit that the test can be beaten. One has even estimated that 50 percent of the population could, through intensive training, learn how to beat the test. Paul Minor, former chief FBI polygraph expert, said: "Can the machine be defeated? Yes, but not easily. When the poly is beaten, it's usually by training or by fluke."
Some polygraph experts estimate that lie detector tests are accurate in 95 percent of all cases. But this figure is disputed and some estimate the range of accuracy as anywhere from 70 to 90 percent.
Support for polygraph testing seems to be growing in the scientific community. A 1982 study of psychophysiologists, scientists who study interrelationships of mind and body, found that 60 percent believed lie detector testing "was a useful tool when considered with other evidence for assessing truth or deception." A second poll taken in 1992 showed that 80 percent of "those psychophysiologists who considered themselves well-informed of the literature believed that the modern polygraph technique was useful when considered with other evidence."
One problem is a lack of standard training and qualifications for polygraph examiners. Only 30 states have set up licensing boards, and no national licensing board exists.
Following the Frye case in 1923, almost all courts—state and federal—refused to admit polygraph evidence. A few courts allowed polygraph evidence if prosecution and defense agreed in advance of the test. But at least one state legislature enacted a law banning courts from admitting even agreed-upon polygraph evidence. After Daubert, several federal appeals courts began allowing polygraph evidence in certain situations.
Another controversial technique is the voiceprint, or sound spectrography. It is based on the idea that each individual's voice is unique. According to the theory, this is caused by the unique shape of each person's mouth, throat, and voice box and a person's unique way of moving the muscles to speak.
Sound spectrography was developed during World War II to identify enemy radio operators. The modern spectrograph, used to make voiceprints, came out of the Bell Labs in the 1960s.
The spectrograph has a revolving cylinder with paper attached to it. As the cylinder revolves, a needle moves across the paper according to the recorded voice's frequency, time, and intensity. The resulting squiggles form a voiceprint, or spectrogram. An examiner can compare voiceprints to determine whether they come from the same speaker.
Comparing voiceprints is different from comparing fingerprints. Everyone's fingerprints remain the same. If Joe makes five prints of his right thumb, they will all be the same. But if Joe makes five voiceprints of him saying, "Put $100,000 in unmarked bills in a brown paper bag," each voiceprint will be different. But, according to the theory, they will resemble each other more than someone else's voiceprint saying the same words.
In 1972, Oscar Tosi of Michigan State University conducted 34,000 tests of voiceprints using 250 male students and about 30 examiners. The examiners, who had only undergone a brief one-month training, were given 15 minutes to interpret each test. Tosi found that false identifications occurred in only about 6 percent of the tests.
In 1976, however, a committee of the National Academy of Sciences concluded that "technical uncertainties concerning the present practice of voice identification are so great as to require that forensic applications be approached with great care and caution." Ten years later, the FBI published a report, which several scientists criticized as deeply flawed. It announced that an extensive examination of 696 FBI voiceprint cases covering 15 years revealed just one false identification and two false eliminations. The report also stated that voiceprints had "yet to find approval among most scientists...as a positive test in comparing voice samples."
In 1992, a committee of the International Association of Identification set certification requirements for voiceprint examiners and set standards for conducting voiceprint comparisons.
A majority of courts today admit voiceprint evidence. Most of these courts cite the Tosi study as showing the reliability of voiceprints (and ignore the report from the National Academy of Sciences committee). Courts that still apply the Frye test are less likely to admit it.
One of the newest techniques in forensic science is DNA fingerprinting. DNA (deoxyribonucleic acid) is the genetic code that determines a person's physical characteristics. Each human cell holds the complete genetic blueprint for an individual. No two people, except for identical twins, have identical DNA.
Under an electron microscope, the DNA molecule looks like two ladders spiraling around each other. The "rungs" of these ladders are made of pairs of molecules called "bases"—the essential ingredients of DNA. There are about 3 billion base pairs in one person's DNA. Of those 3 billion base pairs, 99.9 percent are identical in every human. The differences occur in the remaining .1 percent. This sounds small but this is still 3 million base pairs.
Someday, scientists will be able to make a complete map of a person's DNA from a bit of blood, saliva, semen, skin, fingernail, or hair. But DNA fingerprinting does not, at this stage, make a genetic map of all 3 million base pairs. Instead, DNA fingerprinting charts a few selected areas of DNA.
The process involves recovering a small amount of body tissue or fluid, separating the DNA chemically, cutting and sorting it electrically and chemically, transferring it onto a nylon sheet, and adding radioactive probes into selected areas of the DNA. This produces the DNA fingerprint, which looks like a bar code.
The DNA fingerprint from a suspect can be compared to DNA found at the crime scene. If they don't match, then the suspect is cleared. In the last few years, DNA fingerprints have played a major role in clearing innocent suspects and freeing wrongly convicted prisoners.
If the suspect's and crime scene's DNA do match, this does not necessarily mean the DNA at the crime scene belongs to the suspect. Since the print charts only a few sections of DNA, conceivably other people could share the same DNA print.
Scientists calculate the probabilities. They do this by consulting databases of DNA patterns. They look to see how often each particular pattern, or band, in the bar code occurs in the general population or in the ethnic group of the suspect. Let's say the bar code consists of just five bands and the scientists find that each of the bands occurs in the population as follows:
Band #1—10 percent (or 1/10)
Band #2—5 percent (or 1/20)
Band #3—20 percent (or 1/5)
Band #4—25 percent (or 1/4)
Band #5—2 percent (or 1/50)
By multiplying the odds of having each band, scientists calculate the odds of having all of them. In this case, the odds of a person sharing the same DNA pattern for all five bands would be 1 in 200,000 (10 X 20 X 5 X 4 X 50 = 200,000).
Some critics believe DNA experts often overestimate the odds of having matching DNA fingerprints. They point out that people of the same ethnic group share more DNA and would be more likely to match DNA than people outside these groups. DNA proponents respond that the FBI now keeps separate DNA databases for whites, Hispanics, Asians, and blacks. But critics contend these groupings are too large. In 1992, a National Academy of Sciences report agreed and recommended a temporary ceiling on odds for DNA matches. In 1996, however, a follow-up report found that the databases, combined with new formulas, gave accurate and powerful odds and lifted the ceiling.
Critics also complain of contamination. Proponents agree that contamination can be a problem if people collecting and storing DNA evidence do not follow proper procedures. But they point out that contaminated DNA will lead to false exclusions—not false matches.
Finally, critics cite poor lab procedures. In one study in 1987–88, of 50 samples sent to the three major private DNA labs, two labs made mistakes on one sample. The 1992 National Academy of Sciences report suggested sending separate samples for testing to ensure quality control, but the report concluded that lab procedures were "fundamentally sound."
The overwhelming majority of court cases have allowed the introduction of DNA evidence. Some state legislatures have specifically written laws permitting it.
For Discussion and Writing
1. What problems do new types of scientific evidence pose for a court?
2. What is the difference between the Frye and Daubert tests?
Which test do you think is better? Why?
3. Between polygraph, voiceprint, and DNA evidence, which do you
think is most reliable? Most unreliable? Why?
ACTIVITY: Should It Be Admitted Into Evidence?
IIn this activity, students role play state appeals courts deciding on a proper standard for admitting scientific evidence and applying that standard to three cases.
Form small groups. Each group should (a) review the section on scientific evidence and decide what standard you believe is proper for admitting scientific evidence into court, (b) discuss and decide whether the evidence should be admitted or not, and prepare to report your decision and reasons to the class. In all cases, assume that proper procedures have been followed with the evidence and that the expert is well-qualified in the field.
In each of the following cases, the trial judge refused to admit the evidence and the party trying to introduce the evidence has appealed.
#1. The prosecution wanted to introduce voiceprint evidence that identified the defendant as making a telephoned bomb threat.
#2. The defense wanted to call a polygraph expert to the stand who would testify that the defendant was telling the truth when he denied committing the murder.
#3. The prosecution tried to call a DNA expert who would testify that the blood found on a broken window belonged to the defendant. The judge refused to admit the evidence citing a state law prohibiting all DNA evidence at trial.