This is a somewhat sloppy wording and is worth examining in case someone tries to borrow this argument in a fingerprint case. "Uniqueness" never means literally unique. It always means that something is so variable and random that it can be used to discriminate among sources to a very high degree of certainty. In this issue, there are substantial differences between firearms and fingerprints. Firearms are manufactured with the specific intent that they be as alike as possible. That is the essence of the interchangeable parts process that revolutionized the industry. And the number of different firearms models is very small compared to individual humans whose friction ridge skin could hardly have been designed as a better discriminator. In the actual meaning of "uniqueness," the fundamental assumption for friction ridge skin is indeed generally accepted. It is the scientific validity of threshold that is questioned.* the fundamental assumptions of “uniqueness” and “reproducibility” of tool marks on ballistics evidence are not generally accepted in the relevant scientific community;
Reproducibility is not such an issue for fingerprint identification. We are pretty well able to deal with variations in how fingerprints are impressed. We do not have such a problem of changing features over time as do firearms examiners. And theirs is probably not so great either, since I suspect most cases involve firearms that are rather quickly recovered. The do, however have the issue of mechanical tolerances.
Bad language. Observations of marks and impressions are always objective, so long as bias or prejudice is not a factor in how the thing is interpreted. The things we observe are actual physical events and are therefore objective. Two persons may observe a line of musical notation. They may disagree as whether or not it is good. That is subjective, since it depends on the nature of the observer. But we are entirely in agreement about the tones and durations represented. That is objective. Any other use of "objective" and "subjective" would absurdly render not only forensic examinations based on observation invalid but also all scientific knowledge gained through human observation of phenomena. Only machine determinations of fact would be admitted.* the subjective methodology for identifying so-called “individual” characteristics is not generally accepted; and
This is the core issue in identification. They are essentially stating that a theory and threshold for practical identification has not been validated to any acceptable scientific standard. Validation involving the sort of theoretical foundation common to all comparative forensics inevitably becomes probabilistic. This is the one that must be dealt with adroitly if one is to provide the court with some basis for accepting identification testimony.* claims of identity should not be admitted because there is no generally accepted statistical expression for the probability of a coincidental match.
The NAS Report is specifically described by many members of the committees as not to be taken as authoritative. It specifically did not pretend to have utilized every study of the disciplines it addresses, and it is largely merely a call for development of authority. It does not present itself as claiming any such "significant weight." It calls for more study that can provide weighty support to courts.I. NRC Reports Are Given Significant Weight By Courts In This Jurisdiction As Well As By Forensic Practitioners
Let's look again at:
B. The subjective methodology for identifying so-called “individual” characteristics is not generally accepted
1. It lacks a “precisely defined process”
2. No standard for minimum points of comparison
3. No consensus on the meaning of “match”
4. Unknown error rate
5. Subjective methodology heightens exposure to contextual bias
"Precisely defined process" is so vague as to be largely useless. The "process" is well-defined. Certainly, an individual examiner's "process" should be well-defined and well-articulated.1. It lacks a “precisely defined process”
Here, I think maybe firearms examiners may have a more effective response. There is only one class of human friction ridge skin. There are many, many classes of firearms. Each firearm is capable of producing different discriminatory marks than one of another type. And different types of marks have different types of variations that can have different powers of discrimination. The nature of firearms identification is such that a "point standard" is dangerously simplistic. However, an examiner can confidently exceed an amount of observed features that would establish practical unique identification.2. No standard for minimum points of comparison
The threshold applied by implication in an individual firearms case may be argued against as subjective. But it is not. It is objective. It may be particular to that examiner and the particular firearm and marks observed, but the examiner can presumably state hypothetical lesser combinations of observed data in that case that would have fallen below the threshold.
No. While it is not particularly good language, "match" means that the source of the marks is identified to a very high degree of confidence. The numbers of manufactured potential sources is so small that the probability of two producing the same marks may rationally be presumed to be very small indeed.3. No consensus on the meaning of “match”
Known error rate is simplistically meaningless, since it cannot really be studied. One cannot determine if or when subjective factors may be working. The possibility of error may only be meaningfully examined for individual cases through exploration of the observations used to reach the conclusion. It is subject to alternative interpretations by other experts. In other words, it's a credibility issue that can be contested as any other testimony.4. Unknown error rate
The methodology of observation of features and accumulation of data is objective. Exposure to contextual bias is controllable, and imagined subjectivity is subject, again, to credibility argument and is not a matter for admissibility.5. Subjective methodology heightens exposure to contextual bias
There is citing of the bullet lead affair. That is frankly mixing issues in one discipline with a challenge to another. It is probably an attempt to convey the notion of a previously accepted method being found wanting. I think I would react to that with puzzlement and express wonder that they cited something related to the trial issue by the coincidental involvement of firearms when bullet lead analysis was about metal composition, not mechanical marks.
As the NAS folks pointed out, there is nothing new in this but the Report itself. What is new is that, from the number of mentions of the Report in court filings, it is prompting more thoughtful challenges.