RL Tavernaro wrote:Another way of describing some comparisons that lack sufficiency, yet are consistent, is along the lines of:
"The latent print from item A could neither be identified or excluded as having been made by the right index finger of John Doe."
The specific nature of the comparison is given, without confusing the issue with possibly inappropriate terminology. Even in comparitive analysis disciplines where "consistent with" has historically been used, there has been debate as to its precise meaning, and when or if it should be used.
Essentially what you've done here is define inconclusive similar to SWGFAST:
INCONCLUSIVE
The inability to either individualize or exclude an area of friction ridge detail.
So relative to this, could you not just write inconclusive?
Of course that takes us back to the question of what additional documentation should go along with an inconclusive conclusion.
"Fingerprint Science" traditionally only gives us three choices: Yes, No, and Can't Tell. Other "comparison sciences" allow more options, five or even seven. Five would be: Yes, Probably, Can't Tell, Probably Not, and No. Seven choices offers a greater degree of probability toward either yes or no.
I have seen latent prints that, in my own mind, fit a higher degree of probability than simply, "Can't Tell." My friend RL Tavernaro may remember a case we both looked at a number of years ago. A female examiner in our agency called an ident on a lamp at the head of the rape victim's bed. The rape victim and her entire family were all loops, but one latent on the lamp at the head of her bed was a whorl. Not only that, it was as perfect an overlay with the suspect's fingerprint that you can imagine, to a grand total of three points, as I recall (but who's counting?) I felt that the latent was "probably" that of the suspect, but in line with department policy and latent print dogma, I declined to verify and said simply, "Can't Tell."
I wish we had a less subjective way of saying, "Gosh, it sure looks like him and I think it probably is, but I 'Can't Tell.'" With the work of Christophe Champod, Glenn Langenburg, and others, perhaps someday we will have a more objective way of stating that. I hope so.
I understand Dennis' dilemma. A lot of cops and a lot of juries might misunderstand what a higher degree of probability means, and having a detective use that information to coerce a confession out of a suspect is not an ethical use of the information. An ethical use might be to redouble his investigation efforts in other directions, but not to intimidate the suspect into confessing. And yet, I, too, have worked with investigators like that. But, as scientists, should you withhold information simply because it might be misused?
Pat A. Wertheim
P. O. Box 150492
Arlington, TX 76015
Pat A. Wertheim wrote:But, as scientists, should you withhold information simply because it might be misused?
You leave us with the inference that an ethical scientist should not withhold information from people in positions of power (officers, jurors) who might misuse that information. If you had been, say, a rocket scientist or a nuclear scientist in the late 1930s or early 1940s, what would you have done with your scientific information? In hindsight, now, what would have been the right decision? Do scientists carry a share of the responsibility for "information" that is misused by those in power? I think they do.
Do scientists carry a share of the responsibility for "information" that is misused by those in power? I think they do.
Hmmm....do we then fold up and go home, knowing full well that many of our positive idents will be "misused" and discarded by fickle courts due to technicalities? Do we really believe we are steering the boat, or just providing a compass that may be disregarded at will by the captain?
I think you got it right at the end of your post. I was addressing the concern with providing more comprehensive information in a report that would adhere to SWGFAST and other guidelines. The idea is to convey the information in a manner that is factual and not judgemental, or bias inducing.
Dennis,
I agree, in the sense that giving information that would indicate a percentage type probability could be a very slippery slope, & I think is currently beyond our scope. On the other hand, concerning probative value of a particular latent print belonging to a perpertrator, & the fact that the print comparison is 'inconclusive', could well be important info for the investigator/prosecutor.
Pat,
I do recall the case, although many of the details elude me at present. My recollection is that there were four 'points' (but who's counting?) . I too declined to verify.
Thanks, Bob. Even though we are both a few years older now than we were when we looked at that rape case for Glenda, our memories haven't failed us completely. That's encouraging! But that is a good case to make my point. We both declined to verify the identification, but that is exactly the kind of case in which I would like to have been able to tell the investigator, "I think you're on the right track with this guy." Now, I think it would have been completely wrong for the investigator to go to the suspect and say something like, "We've got a positive identification of your fingerprint. You might as well confess and make it easy on yourself." It would have been even worse for the investigator to go to the prosecutor and say, "The fingerprint guy has an identification, but we can't use it in court because it doesn't have eight points." Do investigators misuse our information like that? Some do, some don't.
On the other hand, let's say the family had only loops, the latent on the lamp was a whorl, and it was obviously an exclusion to the suspect. Then we would, in effect, be telling the investigator not to focus too hard on that suspect, but to broaden the scope of his investigation. Why then, if we cannot exclude the suspect and we see some strong similarities that fall just a tad short of individualization, should we not be able to say, "Look, this is not a positive identification. I cannot testify in court that this latent was made by your suspect and, in fact, it might NOT have been made by him. But my advice is to dig into his alibi a little harder and keep your focus on him for a while before you dismiss him as a suspect."
I agree with Mr. James that we have an ethical obligation not to let our "opinions" be misused, but I believe we should have more choices than simply "Yes," "No," or "I can't tell you anything at all."
Pat A. Wertheim
P. O. Box 150492
Arlington, TX 76015
To respond to some statements.
I don’t think that an officer would drop a suspect of a case based on an inconclusive determination. And regardless of what information that we supply to the officer, they will lie to the suspect. It is all part of interview and interrogation. This should not influence any aspect of our job.
I agree that we have an ethical obligation not to bias the jury. It’s not that I don’t think that there is relevance in wanting to define the degree of inconclusiveness, but how would you determine “higher degree of probability”? Using Pat’s example of a three/four point print that is a whorl- would anyone feel comfortable going to court and stating that it is a higher degree of probability? I admit, I am fairly new in this profession and this scares me a bit. If there is the slightest chance that it is not an individualization I don’t feel comfortable trying to guess what the probability is, but I don’t have a whole heck of a lot of experience to back this up. I have no parameters to determine the probability or degree of consistency. Will statistics solve this?
I'm not sure I see how telling the jury that you could not exclude the suspect as being the donor of the print is overly prejudicial. In my mind, what you've done is reduce the value of your finding to a class characteristic -- it would be no different than saying that the suspect had the right blood type, or the right type and size shoe.
It seems to me if you are looking at a whorl with four points that match the suspect in sequence, shape, etc, but you don't feel comfortable calling it, you should be able to discuss the similarities you are seeing. You have effectively put that person into a "class" of people who have whorls with four points that look like these. Even without using probabilities, a jury should be able to see that the suspect is part of a group that is smaller than all the fingers in the world, or even all the whorls in the world. It is up to them to decide how much weight to give this information.
I agree that if there is clearly excluding information, that should also be used. If you know the latent is a whorl and the suspect has all loops, then that information should also be presented to a jury.
But this divulging of information has to work both ways. If we know something that tends to exclude, even on a print that couldn't be used to make a positive ID, we should tell someone. If the same print that can't be used to make a positive ID tends to indicate that the person COULD have made the mark, we should also tell someone that.
I guess my main point is that we should not hide information because we're not sure what someone will make of it. Our job is to observe and report. As soon as we start making judgement calls based on who will use the information, and how, we are introducing bias, because we're saying this information should or should not be divulged based on who it might or might not benefit. I don't think that is for us to decide.
It seems to me if you are looking at a whorl with four points that match the suspect in sequence, shape, etc, but you don't feel comfortable calling it, you should be able to discuss the similarities you are seeing. You have effectively put that person into a "class" of people who have whorls with four points that look like these. Even without using probabilities, a jury should be able to see that the suspect is part of a group that is smaller than all the fingers in the world, or even all the whorls in the world. It is up to them to decide how much weight to give this information.
But how small is the group that you are placing the suspect in? Sure it is smaller than all the whorls in the world, but if you are the expert in the trial shouldn’t you be able to say how small this group is?
I would like to point out that lying to a suspect during interrogation is perfectly legal in this country.
I have played the numbers game too long in my career to ever give the defense any way to screw me on the stand. "Whaddaya mean, "consistent with"? Is it my client's print or not? Your Honor, it's a simple question; please direct the witness to answer!"
I shudder to think what the state of note-taking is heading to. There is little enough time to actually make comparisons in this business, what with the documentation an ASCLD-lab has to do. I swear, if I spend one hour doing a case from start-to-finish, I will have spent literally around 53 minutes on the documentation, maybe more.
I can just see what it's coming to: "It was a dark and rainy day when I grasped pointers in hand and gazed upon the lift card marked Item 1. After I examined the lift card thoroughly, I determined that there was one latent fingerprint containing sufficient quality and quantity of friction ridge skin detail to make an individualization to a specific subject. I then strolled casually over to the AFIS computer... and thus received a print-out of the tenprint card I requested. It was a live-scan card and the quality was poor...After settling my buttocks gently back into my seat, I leaned forward and placed the printout on the desk in front of me...
I think Heidi makes some great points, and I completely agree that that is the direction we should be heading…at least until the statistics come along.
Will statistics solve this?
Tash
I think statistics will help clear up some of the gray of the ‘consistent with’ area but will disrupt the black and white ‘ID’ area. The ‘ID’ area may become an issue when we get statistics due to increased pressure from our supervisors or investigators, or even from ourselves to call it an ID because the statistical model reports an extremely high probability that it’s an ID. Will we be justified in taking that print we are 99% sure of and bumping it to 100% because the computer is saying there is a 1 in 1,000,000,000,000,000,000 chance it is not a match? Or what about an ID we are 100% sure of and the computer is saying there is a 1 in 1,000,000, chance it is not a match. And while it should always come down to the examiner and not the PC I think putting statistical probabilities to things could open a whole new can of worms in the office and on the stands. I also think it has the potential to increase the number of bad idents due to examiners relying to heavily on it with ‘close calls’.
PG
Perhaps we are overlooking the obvious here and putting undue pressure on our reported findings as the "be all and end all" of the matter. Remember that there are two opposing attorneys operating on every case. Our reports are simply expert findings, made with overt attempts to eliminate bias (ACE-V, independent verifications, etc.) It is the task of the prosecuting attorney to choose what evidence to present to the court. If a report is not able to stand on its own under cross examination, then it should either be withdrawn from consideration, or it should rightfully be defeated and excluded in the courtroom. It is not our job to decide what the court uses or allows. If our findings are included and the wording is unfamiliar to the court or the jury, then it is our duty to explain and clarify our findings when asked to do so. If the court mis-interprets a report or fails to ask for clarification, then they do so at the peril of the accused and the validity of the criminal justice system. If I say a latent print has insufficient detail for positive identification and you use it as an ident, that's your fault, not mine.
Personally, I find it incredibly frustrating that latent print identification, arguably the most reliable and accurate forensic tool available, can be dismissed for reasons unrelated but incidental to the validity of the identification. Courts allow hired "experts" often with obvious bias, eye witnesses and testimony of highly questionable reliability, and a myriad of other less reliable evidence to influence or bias the jury with full intent to do so by both the prosecutor and defense. Yet I've never seen an attack on an eyewitness based on statistical reliability of eyewitness accounts, requests for contemporaneous notes during the sighting, verified visual acuity exams by accredited opticians, and the result of current eye chart readings kept in files.
As we go the extra mile to eliminate bias in our reports, the adversary system of the courts on one side or the other continuously attempts to undermine and distort our best efforts of truth and fact. We cannot expect to submit findings and have them accepted at face value. Can we fail to provide all the information we are capable of finding for fear of mis-understanding by the uninformed? I think this is what's driving the multiple choice approach to reporting of latent print work. It is, it isn't, or it's no good. Pick one, only one, and don't bore us with the details until we get to court, then provide us reams of documentation that tells how you arrived at your choice. No gray areas, no maybes, no "consistent with".
We've all been on the witness stand when there are things we would like to clarify or correct, but we are not given the opportunity. Sometimes this is because the prosecutor has failed to prepare his case and has not asked for a pre-trial conference (I've had one in 15 years!). Often it is because the defense sticks to "yes or no" answers to steer and bias the evidence and does not allow any chance to clarify or correct. "Are you still beating your wife, yes or no?" It's not our ballgame. We can only keep our records clean, honest, and hope for the best. We are a stainless steel link in a paper chain.
. I too declined to verify.