Name: When DNA Isn’t Enough: The Case That Showed Why Activity-Level Interpretation Is Indispensable
Uploaded: 2026-04-29T16:56:08.355Z
Duration: 1 h 40 min 36 s
Description: When DNA Isn’t Enough: The Case That Showed Why Activity-Level Interpretation Is Indispensable

Transcript for "When DNA Isn’t Enough: The Case That Showed Why Activity-Level Interpretation Is Indispensable": Think about the most important case you've ever worked. The one where the evidence was well supported, the results were clear, where everything lined up, where you handed over your report and thought, this is as solid as it gets. Now ask yourself, would that report survive an appeal? Not the identification, the interpretation. If that question makes you even slightly uncomfortable, good. That means this conversation is for you. I'm Matt Gabriel, director of product applications and services at Thermo Fisher Scientific, and this is HIDS 365. For those of you who've been part of the HIDS community for years, you already know what this program represents, A forum built around questions forensic science is actively engaging, led by the scientists closest to the answers. HIDS three sixty five takes us one step further, a year round opportunity for expert led learning. Each episode takes on a real challenge in forensic science straight from the scientists leading the charge, doing the research, setting the standards, and implementing new use cases to improve human identification. The next four episodes are built around a central theme, overcoming interpretation challenges. Each episode approaches the challenge from a different angle, forming a single conversation. The foundational question when DNA evidence walks into a courtroom powerful, accurate, and compelling, how far do we allow it to speak? Today, we begin at the boundary between identification and activity. Subsequent episodes move into determining number of contributors, interpretation under pressure, and finally, what novel methods and the expanding genomic capability demands from laboratories who seek to answer more questions. In 1995, a 17 year old named Birjeet Tengs was murdered in Norway. For decades, the case went unsolved. A cousin was wrongfully convicted through a coerced confession and decades later, in 2022, fully exonerated for the crime. Modern forensic genetics changed the picture. Advanced YSTR analysis and next generation sequencing identified a rare profile on the victim's clothing. The findings were confirmed across laboratories and jurisdictions. The statistical weight was substantial. In 2023, a new defendant was convicted based primarily on that DNA evidence. But later that year, the Court of Appeals overturned the conviction. The presence of DNA was not in question. The debate centered on how the DNA arrived at the scene. The court distinguished between establishing the identity of the DNA and determining the actions of that individual. This episode highlights the significance of that distinction. Every interpretation challenge in this case considered six areas. First, the forensic methods, how a trace stain with a 60 to one female to male DNA ratio led to an identification spanning seven generations. Second, the hierarchy of propositions, the distinction between proving who versus proving what, and how confusing the two questions can influence a conviction. Third, the unknown contributors, at least three unknown male DNA profiles and hair from five unknown individuals. Fourth, the Bayesian network, every possible route the defendant's DNA could have taken to reach the evidence mapped for the court. Fifth, confirmation bias, how it entered the investigation and impacted the outcome of the case. Sixth, practical reporting guidance, what dynamics are at play when the prosecution asks the court to heavily weigh DNA evidence? One thing before we dive in, your questions are a real part of this session. If something comes up as we go, a question about the science, your own practice, anything you wanna hear directly from the experts, please put it in the chat. Anything we don't get to today goes directly to the speakers. We are here to learn together, and we'd love to hear from you. So let's begin. This is Carmoi, 30 kilometers of coastline on the Western edge of Norway. Three small towns, white houses by the water, a place so safe that no one thought twice about a teenager walking home alone. In May 1995, 17 year old Brigetta Tengs walked home from a Friday night out. She never arrived. She was found the next morning in a brushwood clearing beside a village road 400 meters from her front door. Her skull had been crushed. Her clothing had been disturbed. There were hairs clenched in her hands. The community that believed it was safe would never recover. And then the case began to drift away from the truth. The first significant lead consumed six months of investigation before testing revealed it pointed nowhere. With no physical evidence, no witnesses, and under enormous pressure, investigators arrested Big Gitter's 19 year old cousin. After prolonged interrogation, he confessed to a crime he said he couldn't remember committing. He recanted almost immediately, but but the confession was enough for a conviction. On appeal, he was acquitted, but he was still ordered to pay compensation to Brigitte's parents, cleared of murder, and branded with it at the same time. He wasn't fully exonerated for another twenty seven years. One innocent life taken, another destroyed. And the real question, who killed Brigitte Tengs, still unanswered. Meanwhile, the evidence sat in storage, degrading. It was collected before protective protocols existed. But decades later, science had something to say. A new analysis identified a male profile preserved on a single sample from Birgitte's clothing. Genealogical work narrowed the field of possible contributors to a small number of men. In 2023, a court accepted the science, and a conviction followed. But this case was never going to end there. DNA can answer one question with extraordinary force while leaving a with extraordinary force while leaving another dangerously open. On appeal, the higher court accepted that the defendant's DNA was present, but presence was not enough. The court needed to know how it got there, not whose DNA it was, but what activity put it there. Did it arrive through the crime itself or through some other route entirely? In this case, that question became the difference between a scientific conclusion and a legal one, between a profile and a verdict. In Norway, the Tengs case became a turning point in how courts evaluate forensic evidence and how an entire country thinks about the distance between a DNA match and the truth. Nearly thirty years after Birgitte's death, it remains the most studied criminal case in Norwegian history. And the question it left behind belongs to every forensic scientist in the world. When your evidence moves from who to how, does your interpretation hold? Three scientists, three laboratories, nearly three decades inside a single case. Today, the team behind the published findings shares what evidence taught them and what it means for case work like yours. Hello. I'm Walter Parsons. I'm a forensic geneticist, and I would like to talk to you about a COVID case. Birgit Te Tengs, a Norwegian girl that was murdered in 1995. We got in touch with this case in the very early February when the Norwegian police came to our institute and asking us for help in this case. We discussed the fact that, the crime scene was probably not dealt with in a way that was, appropriate for mitochondrial DNA typing, which is very sensitive, which was new at the time. However, we agreed to analyze hair shafts because hair shafts, we can control contamination by cleaning the surfaces of the hair. So we typed the hairs that were in the hands and around the hands of the victim, and we found different mitochondrial DNA profiles. The later defendant provided the reference sample, but his mitochondrial DNA didn't match one of the hairs. There were testings on the waistband and the pantyhose, which didn't yield any interpretable results. So the case went cold again and in 2016 was picked up by the criminal police in Norway because they were recommending new investigations, focusing particularly on the hair samples that gave mitochondrial DNA profiles not identical to the victim. All those mitochondrial DNA analysis did not match any of the reference samples that we received from that particular case. So it's yet unknown to whom these hair belong. There were also other stored items under investigations, and I would like to focus on one particular item, which was a sample referred to as sample a 12 f, which was located near the waistband region of the penthouse. There was a slightly positive tetra base test, which is just suggesting the presence of blood. And the DNA typing of that sample, the autosomal SDRs, resulted in a DNA profile that matched the victim. Y SDR testing of the same sample, however, yielded a y profile that was identical to the y profile of the defendant. This was done by conventional capillary, but also by massively parallel sequencing analysis. A couple of years later, the Netherland Forensic Institute performed rapidly mutating YSDR analysis on samples from closely related men to the defendant. And those male family, members could be excluded based on the Y SDRs, the rapidly mutating Y SDRs based on mutations. The additional statistical analysis were performed by, Michael Mayer Anderson from Denmark. He was using the, y profile in a 12 f. He was using population genetic principles and mutation rate to demonstrate that there is a very limited number of males in the population that are expected to share that y profile. This sparked genealogical investigations by the criminal police, and they were tracing down the male lineage of this defendant to the early eighteenth century and eventually led to the acceptance by the court that the male DNA in a 12 f originates from the defendant. The first trial in Horgesund resulted in a conviction. The case went on appeal, and the NFI examined further examples from the pantyhose, including the areas of a possible bloody hand contact. This produced evidence for multiple male contributors, particularly in this area, including the YSDR profile that matched the defendant. However, the appeal court overturned the conviction while they were still accepting that the defendant's DNA was present in sample eight twelve f. So the issue was not the source attribution, but what the DNA meant at activity level. When we reopened the case, I mean, more than fifteen years have passed. Technology has improved significantly. We went from capillary electrophoresis, mitochondrial DNA sequencing, to MPS, massively parallel sequencing, with much greater opportunities and possibilities, particularly for sequencing mitochondrial DNA. So when we received the hair samples that we were typing with mitochondrial DNA, we needed to exercise extreme caution that we are not typing the mitochondrial DNA on the hair shaft, which could be due to contamination, but make sure that we type the mitochondrial DNA in the hair shaft. And this is why we have rigorous cleaning protocols in place that make sure that the surface of the hair is purified from remnant cells or from remnant DNA that could come from other sources. And only then after this purification step, we would digest the hair completely to get to the mitochondrial DNA that is in the in in the inner part of the hair shaft. I'm Peter Gill, and I'm professor of forensic genetics genetics at Oslo University Hospital. I was approached directly from Kripos, to give a view on the case. This is when they were going through their cold case review, and, I was advising them. The main role in this case was as a court appointed expert along with, my colleague, Jonathan Whittaker. And our role is primarily to review all of the results and agree them effectively or peer review them before they went to court. Our role was also to advise on the activity level, and, a lot of effort was put into that. So during the course of this, we had frequent, discussions with both the prosecution and the defense as well because as a court appointed expert, but you're working for the court, you're not specifically appointed by the prosecution or the defense. You're there purely to to advise, the court. So, we spent a lot of effort, thinking about the activity. How much did the conviction of, an innocent man, although we didn't know that at the time, how much, did that cast a shadow over the the case? I would not say that it completely froze the the search, for other individuals, but it clearly cast a long shadow over it. If someone remains technically guilty, even after acquittal in acquittal in the criminal sense, which is what happened in this case, that inevitably affects the investigative momentum and the mindset of the investigators. We discussed this, in our our paper, and, we conclude that this doubtless inhibited further investigations. Because I think it's just human nature, if you think you've got your suspect and you think he's the correct individual, then you're not really gonna put too much effort into looking at any alternatives. So the problem is that once your case narrative hardens by the this conviction of the cousin, then these alternative lines are less actively pursued. Well, my understanding is that, on appeal, he was actually found to be, innocent, and I I don't think he was actually incarcerated after that finding. But he remained technically, guilty for, civil, purposes. So he wasn't actually exonerated until, 2022. What has been lost in the decades that you could never recover? But before you even start, you know that some things may be irretrievable. You have uncontaminated touch DNA opportunities, the exact trace locations, where did where did the sample actually, come from. You're lacking a reliable reconstruction of who touched what and when. Often, the the original packaging may be missing, etcetera. So even if the later laboratory work is excellent, you are losing, context, of the of the case. And that means that some of these evidential questions, may become harder to to answer. Of course, contamination is, is a a big issue, and this was explored fully in the in the court. But, of course, when you don't have, the controls that we have today, contamination is something that you have to seriously, consider. And, indeed, that was explored, in the court in in great detail. The sample a 12 f, was brought to us as a DNA extract. So the DNA was extracted, at the institute in Oslo. And, we were specifically asked to type white chromosomal DNA in addition to the autosomal STR profile. The autosomal STR profile didn't give any clue that there would be male DNA present. So it was a pure female profile that we found and that matched the victim. So it's only the Y chromosomal analysis that we're able to discover this tiny contribution of Y chromosomal DNA in that sample. Now, in some cases, you would expect that the profile is not interpretable at such an, ratio of, something like 60 to one. But in that case, it was a pretty clean, neat single source profile, as far as we can tell, from micromosomal DNA. And we were further asked if we can use mitochondrial DNA on that particular sample to exclude the suspect, the defendant. So what we did is, prior to using the last remaining volumes of that sample for MPS analysis with mitochondrial DNA, we were conducting an experiment where we produced one to 101 to 200 mixtures of two known sources of a mitochondrial DNA. But we were able to detect the partial profile of the minor component with MPS technology in this one to 101 to 200 samples. So that gave us confidence that MPS would be a feasible technology to look at sample A12F with respect to mitochondrial DNA because we have to expect mitochondrial DNA from the victim and from the male contributor to be present in in the sample. And the result that we got was exactly mimicking or, you know, confirming our initial experiments in a way that we we found a dominant mitochondrial DNA profile that matched the victim, which was expected because that is also reflected by the autosomal STR result. And in addition to that, we saw a very minor component, which resulted in a partial profile, matching the, mitochondrial DNA of the suspect. So in conclusion, a mitochondrial DNA test, which is quite untypical for this kind of samples because we expect a high, contribution from the victim could not give a result that would still include the defendant as a possible donor of that sample. The conclusion, was that the mutation at DYS five seventy six, would be found in fewer than eighteen men in the world. And and the question is, well, is is this decisive? Can we say that this is a decisive answer? Now when we give evidence in court, we have to be very careful. Our first duty is to avoid the prosecutor's fallacy. And what that means is that we have to avoid speaking to the probability of the hypothesis, given the evidence. And if we say something like the DNA came from the defendant, then this is actually an example of the the prosecutor's, fallacy. So we have to avoid saying things like this, so we don't use words like, decisive. But what we can say is that the evidence supports the proposition that the DNA came from the defendant, and indeed we can use a verbal scale. So we could qualify it by saying that the evidence strongly supports the proposition that the DNA came from the defendant. And it's very important to split the the scientist's role from the role of the court. And that means only the court can actually come to the conclusion that the DNA came from the defendant. And the scientist's role is to give the information for the court to make that decision. However, once the the court has made that decision, then the scientists can, adopt it, and then we move on to the activity level, assessment. And that's another reason why we have to be very careful with using words like decisive, because the court may have made a decision at the, identity level, the sub source level, but that's not the same as the decision which must be made at at the, activity level. So, effectively, there are there are two different aspects of the evidence that we have to consider, the identity level, which we call the sub source level, and the activity level, which is the how, why, and when the DNA became evidential. With respect to the technologies we used, we are accredited under capillary electrophoresis systems at the time when we handled the case. And for, massively parallel sequencing, we are accredited for mitochondrial. We were we were accredited for mitochondrial DNA only. So in that case, and and we made that clear in our report, that the Y chromosomal analysis that we performed on MPS was done outside of our accreditation body. Now what we found is that, MPS was very strong in showing the mixture, in interpreting mixtures. And that comes with the nature of MPS because we are looking at individual sequencing that have linked alleles, and we can therefore link the alleles together and attribute it to one source, which is something that you cannot do with capillary electrophoresis. However, the, the system that we used was combining autosomal SDRs with white chromosomal SDRs. And we think that this has limited the ability to detect more Y chromosomal markers in that uneven, unbalanced mixture than we were able to because the the the autosomal part was, of course, consuming the chemistry and the and, you know, the the components of the of the assay for the dominant female profile of the victim. But, you know, the the the strength of massively parallel sequencing certainly lies in its ability to link alleles of the same donor to each other. It's something that capillary electrophoresis cannot do, and this was also a major advantage in that case. Hi. My name is Arnott Kahl. I'm a forensic scientist at the Netherlands Forensic Institute. The NFI was contacted in 2023 by the nor Norwegian police, and their question was relating to the a 12 f f YSTR profile that was generated by, by Walter and his colleagues in Inchbrook. And one of the questions was if it was possible to, get more information out of the sample, more more YSTR load side, and also if we could, discriminate the YSTR profiles of the the closely related males, most most particular, the two brothers of the defendant. The the amount of y male DNA in the sample was was very low, so we had to be very careful, not to consume, that sample. So the first investigation that we did was to, use our different panels of y chromosomal markers on reference samples from the defendant and both of his brothers to see which, if any, of those markers could discriminate between them. And then the idea for a follow-up was if if we were able to identify a locus that was different between the defendant and his brothers that we then could see with with a very dedicated analysis, a monoplex and NPS approach, so a very sensitive method to see what the typing was in the in the a 12 f sample. But it all, went a bit a bit different. We did our analysis on these three reference samples, the defendant and his two brothers. We used different panels of YSTR load side. We used, the forensic panel, which is an NPS, analysis. And we used the rapidly mutating YSTR markers that were originally developed by the Erasmus Medical Center, the the group of Manfred Kaiser and Albanewalff and, and then their colleagues. And, actually, we we work with them to, validate those rapidly mutating markers for forensic case work, and we were, accredited for that. So we were able to, well, do this investigation in the case under, accreditation. And it appeared that's, of the 54 loads I we tested, there were two loads I that were giving difference, results for the defendant and his two brothers. So we could discriminate all all three of them. And quite quite surprising maybe is that one of these loads I was the DYS five seventy six, which is actually present in most of the, commercial kits now available like the the the PowerPlex y 23 or the GlobalFiler. And this made it quite easy for the Norwegian police to test test a large number of the male relatives for this particular locus and to see if they had the the the allele that was present in the in the the trace in the a 12 f f a 12 f sample or or not. But the final conclusion of that investigation was that it is actually the defendant who had the mutation compared to all his male relatives. So he was the only one with a a different allele on this this locus. And when we are discussing, likelihood ratios, it's very important to discriminate likelihood ratios on a sub source level versus, an activity level. And if we have an LR calculated at sub source level, we cannot transfer that to the conclusion at the activity level. Because in this way, the, activity level evidence is, highly exaggerated, and that can, cause miscarriages of justice. So the the likelihood ratio on sub source level is not the same as the likelihood ratio on activity level. To explain the hierarchy of propositions, I would describe this simply, as a rule that you must answer the question which is actually being asked at the right level. So the first question which which you'll be asked is, whose DNA is it? And we call this the the sub source level. And that is obviously a separate question. So how did it get there And what activity does it support? And this is the activity level. They are completely different questions, and we are able to carry out likelihood ratio calculations calculations for each one. So the calculation we carry out with the sub source question is obviously going to be different to the calculation that we carry out for the activity level, calculation. And it's very, very important not to confuse the two, and it's very important not to carry over the sub source likelihood ratio to the activity level, likelihood ratio, because this can cause miscarriages of of justice. And certainly, this case actually depends on the distinction between the sub source and the activity, levels. And we have to be extremely cautious to make the difference clear to the court. After the first, conviction, the NFI was asked to conduct further investigations into the, traces on the penthouse. On the thigh region, there was a a potential, a handprint, and it was thought that maybe this location on the on the on that penthouse could contain, DNA from, well, a perpetrator, the victim herself, and, possibly also other individuals. So we did a thorough investigation on the on the on the patterns, on the on the on the, Pentehaus. And we, took 65, different I extracted the DNA from that and trying to find, DNA profiles and getting, an idea of who has contributed to to these to these traces. And what is very important in this kind of investigation is that if you get DNA profiles that you, try to find out if these DNA profiles, originate from known individuals, like first responders to the crime scene, crime scene investigators, laboratory personnel that has been working on the case, maybe family members or or friends from Birgitte, the the victim. And, we did all those, comparisons with as many different profiles as we could could get, but none of them, matched the the profiles that we got from the, from the Penteos except the DNA profile from the from the victim herself. There was, a further investigation into the the YSTR, profiles that we, obtained. What can be explanations for these these profiles? They can be, backgrounds, DNA from an unknown source not related to the crime, which is quite common to to find on, on clothing, because the clothing can can become, can come in contact with any, many other other persons. But DNA profiles can also point to, another contributor, not the the suspect, not the defendant who might be the real perpetrator. So as long as we don't know the source of these DNA profiles, we cannot discriminate between a background signal and a possible perpetrator. For a number of samples retrieved from the thigh region of the PentiHost, we, got, YSTR profiles which were incomplete, so we did not get the full profile. And also they were, sometimes mixed profiles, so showing, multiple alleles on on one locus. And from that, we we concluded for those samples that, there there was possibly presence of the male DNA from more than one male, maybe up to three. That that really depends on if you, consider an a peak in your electropherogram to be, a real allele from a real donor. And maybe you can be more careful if you can also consider that as, as an artifact, for instance, an elevated stutter peak. So there's some insecurity, some, unknown, number of additional males in those in those samples. And the problem then is if you have those mixed and incomplete, YSTR profiles is that we do not have a statistical model. There's no probabilistic probabilistic genotyping method that can handle those mixed, YSTR profiles when there's an uncertainty about the number of donors and when there's artifacts like a little drop in and and drop out. So we we have mixed profiles. We have individuals that, that match those mixed profiles, but we cannot calculate, the probability of these findings giving propositions of the prosecution or propositions of the defense. And, in this well, for this reason, there's, for those samples, no scientific evidence to support or to refute the proposition that DNA from a certain individual is present or not. And the lack of such a probabilistic genotyping method is was, well, very apparent in this case and, also stimulates us to think further on this, on this and to develop those those methods in the future. For sample 55, it was different because there we obtained a consensus YSTR profile, that was single source and that also matched the defendant. And this particular consensus profile was not observed in the in the collection of over 100,000 YSTR profiles present at the YHRD. So we could, report, evidential value on a verbal scale. And we reported that the probability of serving observing this consensus YSTR profile is far more probable when the suspect is donor of the male d DNA in sample 55 than when another man, unrelated to the suspect, is the donor. What is a Bayesian network? A Bayesian network is best thought of as a structured map. What it does is it lays out all the plausible routes by which a defendant's DNA could end up on the closing. And these possibilities are either a physical assault or by some kind of social contact, which leads to indirect transfer, or it could be contamination, or there could be an alternative offender, as well as background, DNA from unknown individuals. I think it's a very valuable thing to show a court because, you don't need the Bayesian network necessarily to give you an actual calculation. And the value of it used in this case was primarily illustrative. It forces the court to see that the source attribution alone does not collapse all the roots into into one. And the question is, well, did the judges understand it? Yes, but my feedback was that they they did, understand it. The court clearly understood the differences between sub source and activity, for example. And I think the core the court understood the core message, which was to show that there were multiple transfer pathways and that the activity could not be read directly from the source attribution. Quite often in case work, I I see, examples of statements where the unknown individuals are are literally, dismissed because the the scientist may say that, this is something we expect to see. We expect to see unknown individuals, and they may just ignore them. But but, actually, this is the wrong thing to do because, if we think of the the propositions or the hypotheses, we cannot ignore the defense, hypothesis. The defense is saying that an unknown individual is responsible for the crime. So it follows that if we find DNA from unknown individuals, then this actually supports the defense, proposition that an unknown individual, is responsible. So what that means is that if we do find unknown individuals, we can't just dismiss them. What we have to do is to actually take account of that, in our activity level, calculations. For example, let's suppose we find, an unknown individual there, and let's suppose the the quantity of DNA that we recover is the same amount as that has come from the suspect. Under those circumstances, the, activity level likelihood ratios, must be equivalent. If I obtain, a quantity of DNA for an unknown person, which is the same amount as from, the the suspect, then the the activity level likelihood ratio is based on that quantity of presence of unknown, individuals. Ultimately, how much, notice the court takes of this is going to depend upon the other evidence in the case. Court takes account of all of the evidence in the case. But if we are looking at the DNA in isolation, then two samples, which have equivalent quantities, have effectively the same amount of information in terms of strength of evidence. In addition to that, I think the court made the point that the prosecution was focusing on the, on the sample a 12 f, and did not, according to the court decision, pay enough attention on finding the donor of the hair samples that were in the hands or near the hands of the victim and, you know, cannot be ignored as potential relevant samples, in this crime. Yes. That's right. The the court did actually focus very strongly on these hair samples, and they pointed out that these hair samples had clearly not come from the, suspect, and there was evidence that they had come from unknown, individuals. Right. And the investigators were criticized for not putting sufficient resources into investigating these unknown individuals, because, obviously, they support the defense proposition. Well, I guess ignoring is a strong word here. I I would not say that the prosecution or the police were ignoring other evidence. They were submitting reference samples of all possible, reference donors that they, were aware of through the examinations, in this case, for DNA testing. And these the samples of these individuals were tested with Y chromosomal markers. They were tested with mitochondrial DNA and also autosomal DNA. But those donors, they did not, match any of the hair samples or any of the additional Y profiles that Arnold found in, in on the Penthouse. Yes. There were there were a number of, alternative, individuals who were investigated, and the suspect was one of these. He had did have a a previous conviction, which is why he came into the frame. And it just so happens that his profile, aligned with one of the findings from this case. It's certainly as important to think about confirmation bias, as it is important to think about, the prosecutor's, fallacy. It is unfortunately, I think, quite common with the investigators, and certainly, they they need to be trained in avoiding, confirmation bias. And it's because it's extremely dangerous to, work on a case and, to be led by a guiding principle, where that principle is that you're you're just looking for the information which fits your hypothesis. What should, a scientist do if the result contradicts the dominant hypothesis? Well, the the first issue is that, we don't really think in terms of the dominant hypothesis. There were two alternative hypotheses, and neither of these are are dominant. So you'll have a a hypothesis which is put forward by the prosecution. You'll have a hypothesis that is put forward by the defense. But I think it's wrong to think of either of these, hypotheses as being, dominant. So what the scientists will do is to establish the, the support for either one of these hypotheses. But, obviously, they can't both be true. One one hypothesis may be true and the other one, isn't, and the scientist evaluates the strength of the evidence which supports one of the hypotheses. So we have to be very cautious, I think, with the way in which we which we think about this. We shouldn't think about dominant hypotheses or propositions. It's best to think of them as equivalent. I I totally agree with, what what Peter is saying that you should in include those findings in your case evaluation and include the findings in your report as well. During our investigation, we we try to stay blind for that kind of information. So as as Peter already said, we are not discussing this in terms of a dominant hypothesis or what one hypothesis being better than than than the other. So it's just, well, find finding the facts and interpreting them in the in the in the information that you have in the in the case. Also, laboratory personnel, they don't know, the background of the case. They are just no, working with a DNA extract and generating a profile, not knowing anything of the of the context of the case. That's only one stage further when the DNA profiles are compared and are, well, evaluated in with all the information in the case. I I think it illustrates a mindset of the investigators who are trying to solve this. It's a very high profile crime, and they're they're trying to solve it by carrying out, more tests effectively. But in many respects, they're they're trying to answer the wrong question because the question was always always at the activity level, not at the sub source level. So but there the investigation was primarily directed at the sub source level when they should have been thinking more about the activity. The question relates to the, two conviction the conviction, in the first trial and the exoneration in the second trial. And it's been pointed out that the DNA evidence did not effectively change, so so what did? I think the answer to this, question revolves predominantly around the interpretive context. So this context, changed or was thought of differently between the two different courts. So the appeal court, certainly had the additional evidence from the NFI work. And they placed greater emphasis on the unknown contributors, and they put a much stronger focus on confirmation bias. Even though the evidence in the first court certainly spoke, about unknown contributors, and the first court was certainly warned about confirmation bias. So the evidence in that respect was was was more or less the same, but the courts actually thought about the evidence in slightly different ways. So the appeal court placed a clearer distinction between the source attribution and the activity level meaning. Attribution and the activity level meaning. So, obviously, the the evidence from a 12 f, was was crucial in both, courts, but the evidential landscape around it, changed between the two. The appeal court focused on the defense not needing to provide an alternative, explanation, And they they said that it's the responsibility of the prosecution to prove the case beyond reasonable doubt. And it was really important in the court's mind to not place the onus of, of of proof on the on the defense. And this is known as reversing the burden of proof, and, that effectively, is is not allowed. The second court was much, much more focused on the issues of confirmation bias, issues around the prosecutor's fallacy, for example. And, they were very focused on the burden of proof. So does that mean that innocence, is is unlikely? Perhaps the the the best way to answer this kind of question scientific test, as in this case, you do it for a reason, and you will have an a priori expectation if a proposition is true. So if the defendant assaulted the the victim, I would expect to find, his DNA. If somebody else, assaulted the victim, then I would expect to find someone else's DNA, for example. So what happens if you don't find the DNA? Well, that actually supports the, the defense of proposition. And, there are actually mathematical calculations that you carry out which shows that this, is true. So the answer to the question is, if you find something which you don't expect to find given some proposition, then this will support the alternative. So if you don't find DNA, for example, then when you expect to find it, then that can't be simply dismissed as as neutral evidence. The the question is whether a properly structured activity level likelihood ratio, would have helped. But definitely, this would have helped. The, the calculation was not carried out, however, because because there was no agreement on the experiments, to to carry out. So this was certainly something which I discussed with the prosecutors and the, defense. Whereas the experiments to carry out are quite obvious from the, the the prosecution point of view, it becomes much more difficult from the defense, perspective, because, as I've already mentioned, the defense are under no obligation to supply, an alternative. And so you can put forward various experimental designs, which are designed to maximize, the the transfer of DNA given the defense, proposition. But, ideally, you would like to get, an agreement between the defense and the prosecution before you carry out these tests. So, to cut a long story short, it wasn't possible to carry out these tests. And this is quite an unsatisfactory position to be in because what it means is when you give evidence, you are not in a position to give an activity level likelihood ratio because you don't have the data. One of the interesting features of this court was, they actually said the experts had a duty to speculate in favor of the, accused. But I have to say, actually, having spoken to, lawyers, since, this trial, They say that this is an unusual thing for a court, to say, and it seems to be more more like a a a one off as far as I can gather. The overriding responsibility is that of the prosecution to prove the case beyond reasonable doubt, and this seems to be pretty universal. But if we think a bit more about speculation, of course, scientists are not allowed to speculate. We do not speculate. A court may speculate in terms of propositions, but scientists cannot speculate. They can only interpret propositions which have been handed to them. Yes. There, appears to be a discrepancy in the in the conclusions that the courts, has on on sample 55. To remind, sample 55 was a a trace sample on the thigh of the pantyhose, and it contained a YSTR profile, which was not as complete as the the YSTR profile from a 12 f, the the stain that was close to the the waistband on the on the pantyhose. And the the court concluded that, 55 could be transferred to a 12 or the other way around during folding and and handling of the the pantyhose, during packaging, or during any of the investigations. And where where the court was clear that the DNA in a 12 f, was originating from the defendant, they did not reach that conclusion for sample 55. And I'm not sure how how they reached those different conclusions. But what what I can say is that the YSTR profile of sample 55 was not as complete as the profile of a 12 f. So a 12 f had 26, confirmed load size, typed from the YSTR markers, where for 55, there were only 12 load size confirmed. So there's less information in, in the 55 sample. And also, well, first, I need to say that despite this lower number of of loads high in 55, there are still a lot of weight, to this profile if you compare it to, YSTR profiles in the YHRD. So this this 12, locks profile was not found in the YHRD in over 100,000 profiles. And also it was, quite interesting to to read that the court considered that, profile in '55 could be a a composite profile. So containing alleles of different contributors that then were put together in a composite profile that by coincidence aligned with the profile of the suspect. And this was, an interpretation that was purely from the court and was not proposed by any of the experts during the trial and also not supported by any statistical evaluation. I'm going to give, a step by step instructions for reporting when DNA is primary evidence and there is no specific defense proposition. Effectively, the defendant is saying that he had nothing to do with the crime. He wasn't there. He doesn't know why his DNA was there, and he has no explanation. But there we can still, look at this in a in a structured way, however. First of all, we have to define clearly what proposition level you can address, and this comes under the hierarchy of propositions. Can you address the source? Can you address the activity? Or or both? Of course, when we give evidence, we have to talk extensively, about what the evidence does not tell us as well as what it does tell us. So the second step, we must state explicitly what the DNA does establish and what it does not. And at the same time, we have to warn about the prosecutor's fallacy and also about confirmation bias. Thirdly, we have to identify if there are unknown contributors present and discuss their impact on the evidence. Fourthly, we have to warn the court about carrying over a sub source likelihood ratio into an activity, conclude conclusion. At the same time, warning about relying solely on DNA evidence alone, because this is where confirmation bias also kicks in. Fifthly, if the defense position is nonspecific, we have to explain that directly in relation to the limitations of any assumptions that are being made in this case. Sixthly, I would consider proxy experiments, in these kinds of of cases. It is possible to think of experimental designs, which are designed very conservatively to maximize support for innocent indirect secondary transfer. The court will need to decide whether the experiment itself is relevant to the case in hand. The seventh step, you must disclose the assumptions, caveats, and any missing data. And finally, make it very clear that the court must decide the relevance and overall weight of the evidence. The scientist's job is evaluated clearly, given the propositions. We are not adjudicators. That is the job of the court. Our paper identified a specific forensic difficulty. In, a type two case, the defense may simply say he was not involved. This differs from the type one case where we do have information from the defense, and I give an example in the paper about a case I was involved with where a watch was recovered from a crime scene. And there was no doubt that this watch came from the defendant because there there was actually a photograph of the watch, which the the police had, and he said that he'd lost it. Now, under that circumstance, it's fairly easy to find, or to think of a defense proposition. What you can do is you you simulate, someone losing a watch. Maybe it was it was always stolen, and someone else wears it for a week. And the question is, will you still find their defendant's DNA if it's been lost for a week and someone else has been wearing it? So these experiments are are fairly straightforward. The difficulty comes with a type two case where there, where there is no, defense alternative. Now the question there becomes, well, what sort of experiments can you, design to help, with this kind of, case? And, certainly, this is something which we discussed at the, at the the TENG's, case. We looked at various, scenarios. One of the, scenarios was that, the the victim visited a a shop, went to a toilet in this shop shortly after the defendant had been. So she the the proposition was that she held the door handle. She went to the toilet, pulled her pulled her pantyhose down, and that's how the defendant's DNA got on the pantyhose. Now these kinds of experiments, we can, of course, simulate, but, ideally, you need a court agreement to to be able to do this. And as I pointed out that, at the time, we didn't have the agreement of the court to do this. But I I've since thought that, really, if if we had a series of, tests which we've already carried out, then, of course, we will be perfectly at liberty to disclose this information, to the court and also give them, activity level likelihood ratios, even though we didn't know what the precise activity was. Of course, it's the final decision of whether the this is acceptable to the court is the court's decision. It's not the decision of the scientist to decide whether these kinds of experiments are acceptable, to the court. The role of the scientist is is to provide the information along with the caveats, and the court ultimately decides its relevance. But but, of course, the it is up to the court. If the court wants to explore this further, of course, then the court actually could tell the scientists what sort of experiments which, it would like you to carry out, and this would be the ideal position to be in. If there is a case where the defense, has no specific alternative proposition, how do you move forward? Can you design proxy experiments? And do you need, parties to agree, on those experiments? Well, in The Netherlands, we have an inquisitorial system, and that gives us the opportunity to arrange meetings with all the parties involved. So in the situation where the defense has a a specific alternative proposition, we will set up a meeting with the prosecution, the defense, and an investigative judge. And we discuss, the options of the experiments, in in view of the the the case at hand. And if we make those experiments, all of those parties need to agree on these experiments and that they are relevant for the case and that they they can be used in in, further proceedings. And the other way, when the defense does not have a specific alternative proposition, we will still try to set up such a meeting and discuss experiments. And those experiments, of course, cannot be as detailed and specific, as as in the situation where there is an alternative an alternative proposition. But as Peter already explained, some some experiments are quite straightforward and can be conducted in more general terms such as a secondary transfer experiments, for instance. But still, if the defense does not agree on those experiments, well, there's no point in conducting those experiments anyway. So we then will not move forward to those experiments because you cannot use them. The the question is, how do legal systems shape correct interpretation or misuse? So one is the inquisitorial, and the the second is the adversarial, which, is prevalent in the in UK and North America. The main difference, I think, is that in the inquisitorial system, judges actually play an active part in the investigation, and they may help to build the evident picture. It's quite different in the adversarial system because, judges do not investigate, and they must work within the picture which their parties present to them. I I think of the adversarial system as more like a boxing match between two two lawyers. The key feature, I think, is that in the inquisitorial system, judges can be more engaged earlier, and that may help define the questions which science should answer. Whereas in the adversary system, much depends on what the parties, ask for, what the judge admits, and how well the scientific limitations are the are communicated. I think neither system is is perfect, but, my experience in working in both systems is that I certainly prefer the inquisitorial system because it is more open. You are for example, in the Tang's case, we were allowed to listen to each other's evidence, and the judges even allow us to ask questions of other experts, for example. And this is something which wouldn't happen in the adversary. The question is whether defendants should be prosecuted on DNA evidence alone. Is this a scientific or a moral issue, or is it both? I think it is certainly both. From the scientific, point of view, we've already discussed the hierarchy of propositions. And if a scientist goes to court and just discusses the, the sub source evidence, Maybe he'll talk of statistics of likelihood ratios of the order of billions, for example. The danger is, that this figure is just simply carried over to the to the activity, level. And the problem is, of course, if there is no other information in the case, the court only has the sub source information, and therefore the temptation will be just to use that that number, that huge number to decide the ultimate issue. And of course this is also a moral, issue, because you're basing convictions on incomplete information, if that if that happens. So we call this Sol Plank evidence, and it does create a serious risk that the court will overread what the science, actually says. And I think the Tengs case shows why this is dangerous. But I guess the, the Tengs case makes this very clear that silence on the activity level part, can mislead the court significantly. And and it's because in this case, we only had DNA evidence, so there was no other evidence. There was no eyewitness testimony that would place the defendant at the crime scene. So it's it's it's based on DNA alone. And the fact that the y chromosomal profile of the defendant was so rare, and in addition to that, the constellation was so lucky from the prosecution perspective that we were able to exclude close paternally related individuals like the father or the brother of the defendant made the Y SDR profile and the DNA evidence very strong. And in addition to that, the genealogical analysis took place, which is rarely happening in such cases, at least to my knowledge. And they, you know, they spent the full year on trying to trace back the the the paternal ancestors of of that lineage going back to the early eighteenth century. So all this is strongly suggesting, and which was also later accepted by court, that although it's a y chromosomal profile, which is uni parentally inherited, there was no doubt that this DNA came from the defendant. Now if we would only if we would stop here and not discuss or warn about the fact that we don't know how this DNA came onto the pantyhose and, the area of the waistband, if we don't tell this to the court, they could certainly be misled that the defendant is the actual perpetrator in that case. So recently, Ted Hunt, published a a paper which questioned the admissibility of activity level, in court. And my understanding in The US that this paper has, gained some, traction. However, I I think we need to be extremely, cautious because I think the the paper is really concentrating on the admissibility of, evidence, in court. But just because something may be judged inadmissible is not a reason for a scientist to withhold the information. And my concern with the paper is that this was the kind of message which I felt was being, propagated. It's not admissible in in court, and therefore, you cannot report it. It. But I think that's a conflation of the responsibility of the the expert with the court admissibility issue. Clearly, the responsibility of the the scientist is to give the best science possible. And even if something is not admissible in court, that does not relieve the scientist of the his or her responsibility to set out their full scientific opinion in writing. And certainly, no court can stop a scientist giving their scientific opinion, in writing. Of course, the the scientist cannot control the court proceedings, that is a matter for for the judge. On the contrary, full disclosure is extremely important precisely because it may well assist the defense. So Hunt may be right on the narrower issue of whether certain activity level, testimony will be admitted. But we must not confuse that with any suggestion that the scientist is absolved of the responsibility to identify the relevant scientific issues and make them known to the court? Well, if I were to, articulate, the most relevant findings in a single sentence here, I we were considering whether or not it was clever or it was, justified to conduct experiments on in a COVID case where we already know that the contamination precautions that were taken at the time and that were fine at the time are not meeting current standards. And I would still argue that, yes, we should invest investigate such cases. We should perform the experiments based on the technology, on the latest version of the technology that we have. Because first, we need to get the data, and, we need to see the results before we can come to conclusions. And, of course, we need to factor in the, you know, the the the fact of the situation at the time that, the contamination precautions were different compared to now. So dealing with that case, doing the experiments, I would still repeat, and and and I would still do this. However, what this case is teaching us is, and probably the take home message for many that were involved in this case, do not fall victim of the confirmation trap. For for me, the most important, issue for from this investigation is that we should never give up on a cold case and and keep on trying. The technology keeps improving. So over time, we can use newer technology to investigate cold cases and come up with a relevant result that was not possible a couple of years ago. And also another important thing, we need to give, all our attention to the DNA profiles of unknown individuals. So we need to try to identify them and try to explain them in the in the in the context of the case. Because they could be potential perpetrators, or they could be completely innocent. And that is very important for the, for the course to to address. I think it's really important to, discuss difficult cases because, if we are having these issues, then I'm sure the other other colleagues are having exactly the same issues. So I think that the activity level is still at quite an early stage of development, actually. For example, the research that we recently carried out with the React project, we have we have a collaboration of 20 laboratories all doing the same experiments, but the results are actually quite different between the laboratories. So then we have this fundamental question as well. Okay. Can we actually, you know, trust results from laboratory a if I'm laboratory b and my methods are different? Can I actually report their results? And it's these sort of questions which we are, dealing with, and they actually seem a lot more research needs to be be done in this, area. And I I think that is the reason why that we see this disparity between, countries and and even laboratories. I mean, I think it's not true to say that there are many laboratories in Europe that routinely do activity level reporting, for example. And I think this is just something which we're working through at the moment. I want to come back to the silencing of of, activity level in this particular case. You know, in a nutshell, there is a DNA profile found on the victim that not even the court disputes that it comes from the defendant. The the the first court ruling, leads to additional analysis, and this analysis confirms the presence of this profile and additional profiles. I would not be surprised if some courts or people acting in this scenery would see this as a confirmation that the defendant was at the crime scene. And I think it was a very thoughtful decision of the court of appeal not to put too much emphasis on the presence of the, you know, the confirmed presence of this profile. But on the contrary, this additional finding was weakening the position of the of the prosecution, really. The other profiles in that case became almost irrelevant in the line of the investigations. This clearly shows that it is our responsibility to clearly state the limitations of the results that we report to the court. And, they might be statistically very powerful, and they may may have, confirmatory quality, but that is not the full picture of this case. And, and since we don't know whether or not the defendant was the true perpetrator, we have to weigh the evidence in its full complexity. And in this particular case, there were DNA profiles, quite different DNA profiles that couldn't be assigned to any source. I think the the case does, speak to the importance of activity level propositions. And and the point was made that, the the level of activity level proposition reporting is is very, very different between laboratories. And this is something that which we really need to to focus on. To do activity level reporting, it's not actually necessary to do the experiments. You can still talk to the subject. The important thing is to talk about, the prosecutors fallacy, to talk about confirmation bias. You must talk about the limitations of the evidence. You must talk about what the evidence tells you and what it does not tell you. That's really important. I gave an example of a a a Bayesian network, but I didn't have any probabilities in this. And I just used think of it really as a flowchart because these networks are really good ways to explain, to a court the various modes of trans transmission of of DNA via, secondary transfer or direct transfer. So I I do think it is really, really important that laboratories, you know, start reporting activity level. They don't need to give statistics, like, when they do this, but they do need to understand it. So I just have a very simple file message, and that is never let a strong answer to the source question masquerade as an answer to the activity question. I was approached by the police in Norway about this case, so, some years ago, and and, and they they they did ask me, you know, whether whether they should continue to investigate it. And I said, well, yes, of course of course, you should. Even though we accept that there are all these problems with, potential contamination, etcetera, it's it's not really an option not to investigate because you just don't know where it's going to lead to. In in a case where a a cold case, quite often, it may lead to other evidence. The DNA evidence is just part, of the, of the case, of course. But but often the DNA evidence leads to other physical evidence which can be used in the case and the case builds on that. It didn't happen in this case course, of course, but of course, when you carry out cases like this, you just don't know what's going to happen. But that's never a reason not to try. This episode asks what happens when an identification meets a court that demands more than a match. Hierarchy of propositions, unknown contributors, and confirmation bias all played a part in this case outcome. Peter, Arnaud, Walter, your combined decades of work have shaped how the forensic community draws the line between identification and activity, and we're honored to have you open this series. Before we close out, this session will be available on demand for you to rewatch anytime. You'll also receive an email with a link to your continuing education certificate. Please complete the survey to download. Now we want to hear from you. Let's get to your questions. Walter, Arno, Peter, so good to see you guys, and thank you for giving us some of your time to kick off HIDS three sixty five. How are you all? I'm pretty good. Thanks, man. Pretty good. Yes. Yes. Good. Thanks. Great to see you. So we only have a few minutes for questions and answers. I would just wanna jump in for some housekeeping for the attendees who are joining us today. Please put your questions in the chat if you have them. I have a few questions for you all, and I have to say after this, seeing this video for the first time, my head is absolutely spinning. Thank you for the work that you did on this case and for giving us your time. When I think about the Tangs case, it has it's so rich from a forensic DNA perspective, but also it introduces many key questions for forensic scientists to think about as we, introduce evidence into court, and I wanna ask a few questions about that. First off, we had autosomal STRs, Y STRs, mitochondrial DNA, using both CE and massively parallel sequencing, not to mention the sub source and activity level questions. So when you think about, the case overall and the work that each of you did, how do we how do we set up the forensic scientists in relationship to the court? I think the the comment was made, scientists evaluate, courts adjudicate, and I wanna start with you, Peter, first. Can you expand a little bit on that comment? Yes. I I I think I made the the point, which I think is really, important is that there there is a specific role for the scientists to carry out. But there's also a very important role for the court, as well. The court I think the court has to be, much more involved with this this process. For example, It's very difficult for the scientist to decide, you know, what propositions need to be evaluated if there is no communication with the defense, for example. And and this is why I I talk about proxy, proposition. But, of course, if we got start going down that route, then, of course, the court needs to be involved to decide whether these proxy propositions are relevant or accepted in the particular, case case in hand. And and so this is where courts need to be, much much more involved, I think. Thank you for that. Walter, can can you add any any points to that as well? Well, one of your question or part of your question, Matt, was, in the direction of the technology, so talking about capillary or electrophoretic system for that matter and massively parallel sequencing. And it has been mentioned in this episode multiple times that, rather than giving up on a case, we should try because first we need to see the data in order to make, the interpretation. I would like to add to this that, it is, of course, important to understand that we are using, evidence. So when when we do a DNA analysis, we are using up evidence. So what we mentioned in the in the, the video still holds true, but it has to be covered with the reasoning that the the methods are fit for purpose in order to perform these kind of analysis. Now, with respect to the different platforms, the electrophoretic system that we used, around the accreditation accreditation in our place, so these are the results that we report in the expert opinion, but they do not, convey the full picture that we get we would get this massively parallel sequencing. And, this tanks case also shows, that we can get additional information, like additional markers with them yet. But we can also see kind of attributions in a mixture that we wouldn't see with electrophoretic systems. So that is catalysts from mitochondrial DNA here. So that speaks for the fact that, there is a reason why we move technology and why we invest in new methods because it will help us to, to solve cases that are not solved with old methods. Excellent. Thank you, Walter. And Arnaud, a question for you. When you think about the Tangs case or any case that has, multiple different biological markers, autosomal STRs, Y STRs, mitochondrial DNA, how should scientists and the court be thinking about the relative value of those markers and then the weight of that evidence when it goes into the courtroom presentation? Yes. There's clearly a difference between, the the weights that autosomal markers have and YSCR markers, can have or even snips or mitochondrial DNA. Each has their own strength in, supplying information in the case and should be evaluated accordingly. For instance, autosomal markers are great for individualization of of samples to to be very specific in who the donor can be, while lineage markers, the mitochondrial DNA and the y DNA can give you good information about the the the families, the the lineages. So clearly, there's there's differences that need be to be evaluated. Excellent. Thank you so much. I have a question in the chat for you, Peter. I'll read it and you can respond. It says, Peter, you mentioned the varied levels of reporting globally and the importance of giving limitations to the court. What do you see as the most feasible approach now for DNA scientists when asked activity level questions during court? I think that, every statement should, have a discussion about the dangers of the, prosecutor's fallacy, and I think you need to outline the, the issue of confirmation bias. I think these are the the the two most important things. You don't necessarily have to have data in order to discuss the activity level. And I realize that a lot of labs will be in this position, but this does not prevent you from actually pointing out to the court the limitations of the evidence. So the best way forward really is to have this all in your statement. So the court must be aware of the limitations of the evidence, in particular, that you cannot transpose the sub source likelihood ratio to the activity. You have to warn the court that these are two completely different likelihood ratios, etcetera. You have your caveat in there about the prosecutor's fallacy, and you also have, information, about confirmation bias. And I think that, for example, if you have found unknown individuals in your profile, then you should point you should not dismiss this as purely as background, and you shouldn't write in your statement that this is what we expect to see, because because it might those pro those profiles might actually come from a a true perpetrator. You just don't know. So you you have to, lay all of this out in your state. And you can do this without necessarily assigning probabilities. Thank you for that. Yeah. I think, Walter, you made the statement that there was too much of a focus on a a 12 f versus the hair samples. And when you think about a case like this that has many investigative questions but also a lot of different types of evidence, what is the role of the forensic scientist to answer those questions through DNA testing? How far do they go, and when do they, feel like they've done enough for the for the questions at hand? Well, first of all, I should mention that at least, the way how we work here in in our laboratory is that we can only perform the analysis that we are ordered by either the court or the the prosecution depending on who is our our, client. But what we do, do and that is, in line with what Peter said about the system, we are discussing the analysis upfront with the court, with, the prosecution, and sometimes with the defense depending on the case. And, in in most of the cases that I was involved in, they don't they are not experts in in in the in the methods that we use. They they know some parts about the DNA analysis from hearsay, but they are certainly not experts. So I think our responsibility, also besides in providing them information which kind of analysis is fully makes sense and therefore, suggest to be done in a particular case, but we would not perform this without the explicit order. Thank you. And then back to you, Arnaud, when you think about the differences in the court systems across countries, the adversarial versus the inquisitorial, when a case is complex like this, is there a system that's in a better position to get to the truth? Oh, that's a hard question. What what I particularly liked in the in the in the court in Norway is that the scientists were allowed to be present, during their testimony. So we could, as scientists, listen to each other's testimony and, respond to that, ask ask questions. And I think that gave a better explanation of all the science behind, our our work, and a better, information for the courts to to base their judgment on. So I think that is quite particular, for for the Norway situation, but it was really helpful. Thank you for that. Yeah. Peter, I think you said that the adversarial is quite quite like a boxing match. And for those of us who have testified. in court, you know, sometimes we as scientists are in the middle of that match. You're you're you're very that's very true. Yes. Yeah. Yes. So, I mean, just just hearing you all talk about this case, I think my preference would be, go the more of the inquisitorial route, be able to have that open dialogue, evaluate the science, but also the investigative questions, and ultimately arrive at a consensus, right? That's what we should all be doing as part of, part of a criminal investigation, would you agree? Yeah. Yes. Very very much so. The, adversarial systems can be more friendly. For example, in in The UK, the if there is a dispute between the defense and the prosecution, the judge can order the two parties to get together and and prepare a joint statement. But my understanding is that in The US, this would would never happen. But I think that is a a very good way forward is that you you need to get the experts together. The the court will probably be confused by a lot of this evidence. That's the problem, and it doesn't need to be burdened by the areas which the experts disagree on sorry, agree on. It doesn't need to be burdened by the areas which the experts agree on. The court really should only be concerned where there are points of disagreement, and it needs to have these points of disagreement laid out. And the best way to do this is is with the actual experts to actually get together and, and discuss it. So the the it's really important for the courts to provide an environment for that to happen. If it doesn't provide the environment, then I think it then justice will will suffer as a result. So that is that is my message. The court must provide the environment for all of this to happen, and my worry is that the court is not necessarily providing this environment. So there must be a wider discussion. It's not just the sciences. It's the environment that that we work in. That's a great that's a great takeaway, I think, for for myself and the attendees, and, we're gonna close on that comment. But I another you do have another question there, actually. Sorry to butt in, is you've you've got, that's okay. question there which you which you've missed. Have we got time to do that? It's you're looking. about the Bayesian network. Yes. I can. cover you wanna comment. quickly on that one, we have about a minute. left. Yeah. Because I I think it's a very important question, actually, which is why I butted in. But, obviously, my point is that you don't necessarily have to have the probabilities to help the court with this kind of reporting. And in this case, I did not have the the the probability. So the question is how can you give the court the information? Now a Bayesian network, can be thought of as a sort of a flowchart. And in that flowchart, you've got all of the possible means of transfer. And so effectively, that's that's what I showed. But you can see it. in the paper. Sorry. I think we've run We have run out of time, but I I wanna thank you all again. I just wanna close with a comment that our node made that as scientists, we should never give up on a cold case. This is a perfect example of that. Identify as many unknowns as we can and and do the best that we have with the evidence in front of us. So I wanna thank you all again for your time. For the attendees, I just wanna remind you that episode two of hits three sixty five will be coming up on May, and we hope to see you all there. Thanks so much.