All right. Good morning, counsel. Back on the record in the Simpson matter. Mr. Simpson is again present with counsel, Mr. Cochran, Mr. Scheck, Mr. Neufeld. The People are represented by Mr. Clarke and Mr. Darden. The jury is not present. Counsel, is there anything we need to take up before we invite the jurors to rejoin us and the conclusion of Dr. Cotton's testimony?
Two matters, your Honor. One is--one has to do with the cross-examination of Dr. Cotton. In particular, with regard to the National Academy of Sciences book "Reporting DNA Technology in Forensic Science." What happened on Thursday and Friday, your Honor, is that the witness, who obviously is extremely familiar with this book, has read it very carefully, in fact, carries her own copy with her in her briefcase, was allowed to avoid being impeached by that book by simply saying as to those portions of the text which she disagrees with, she doesn't rely upon. I think if we allow that type of approach to continue--
Do you have any case authority indicating that the Court's rulings on this were inappropriate?
Your Honor, I haven't seen any case that stands for the proposition that a witness can selectively pick and choose which sections of a learned treatise she relies upon solely to avoid saying that--I'm sorry--solely to enable her to say that for every section that I disagree with I don't rely on it so you can't bring it to my attention. It would stand the entire rule of impeaching a witness by learned treatise on its head. That is what is so remarkable with this and that is what is going on here, and I don't think that should be allowed to continue. And I have never seen any case anywhere that allows a witness to draw those kind of distinctions between, well, page 365 I agree with, but page 366 I don't and so forth and so on. Obviously she is only going to then rely on those sections she agrees with so one could never under that approach ever use a learned treatise to impeach a witness.
Your Honor, I think Mr. Neufeld mischaracterizes what the witness has done when he states that she has done this intentionally to avoid impeachment. The truth is she has relied on only portions of that material because those portions are scientifically appropriate and other portions are scientifically inappropriate, so I think to characterize it as selective reliance to make testimony sound better is absolutely absurd in the context of this witness' testimony. The Court asked Mr. Neufeld do you have any authority to demonstrate why the Court's previous rulings are incorrect and Mr. Neufeld simply asked the Court I don't know of any authority that says to the contrary. That is what demonstrates that it can be done the way the Court has said. This Court has previously ruled on this matter now a number of times and I think the Court's rulings are absolutely appropriate under the law, because otherwise the Court would be allowing inadmissible hearsay to come into court.
Your Honor, the only point on that, very briefly, is that it is her opinion that certain portions are scientifically inappropriate, but clearly those portions of the report that she regards as scientifically inappropriate were written by the same people who she felt wrote other portions of the report that were scientifically appropriate. The only point I'm trying to make, your Honor, is I can't believe that it is an appropriate attitude within the scientific community to simply reject out of hand, which is what the rule requires so far, reject out of hand those statements, those recommendations written by people in a book in a learned treatise in the same book in which you agree with other portions of it and so those statements can come in. It is setting up some kind of artificial construct which I think, you know, completely dilutes the entire purpose of impeachment with a learned treatise.
Yes, your Honor. The second point has to do with the Court's ruling that the People are required to produce the aggregated frequency for mixtures, and that issue will be addressed by Dr. William Thompson.
Good morning, your Honor. Yes, we do have a problem regarding the statistics to be presented in connection with mixtures, such as the mixed stains in the Bronco and stain 78. Last week the Court ruled that statistics must be presented in connection with those mixtures so that the jury would know what to make of them, what to make of the fact, for example, that Nicole Brown Simpson is included in the Bronco mixture.
Over the weekend we received reports from the Prosecution regarding the statistics that they intend to present and we have serious problems with them. We think that the reports that they have shown us of what they have learned to do failed to comply with the Court's ruling on this matter and we have a series of objections to them that we want to state now.
I think at the time of the Court's ruling last week I think we all contemplated that the statistics in question would be frequency statistics to be computed in accordance with the national research council's guidelines. So, for example, to allow the jury to evaluate what it means that Nicole Brown Simpson has genotypes consistent with certain stains in the Bronco, the jury needs to know what percentage of people in the population at large have genotypes consistent with those stains and what percentage would be excluded. And the national research council provides a very simple straightforward way of doing those calculations by summing the matching and non-matching genotypes. What we got from the Prosecution and what they intend to present apparently is not frequencies, but something called likelihood ratios. Likelihood ratios are a ratio of two conditional probabilities, where a conditional probability, as you probably know, is a probability of one event calculated on the assumption that certain other events occurred. These likelihood ratios--we know of no precedent for the admission of likelihood ratios in a criminal matter. In the notice of objections we filed with the Court in March we stated a series of objections to likelihood ratios. All of the objections we stated in our previous memo apply to the likelihood ratios they are offering here, and more. We think the particular likelihood ratios that they are offering that are found in a report prepared by Professor Weir of North Carolina University are lacking in foundation because the assumptions that are predicate to the conditional probabilities don't necessarily jive with the facts of this case and in some ways are completely inconsistent with the Prosecution's theory of the case. So what they have come up with is not something that complies with the Court's order; it is a convoluted attempt to do an end run around the Court's order and present a complex form of statistical analysis that is unprecedented and that we think is quite misleading. Should they--should they be allowed to present this or attempt to present it, we think a 402 hearing may be necessary to evaluate whether these likelihood ratio statistics are in fact admissible. We think this is a failure to comply with the Court's order. We ask that they be ordered to present the appropriate statistics computed in accordance with the national research council's method, or that the DNA picture evidence be struck from the record.
Yes, your Honor. Again this raises the question that I broached with the Court Friday. What numbers does the Defendant want? I mean, perhaps we can agree on that. But more importantly, what the Defense received was materials from Dr. Bruce weir that this witness, I believe the testimony will show, understands the calculation process, can perform it herself and will do so. Contrary to the Defense's assertion, there is no intent on our part to introduce likelihood ratios. There is a good deal of data in this report that we are referring to, it is approximately twelve pages in length, and it includes frequency information as well that we plan or I plan to elicit from Dr. Cotton during testimony.
Your Honor, would it be possible for Dr. Cotton to answer the Court's question instead of it being through myself as a conduit?
No. Dr. Cotton, good morning. What is it that we are going to be hearing and how will it be presented?
Let's take the RFLP separately from the PCR for a moment. Although the calculations are done the same, for the RFLP we intended to just discuss the two genetic locations where on the individual film you have evidence of a mixture that will be MS1 and G3. And to simply give the figures for what is the probability, you would pick a pair of people from the population that would have this combination of genotypes that would give you this result. The same--so for the RFLP it is simply giving that figure for the MS1 location and giving that figure for the G3 locus. For the PCR the figures are done likewise. That is, what is the probability you would pick a pair of individuals from the population, so it takes into account all possible pairs that could have given this distribution--this set of alleles, and sums those, and then says what is the probability it would pick two pairs of people that could give this? In his report Dr. Weir does provide information about likelihood ratios, but it was not our intention to discuss that.
All right. What are the numbers that we are going to get on the RFLP and the PCR aggregation?
The ballpark figures for the RFLP range sort of like 1 in 80 to one in either--it is either in the hundreds or thousands. It is not in tens of thousands. The PCR numbers are similar. The reason that they are similar is that for the RFLP you are just looking at a single locus and giving information about that locus. The PCR you are giving the composite information across all of the loci that we have done for PCR.
To give the Court an example of sometimes the complexity of this, if one were to take the boot drop, item no. 78, and determine what is the probability of finding the 14 bands that were located, it is, for instance, approximately two with about twelve zeros behind it. I mean, they are astronomical, so it all depends on how the question is framed. It is not our intent to present that type, at least during the direct examination, but I think the Court is getting--I'm sure the Court is getting a feel for characterizing this question can be difficult, but I think we are--at least it is our intent to approach it in a very conservative fashion.
Your Honor, I think it is clear from what has been said that what they are offering is not what the national research council called for.
Counsel, you have the same information before you. What numbers do you come up with?
Let me show you. You know, I prepared a chart, your Honor, showing what we think is the correct computation with regard to stain no. 29, the Bronco steering wheel, and with your permission could I show you and counsel a copy?
Now, your Honor, the frequencies in this chart are taken directly from Cellmark's own frequency data that was disclosed to us in discovery from DNA page 1520. And in order to compute the probability or the percentage of the population that would be consistent with this mixed stain in the Bronco, all that is done is that one lists all of the possible genotypes that one could have that could be consistent with that stain, computes the frequency--adds up the frequency of each of those genotypes in each of the relevant populations and sum them. So we find that among Caucasians 45.4 percent of Caucasians would match stain 29. 59.2 percent of African Americans, 48.8 percent of Hispanics would--would be included--included in the stain. And the percentage excluded would be the complement, simply taking these numbers and subtracting them from one. This is what the jury needs to know in order to evaluate, for example, the fact that Nicole Brown Simpson's genotypes match the stain. These statistics are not what the Prosecution is offering. They are offering conditional probabilities. Even if they don't frame them as a ratio of conditional probability, they are still conditional probabilities. That is the probability of one event conditioned on a set of assumptions. What Dr. Cotton discussed was the probability of a pair of people giving certain results under certain assumptions. Now, the problem is the assumptions. The conditional numbers that Dr. Weir computed, for example, were all conditioned on the assumption that the number of contributors to the mixed stains was two individuals and only two and that somehow we knew that. Now, that might be an appropriate statistic in a case where we know the number of potential contributors to the stain, for example, a rape case where a woman says that she was raped by two men, but in a case like this where we don't know the number of contributors, it is without foundation and irrelevant, prejudicial and completely inappropriate to allow them to present statistics premised on certain assumptions when they cannot demonstrate those assumptions are true. The arguments become circular. In order to--in order to prove the Prosecution's theory of the case is correct, they offer statistics that are premised on the assumption that the Prosecution's theory of the case is correct and we go around and around in a logical merry-go-round and that is why they are able to come up with numbers like 1 in 100 or 1 in 50 when the NRC numbers are 45 percent, 59 percent, 48 percent. So, you know, this is--this is not appropriate, it doesn't comply with the Court's order, and it is not even consistent with their own chart. Their own chart, when they list numbers is headed "frequency." They are not offering frequencies; they are offering these complex conditional probabilities premised on sets of assumptions which are going to be controversial and which I think may not match the facts of this case. For example, if the Prosecution's theory of the case is correct, the number of contributors to the steering wheel stain has to be at least three, and yet the statistics they are offering on that are probabilities based on the notion that there are only two contributors and the statistics ask what is the probability that by randomly choosing two and only two people you would produce the exact results that we see here? That simply is not relevant to the facts of this case.
Yes. With regard to--first of all, Mr. Thompson's characterization is wrong as to the Prosecution's theory that there is three donors on the steering wheel. All that has been reported by Dr. Cotton thus far is that there are types from which two individuals cannot be excluded and there is another allele and that is not inconsistent with two contributors. We know there is at least two because there is a picture. It may very well be only two.
I'm more interested not so much in the theory of the Prosecution's case, but in the method in reaching this calculation in the first place. Have you had the opportunity to review the proposal by Professor Thompson with Dr. Cotton? Because this appears to be the aggregation that is recommended at page 59 of the NRC report.
That is flexible with the Court's direct direction. If cross-examination is finished and then I can conduct my redirect, which at this point is not substantial, then perhaps--I see a smile on the Court's face--then perhaps we could address this issue once Dr. Cotton and I have had a chance to discuss it.
All right. Let's do that. All right. We have framed the issues. I think counsel should have the opportunity to go over with Dr. Cotton what these numbers are. Professor Thompson, do you have a similar calculation for the other stain?
I don't have it right now, your Honor. I would be happy to discuss with Dr. Cotton how we think that should be approached.
Okay. All right. Then I propose that we proceed with the jury, that probably what will occur is Dr. Cotton and Mr. Clarke will have the opportunity to confer over the lunch hour, and let's start with the jury now.
Your Honor, if I may, could I ask that Dr. Weir's report be marked as an exhibit for the Court's purposes only?
I can't believe that it is an appropriate attitude within the scientific community to simply reject out of hand...those statements, those recommendations written by people in a book in a learned treatise in the same book in which you agree with other portions of it.
In order to prove the Prosecution's theory of the case is correct, they offer statistics that are premised on the assumption that the Prosecution's theory of the case is correct and we go around and around in a logical merry-go-round.
45.4 percent of Caucasians would match stain 29. 59.2 percent of African Americans, 48.8 percent of Hispanics would be included in the stain.
If one were to take the boot drop, item no. 78, and determine what is the probability of finding the 14 bands that were located, it is, for instance, approximately two with about twelve zeros behind it.