Bad Science
Learn from yesterday, live for today, hope for tomorrow. The important thing is to not stop questioning. Albert Einstein
Chapter 16
Bad Science
Sitting in the courtroom listening to the Judge’s charge, my mind would sometimes wander.
This trial had forced us to all live together again. Six adults in the same house – for five weeks. With so many living under one roof in our house, I needed to remain organized. Sometimes my mind ruminated about making grocery lists of what to pick up after the day in court. It was usually milk – we needed more milk. We might need a little ice cream. We might need comfort food after today.
Living through an emotionally stressful time, we all needed our comforts.
One of our adult children had decided to bring a DVD with three seasons of the TV show, The Big Bang Theory, about Sheldon and Leonard, their friends and, of course, Penny.
We needed to be together without actually having the energy to interact; so we would sit around the living room, some on the floor, some on the sofas and some at the living room table, with our individual computers and do whatever. Larry was usually on Facebook, skyping, listening to sermons and watching YouTube. Odia read books on her iPod touch, played Angry Birds and other games. Our son, Syras, read his Kobo and Natasha looked up healthy recipes and exercise programs.
Cliff was studying King David and I blogged daily throughout the trial.
However the Judge didn’t seem to struggle with his focus as he continued his charge.
“It is the position of the Defence and its expert, Dr. Waye, that the results themselves are an affront to the scientific method and that the results seen as they are, are not sufficiently conclusive so as to permit a determination that the Crown has proven beyond a reasonable doubt that Mark Grant was in the shed, that he forcibly confined Candace Derksen, or that Mark Grant abandoned the hogtied Candace Derksen in the shed.”
Dr. Waye! I remembered the name.
It was February 19 at 9:21 in the morning. The Judge opened the morning proceedings. “Good morning, ladies and gentlemen.”
“Good morning, My Lord.”
“Mr. Simmonds?”
Simmonds stood up, there was a new energy around him. He seemed excited.
It was odd to see him take the lead.
Apparently when the science of DNA was first introduced into the courtroom, it had been seen as such an incontrovertible truth that the Defence lawyers had no way of even challenging it.
Now it seemed as if Simmonds had a breakthrough. We heard that he was championing a new strategy. He had a new strategy of creating doubt and was already giving workshops on DNA as the trial was proceeding.
“Ladies and gentlemen,” Simmonds turned to the jury. “It's been a number of weeks already, but I’m going to ask you to bear with us for a little bit longer. From our perspective, the evidence that you've heard should already leave you in significant doubt with respect to the issues of whether or not Mr. Grant is in any way responsible for the events that are before, before you. From our perspective, the DNA evidence is lacking, that the evidence and the theories that have been put to Dr. Chahal should leave you in significant doubt with respect to everything that you've heard up to now. However, there have been many things that we put to Dr. Chahal with respect to that that you may still have wonder about. And to ensure that you're in a position in which you have clarity, it is our position that a doctor of genetics, with significant experience in this area, with a long history and an impeccable - from our perspective, an impeccable résumé, will assist you in making that decision and the correct decision, from our perspective, which is a not guilty finding with respect to Mr. Grant. Without being any longer or without belaboring the point, we would call Dr. Waye. I have for the court a copy of Dr. Waye's CV, which we're going to be asking to be entered, and copies for the members of the jury, My Lord.”
It was impressive.
Dr. Waye was introduced as a tenured professor in the department of pathology and molecular medicine at McMaster University in Hamilton, Ontario with a Bachelor of Science in Microbiology from the Department of Microbiology at the University of Guelph in 1981.
He had received his Masters in biology from the Biology Department at McMaster University in1984, his Ph.D. from the Department of Medical Genetics at the University of Toronto in 1987 and had the prestigious position as Fellow of the American College of Medical Genetics. He was also a diplomat on the American Board of Medical Genetics and held a number of professional positions, such as the head of the Molecular Diagnostic Genetic Service of the Hamilton Regional Laboratory Medicine program, Hamilton, Ontario, from 2001 to the present. He also served with the RCMP at the Forensic Biology Section from 1988 to 1990. He was member of the Ontario Genetic Services Strategy Committee, on the Biology Section Accreditation Committee for the Canadian Society of Forensic Science and on the Technical working group on DNA analysis method.
Probably most telling was that he had testified in 41 court cases – most well-known was the Robert Pickton case where he was a consultant to the Defence team from 2002 to 2007.
It felt like the battle of the resumes.
Simmonds stood close to the witness stand. “Dr. Waye, I'm assuming you saw the electropherograms reports prepared by Molecular World?”
“Yes.”
Then Simmonds went to the old-fashioned flip chart that had been standing in the courtroom when we had arrived in the morning. “When we look at the chart, for our purposes we’re looking at Exhibit # 25. Are you able to see that from there? Would it be better if I bring it closer? How about if we try it this way so the jury's also able to see it.” He shifted the flip chart this way and that way.
We watched hungrily. The chart was in the sacred place – that open space between the bench and the counsel tables. This space is called the “well.” It is extremely disrespectful to the court for persons who are not court employees to directly “traverse the well” without permission. During trials, attorneys will ask the court's permission to traverse the well or “approach the bench” for “sidebar” conferences with the judge.
“Is Your Lordship able to view it from that position?”
“I see most of it. Actually, maybe if you just ….”
They turned it this way and that way. But we couldn’t see it at all.
And then it began – the discussion of the DNA.
The first question was about the amplified DNA profile, whether anyone who didn’t match that profile on the twine should be excluded.
Dr. Waye agreed. “When you’re making a comparison, if you amplify up somebody’s DNA and compare it to another sample, they either match or don’t match.” He nodded.
I was shocked that they used the forbidden word “match” in the opening of the direct examination. I wondered if Bell might object but apparently this line of questioning would be different. The rules had changed.
Simmonds continued, animated. “And just so that the ladies and gentlemen of the jury understand, what exactly, again, is SWGDAM?”
Waye answered. “It's a group of international scientists who work on forensic DNA, and they meet periodically, again, to come up with recommendations and standards for the forensic community analyzing DNA for forensic purposes.” One of their guidelines read, “... statistical analysis should be made prior to comparison to the known profiles.”
“What does that mean?” Simmonds asked.
“It means when you analyze a sample you should look at the profile of the evidence sample without any knowledge of what you're comparing it to, and make your judgments of which, which parts of, of that profile are suitable for comparison,” Waye answered.
“Now, why do you do that before you look at the known sample?”
“Well, the human tendency - it happens with scientists, happens with everyone - is that if you know two things that you're comparing to, you're drawn, you’re drawn to, to making a match. So you can't be biased looking at and saying, I've got the known profile here and I've got this here, I've got things here that don't line up. And then you, then you try to find reasons to exclude that, that information.”
“Why is the nature of the way in which they structure the investigation or experiment so important, from your perspective?”
“It's the basis of science. In science, you ask questions and then you conduct experiments to try to answer those questions. It's called the scientific method. And it doesn't matter whether you're - what system you’re working in, whether you're working forensics or medical genetics or basic research, it's always the same process. You come up with a question, you form a hypothesis, and you design an experiment to test the hypothesis. You get your results, you analyze the results and either accept your hypothesis at the beginning or you reject your hypothesis. If the data rejects your hypothesis, you go back and you revisit your hypothesis and you start again. You try to come up with reasons why; you do another set of experiments. So if you follow that method, you're not going to make mistakes. You're not going to come to false conclusions. That's the backbone of science and all science works this way…. You do this so your interpretation won't be influenced by your expectations.”
“And from your perspective, Dr. Waye, do you have any opinion with respect to the way this particular investigation was conducted on that issue?”
Waye answered strongly. “It's, it's flawed. It doesn't follow the scientific method.”
In other words it was bad science.
Simmonds continued. “Now, when you look at those results and compare those with Mr. Grant’s analysis, there were three places he should be excluded. How do we understand what's happened here? Can you explain to us what's going on? Because if he should have been excluded, that's not apparently what Dr. Chahal says.”
Waye answered. “Well, again, in the scientific method, if you compare the two profiles, you're either going to accept your hypothesis that the DNA originated from Mr. Grant or you're going to reject your hypothesis. In three locations here, you're not amplifying up alleles that are present in Mr. Grant's, so when you look at that data, you're left with rejecting your initial hypothesis or reinvestigating your, your hypothesis, doing further experiments to try to explain why those alleles are missing.”
“…but what did Dr. Chahal do?”
“He excluded all of the data from those three loci.”
It took me awhile to understand this. Apparently they were talking about three markers that had been overlooked deliberately. Chahal and Hildebrandt had found thirteen – ten which matched Grant, three which didn’t.
Simmonds moved the flip chart to the side, still in view of the jury but not quite as prominent in the room. He stood close to the witness, still looking at the jury. “Perhaps we can go on, then, Doctor, because from our perspective, somehow in the Chahal evaluation, he, having the same findings, then goes to a theory about 250 base pairs.”
“Correct.”
“Now, is there any foundation in science for his position on 250 base pairs?” Simmonds repeated the question a few times and in different ways. “Are you aware of any theory that would allow someone to just reject everything at 250 base pairs?”
“No. If you knew for a fact that your DNA fragments fall below that level, you're not going to amplify things that are over 250. Then you wouldn't need an imposed cut-off like Chahal did, because you wouldn't get results for over 250.”
At this point, Dr. Waye began to cough and reach for his glass of water.
Simmonds turned to the Judge. “Dr. Waye left his flu bed to be here.”
We were given a break. We walked out into the hallway, all very quiet. This was new information for us. It unsettled us – mainly because we had no idea what was going on. We were all disappointed that we could not see the flip chart. And we were all very curious about these three markers.
When we came back, Simmonds picked up the question.
“I thought either there was a scientific rule or there wasn't a scientific rule. Do you have an opinion?” Simmonds asked.
“Yes. You set that rule at the beginning, not at the end. If you wanted to say at the beginning everything over 250 base pairs is unreliable, you have to actually take those tests out of your kit because once you're faced with data you have to deal with it, you can't ignore it.”
“He's ignored data, Doctor?”
“He's got data on three of those loci that he is ignoring.”
They then went back to that flip chart and began to talk about alleles.
Referring to the chart, he asked Waye, “If alleles were going to drop off, which ones would normally drop off first?”
“The largest alleles - so if you're looking at locus D7, larger alleles would be 11 and 12. So if you’re looking at what was amplified, those would be much – those would be more likely to drop out before 8.”
And off they went into a world I could neither see nor understand. It was a hum of numbers for me.
Finally, Waye came to his conclusion. “So you've got two things happening there. The minor source where you do get an allele amplifying, and the major source you get a larger allele amplifying. So again, allele dropout I would consider unlikely, given these results.”
Simmonds put the next question clearly. “So now we have three different locations with unlikely positions that are being posed by Dr. Chahal where, from our - from your perspective, he should have been excluded. What do we do with that, then? Should have been excluded in three locations; Dr. Chahal just seems to throw that evidence out. Is that scientific from your perspective, Doctor?”
“Well, based on, on how we know these, these systems behave with degraded DNA, the idea of allele dropout here sort of flies in the face of how we know they behave. So it's counter to your expectation.”
Bell was on his feet. “Objection. That’s a leading question, My Lord, in a critical area.”
But the Judge overruled the objection. “He's seeking an opinion from an expert. For the purposes of clarity, I will let Mr. Simmonds pose the question as posed.”
Simmonds conceded. “Okay, I'll put it in a different - way ….”
Simmonds moved on. “I thought either there was a scientific rule or there wasn't a scientific rule. Do you have an opinion?”
Waye responded. “Yes, you set that rule at the beginning, not at the end.”
And they went through different angles of the scientific process again and again.
Simmonds asked, “When you approached all of this material, Doctor, and reviewed it all, from your, from your perspective, is this a scientific process that you have seen here?” And they went through it again.
Finally, Bell was on his feet in frustration. “With respect, My Lord, that's been covered about seven times during the course of my learned friend’s direct.”
Simmonds hesitated. “I’ll put it in a different way.”
There was another pause as Simmonds studied his notes, then looked up.
“What is reproducibility, Dr. Waye, the concept of reproducibility?”
The minute Simmonds said that, I knew we were being reminded again of the “last shot.”
What had seemed like one of the most courageous moves by Bradbury and Lutz, our champion investigators, was now under fire.
Waye sounded like an impatient science teacher at this point as he described the importance of reproducibility. “It is another hallmark in science - one of the things that’s demanded. One needs to replicate it especially if it's a result that's unexpected or may be controversial. So you either replicate it in your lab or have someone else replicate it for sure.”
The Judge leaned over… “Just”…. then thought better of it. “Sorry, go ahead. Go ahead.”
Waye continued. “In a case like this, if you have a theory that allele dropout has occurred, one of the things your lab would need to do is replicate it.”
“But in a case like this when in fact there is none left to try and run this again, how do you feel about these results under those circumstances, Dr. Waye?”
Waye shook his head. “That's unfortunate. Again, you're stuck with an untestable hypothesis. Unless… you do a validation study?”
“What are validation studies?”
“Validation studies are when a lab tries to simulate what they will see in a forensic case, using samples that they know what the results should be. So you’ll take a sample that you know exactly what the pattern is. If everything works fine, and then you do things to that sample. You'll subject it to insults; you'll expose to sunshine, bleach; you'll, you’ll intentionally add enzymes that will degrade the DNA. And you’ll see how all of that affects how your tests work. So you try to mimic what, what's going to happen in the, in the forensic world in the real world, but you know what the answer should be if it works.”
“Do you do validation studies in your lab, Doctor?”
“On every test before it's used on patients, it gets validated.”
“According to Dr. Chahal, there were no validation studies here. How does that affect the results we've been told about?”
“A lab shouldn't be doing a test unless you do your own internal validation studies,” concluded Waye.
And then Simmonds went back to the question of scientific processes.
“From your perspective, can you give an opinion as to the scientific validity of what you're seeing, then?”
“No validity whatsoever,” Waye answered.
“Because, of course, we've now seen the data and you've looked at the data that deals with exclusion. Now would you look at the final report that he produces?”
“The final report again at those three loci - D7, D16, and D18 - takes out all the results. So it takes out 19 what you did see in the actual profile, it takes out what in the previous profile he assumed was there and had 1 dropped out, and it just leaves it a blank slate and says that that data's unreliable because it's over 250. So now he’s come up with a rule that excludes that data.”
“Again, from your perspective, is that defensible in any way as a scientist?”
“No.”
“Those are my questions. Thank you, Doctor.”
Bell stood up.
It was now time for the cross-examination.
I wondered if Bell would change into a “Defence lawyer” mode, but he didn’t. He was very precise in his language, pointed and brief.
“Well, with respect to your position on scientific hypothesis, I noted Dr. Chahal was asked that question on direct and he said that what he was really doing was a test. Would you disagree with his interpretation of what he did?”
“There's a difference between a test and an experiment. A test - if I analyze my DNA and develop my profile, that's a test. It doesn't answer a question; all it does is tell you what my profile is. And you can test my height as well; it doesn't test the hypothesis. An experiment, which is exactly what's done here, is a test that’s done under controlled conditions to test a hypothesis, to test the validity of a hypothesis.”
“Well, when you testified in direct this morning, Doctor, you - your hypothesis seemed to be that Mr. Grant fit the profile. Now, if you're commencing, would you agree or disagree with that?”
“That his ...?” Waye was stunned.
“That appeared to be your hypothesis prior to testing.”
“Parts of the data, he fits. Parts of the data, he doesn't fit.” Waye said.
Bell continued. “No, no, maybe you misunderstand my question. You told Mr. Simmonds, I believe – and correct me if I'm wrong – that your hypothesis for this example was that Mr. Grant fit the profile, and then you said that you would conduct an experiment and then if the results didn't come out the way that you anticipated, you would amend or change your hypothesis. Was that what you told Mr. Simmonds this morning?”
Waye was horrified. “Not in this lifetime. No, that's not the way I do science. You start off asking a question: Could Mr. Grant have contributed to the sample or if it’s somebody other than Mr. Grant. You do the, the test.”
Bell persisted. “I'm not suggesting for a second, Doctor that I’m trying to get you to include Mr. Grant in anything here. All I'm saying is when I took the notes with respect to what you said about hypothesis, it struck me that your hypothesis seemed to be, at the outset, and Mr. Grant fits the profile.”
“No, it's something that you're going to test. It's, it's a hypothesis.”
“Yes?” Bell was attentive.
“It may be right, it may be wrong.”
“I understand that, but during your example this morning was that not your hypothesis? I'm not saying it was - ended up being….”
“It's the only question on the table: Could Mr. Grant be a possible contributor to this?”
“Okay.”
“That's, that's different than saying my hypothesis is that Mr. Grant is a contributor to this.”
Judge Joyal said, “Okay, you're answering my question now. That was the same question on the table for Dr. Chahal as well.”
Waye answered, “Correct.”
“So it wasn't his hypothesis, it was the question on the table, correct?” said the Judge.
“Right.”
“Okay, good,” nodded the Judge.
The penny dropped. It was a test – not an experiment, and a question rather than a hypothesis. It wasn’t bad science – just a different science.
Pure science is used to develop information to predict and perhaps explain phenomena in the natural world. But applied science uses existing scientific knowledge to develop more practical applications, such as technology or inventions.
Again, I didn’t understand all of the conversations regarding the markers on the flip chart that was turned away from us, but it seemed as if the bravado exuded by Dr. Waye in the beginning was very carefully being punctured by the Crown lawyer, now a fencing expert wielding a foil so quick that it was hard to see – except every once in a while, it seemed a light would go on when the foil found its mark.
We all knew that the “last shot” by the investigators was controversial. This also included “boosting” using a stronger primer – which was also controversial.
When Bell asked Waye if the use of the primer was legitimate, Waye answered, “That's one, one thing you can do. You can load more product onto the machine which will amplify the small peaks become larger peak heights.”
“Okay….” Bell nodded.
“It's called boosting.”
“Then you'd agree with me that this isn’t simply a situation where you load the machine, press a button, and you get the results. Sometimes you have to have a look and see what you get from the first test?”
“Yes,” answered Waye.
“And you have to make some adjustments with respect to the amount of PCR product that you might introduce into the machine.”
“That's one manipulation you can do, yes,” answered Waye.
Bell was getting his answers. “Okay. So you said that if you don't put enough PCR products in, you might get a small peak, but on the second test, if you put more PCR product in, you might get a bigger peak?”
“You will get a bigger peak.”
“Okay. Well, let's say you put a whole bunch of PCR product in for the third test. What might happen there?”
“Well, as, as you increase the peaks that, that you’d like to see, you also increase your background.”
“Okay.”
Waye continued. “So if you've got a peak that's below threshold, adding twice as much material will bring it over threshold, you may, you may actually create some other peaks.”
In other words, it was possible to push a sample so far it begins to pick up unrelated data.
Bell continued. “I think it was pretty clear both today and last week that Molecular World has sort of a minimum of 100 - base pair? Would you agree that's an acceptable minimum?”
Waye did not deny it but said, “In their lab, if that’s what they decide to do, it should be something that they've established as a minimum.”
“Fair enough.”
“It varies from lab to lab,” Waye said.
“Then you said if you put too much PCR product in, it might go over the top, and where would the top number be?”
“Again, that's something that each lab sets.”
“Okay. So you'd agree with me, though, there's some fine-tuning involved in an effort to get the, the best possible data.”
“Yes.”
With one deft stroke of his foil he had just confronted Waye’s assertions that one must always have a predetermined 250 base pair limit before doing a test.
Sometimes one experiments.
Bell continued. “And sometimes - although it's not encouraged - sometimes in an effort to get the best possible data, you have to use up all the DNA you have.”
Waye nodded. “Yeah, that's the reality of this business. You're starting off with a certain amount of material at the beginning and you have only so many chances to analyze it. And when you're done, you're done.”
“When you're done, you're done,” Bell repeated.
This took care of the validation and reproducibility aspect. In a perfect world, one does take those precautions, but this wasn’t a perfect world – the DNA was old.
“Now, with respect to calculations, Dr. Waye,” continued Bell. “I understand that something commonly used is called the Hardy-Weinberg Theory.”
This theory is a crucial concept in population genetics. The Hardy-Weinberg principle states that in a large, randomly breeding population, allelic frequencies will remain the same from generation to generation assuming that there is no mutation, gene migration, selection or genetic drift. This principle is important because it gives biologists a standard from which to measure changes in allele frequency in a population.
They went into a lengthy explanation of what this would look like given the DNA findings on the chart.
“You would look at the frequency that you would expect to see those results at those places, and then you would multiply all those results together. Is that correct?” asked Bell.
“Yeah,” Waye said. “So if you use the Hardy-Weinberg Theory, you take the frequency of that in the population, so say that's ten percent have a 13 allele.”
“Okay. So what you do is you take that number, you look it up in the database.”
“Right.”
“That gives you the frequency as to how many times you expect to see that in the general population,” suggested Bell.
“Right.” Waye was into it now. “If you assume that there could be something in there that I didn't detect and you want to be very conservative, you could say that there could be a 13 and any other allele.”
“I thought we were keeping it simple.”
“Nothing's simple in genetics.”
“Yeah, I get that.” Bell smiled.
“And the way to calculate that, again using the Hardy-Weinberg equation, is you take the frequency of the 13, multiply it by one minus the 13 which gives you the frequency of all the other alleles, times two. So it’s two - 2P times one minus P times two. And then you add those two together, and that’ll give you the proportion of the population who could have contributed a 13 allele.”
“You'll get a number from those calculations.”
“Right.”
Bell continued. “And then you take the number from, say, D3, TH01, et cetera, and multiply all those numbers together.”
“That's why the technology is so powerful, is that you take things that aren't individually very discriminating, one in ten, one in eight, but the combined process, going through all of the markers, you multiply them and you end up with one in a trillion.”
One in a trillion!!! I was overwhelmed with the numbers.
“So can you tell us, then - or would you agree with me that Dr. Chahal was looking for 15 markers when he did the tests that he did?” Bell’s voice was soft.
“That's what's included in the test, yes.”
“Okay. And that comes from the kit; is that correct?”
“Correct.” Waye seemed very cooperative.
“And if you got - if he had indications on all 15 markers using the formula you described, you'd multiply all those numbers over and over and over, basically 15 times.”
“Correct.”
“So you could conceivably get a number in the trillions. That fair?” asked Bell.
“If it was a single source sample, you would get something that is - the denominator's very, very large.”
“Crazy numbers,” said Bell.
“Yes.”
“Okay. And that's with 15 markers or loci?” Bell asked.
“Yes.”
Bell stepped back. “Now, when you worked for the RCMP you said that you sort of got them on track to do DNA work back in the late 1980s. Was that with PCR technology or RFLP? What was that?”
“RFLP,” answered Waye.
“How many markers were you looking at back then?”
“Four to six.”
Bell seemed surprised. “Four to six. And I take it once the RCMP got into PCR technology, certainly for a while at least, a long time, they were looking at nine markers?”
“Yes,” nodded Waye.
“So you'd agree with me, then, that the more markers you look at, the more discriminating it is.”
“Correct.”
“So looking at nine markers is better than, say, seven markers?”
Waye nodded. “It'll give you a more discriminating profile.”
“Okay. And obviously ten markers is better than nine markers?”
Waye had a hard time admitting it, but he couldn’t deny it either. He was quiet.
Bell consulted his notes. “Dr. Waye, just one last question. If you assume for a moment that when Dr. Chahal, with respect to the nuclear gene, the STR, made the cut-off at 250 base pair, if you assume that he's correct in doing that, is it true that the known profile for Mark Edward Grant would be the major contributor to the mix profile on the twine?”
Waye’s voice was strong. “If you blindly assume that he is correct in that assumption, then your conclusion that Grant cannot be excluded is valid, and the numbers that he gets for the nuclear DNA, one in 33,000 is valid and multiplying it by the Adam gene Y-STR is valid.”
“Correct.”
“Thank you, sir. No further questions.”
It was 12:30. The direct examination and cross-examination had taken all morning.
Here is another blog – some random thoughts.
*****
A string of miracles: February 2011
Over these last three weeks we have seen the Crown very carefully lay down the outline of the story of what happened November 30, 1984.
We now know that the accused frequented a party house of drugs and drinking behind the little convenience store next to the school that Candace attended.
We know that his girlfriend came forward with lies that could have sabotaged the six-week search for Candace by lying to the police. We know they were both running from something at the time and temporarily lived in an abandoned hole in the ground near railway tracks.
We know that Grant is a convicted, registered sex offender who raped another girl under the Nairn overpass on the same path where Candace was taken by force.
Whether everything we are learning will come to light, I have no idea and it no longer matters. It is enough for me.
I consider it all a miracle, actually a string of miracles, starting with the police investigation 26 years ago that ended in this court case.
It doesn't have to be conclusive for me to be satisfied.
I don’t think any of us will ever be the same again. We have the front row seats for the most profound drama of life.
There are very few human beings who receive the truth, complete and staggering, by instant illumination. Most of them acquire it fragment by fragment, on a small scale, by successive developments, cellularly, like a laborious mosaic. Anais Nin
Chapter 16
Bad Science
Sitting in the courtroom listening to the Judge’s charge, my mind would sometimes wander.
This trial had forced us to all live together again. Six adults in the same house – for five weeks. With so many living under one roof in our house, I needed to remain organized. Sometimes my mind ruminated about making grocery lists of what to pick up after the day in court. It was usually milk – we needed more milk. We might need a little ice cream. We might need comfort food after today.
Living through an emotionally stressful time, we all needed our comforts.
One of our adult children had decided to bring a DVD with three seasons of the TV show, The Big Bang Theory, about Sheldon and Leonard, their friends and, of course, Penny.
We needed to be together without actually having the energy to interact; so we would sit around the living room, some on the floor, some on the sofas and some at the living room table, with our individual computers and do whatever. Larry was usually on Facebook, skyping, listening to sermons and watching YouTube. Odia read books on her iPod touch, played Angry Birds and other games. Our son, Syras, read his Kobo and Natasha looked up healthy recipes and exercise programs.
Cliff was studying King David and I blogged daily throughout the trial.
However the Judge didn’t seem to struggle with his focus as he continued his charge.
“It is the position of the Defence and its expert, Dr. Waye, that the results themselves are an affront to the scientific method and that the results seen as they are, are not sufficiently conclusive so as to permit a determination that the Crown has proven beyond a reasonable doubt that Mark Grant was in the shed, that he forcibly confined Candace Derksen, or that Mark Grant abandoned the hogtied Candace Derksen in the shed.”
Dr. Waye! I remembered the name.
It was February 19 at 9:21 in the morning. The Judge opened the morning proceedings. “Good morning, ladies and gentlemen.”
“Good morning, My Lord.”
“Mr. Simmonds?”
Simmonds stood up, there was a new energy around him. He seemed excited.
It was odd to see him take the lead.
Apparently when the science of DNA was first introduced into the courtroom, it had been seen as such an incontrovertible truth that the Defence lawyers had no way of even challenging it.
Now it seemed as if Simmonds had a breakthrough. We heard that he was championing a new strategy. He had a new strategy of creating doubt and was already giving workshops on DNA as the trial was proceeding.
“Ladies and gentlemen,” Simmonds turned to the jury. “It's been a number of weeks already, but I’m going to ask you to bear with us for a little bit longer. From our perspective, the evidence that you've heard should already leave you in significant doubt with respect to the issues of whether or not Mr. Grant is in any way responsible for the events that are before, before you. From our perspective, the DNA evidence is lacking, that the evidence and the theories that have been put to Dr. Chahal should leave you in significant doubt with respect to everything that you've heard up to now. However, there have been many things that we put to Dr. Chahal with respect to that that you may still have wonder about. And to ensure that you're in a position in which you have clarity, it is our position that a doctor of genetics, with significant experience in this area, with a long history and an impeccable - from our perspective, an impeccable résumé, will assist you in making that decision and the correct decision, from our perspective, which is a not guilty finding with respect to Mr. Grant. Without being any longer or without belaboring the point, we would call Dr. Waye. I have for the court a copy of Dr. Waye's CV, which we're going to be asking to be entered, and copies for the members of the jury, My Lord.”
It was impressive.
Dr. Waye was introduced as a tenured professor in the department of pathology and molecular medicine at McMaster University in Hamilton, Ontario with a Bachelor of Science in Microbiology from the Department of Microbiology at the University of Guelph in 1981.
He had received his Masters in biology from the Biology Department at McMaster University in1984, his Ph.D. from the Department of Medical Genetics at the University of Toronto in 1987 and had the prestigious position as Fellow of the American College of Medical Genetics. He was also a diplomat on the American Board of Medical Genetics and held a number of professional positions, such as the head of the Molecular Diagnostic Genetic Service of the Hamilton Regional Laboratory Medicine program, Hamilton, Ontario, from 2001 to the present. He also served with the RCMP at the Forensic Biology Section from 1988 to 1990. He was member of the Ontario Genetic Services Strategy Committee, on the Biology Section Accreditation Committee for the Canadian Society of Forensic Science and on the Technical working group on DNA analysis method.
Probably most telling was that he had testified in 41 court cases – most well-known was the Robert Pickton case where he was a consultant to the Defence team from 2002 to 2007.
It felt like the battle of the resumes.
Simmonds stood close to the witness stand. “Dr. Waye, I'm assuming you saw the electropherograms reports prepared by Molecular World?”
“Yes.”
Then Simmonds went to the old-fashioned flip chart that had been standing in the courtroom when we had arrived in the morning. “When we look at the chart, for our purposes we’re looking at Exhibit # 25. Are you able to see that from there? Would it be better if I bring it closer? How about if we try it this way so the jury's also able to see it.” He shifted the flip chart this way and that way.
We watched hungrily. The chart was in the sacred place – that open space between the bench and the counsel tables. This space is called the “well.” It is extremely disrespectful to the court for persons who are not court employees to directly “traverse the well” without permission. During trials, attorneys will ask the court's permission to traverse the well or “approach the bench” for “sidebar” conferences with the judge.
“Is Your Lordship able to view it from that position?”
“I see most of it. Actually, maybe if you just ….”
They turned it this way and that way. But we couldn’t see it at all.
And then it began – the discussion of the DNA.
The first question was about the amplified DNA profile, whether anyone who didn’t match that profile on the twine should be excluded.
Dr. Waye agreed. “When you’re making a comparison, if you amplify up somebody’s DNA and compare it to another sample, they either match or don’t match.” He nodded.
I was shocked that they used the forbidden word “match” in the opening of the direct examination. I wondered if Bell might object but apparently this line of questioning would be different. The rules had changed.
Simmonds continued, animated. “And just so that the ladies and gentlemen of the jury understand, what exactly, again, is SWGDAM?”
Waye answered. “It's a group of international scientists who work on forensic DNA, and they meet periodically, again, to come up with recommendations and standards for the forensic community analyzing DNA for forensic purposes.” One of their guidelines read, “... statistical analysis should be made prior to comparison to the known profiles.”
“What does that mean?” Simmonds asked.
“It means when you analyze a sample you should look at the profile of the evidence sample without any knowledge of what you're comparing it to, and make your judgments of which, which parts of, of that profile are suitable for comparison,” Waye answered.
“Now, why do you do that before you look at the known sample?”
“Well, the human tendency - it happens with scientists, happens with everyone - is that if you know two things that you're comparing to, you're drawn, you’re drawn to, to making a match. So you can't be biased looking at and saying, I've got the known profile here and I've got this here, I've got things here that don't line up. And then you, then you try to find reasons to exclude that, that information.”
“Why is the nature of the way in which they structure the investigation or experiment so important, from your perspective?”
“It's the basis of science. In science, you ask questions and then you conduct experiments to try to answer those questions. It's called the scientific method. And it doesn't matter whether you're - what system you’re working in, whether you're working forensics or medical genetics or basic research, it's always the same process. You come up with a question, you form a hypothesis, and you design an experiment to test the hypothesis. You get your results, you analyze the results and either accept your hypothesis at the beginning or you reject your hypothesis. If the data rejects your hypothesis, you go back and you revisit your hypothesis and you start again. You try to come up with reasons why; you do another set of experiments. So if you follow that method, you're not going to make mistakes. You're not going to come to false conclusions. That's the backbone of science and all science works this way…. You do this so your interpretation won't be influenced by your expectations.”
“And from your perspective, Dr. Waye, do you have any opinion with respect to the way this particular investigation was conducted on that issue?”
Waye answered strongly. “It's, it's flawed. It doesn't follow the scientific method.”
In other words it was bad science.
Simmonds continued. “Now, when you look at those results and compare those with Mr. Grant’s analysis, there were three places he should be excluded. How do we understand what's happened here? Can you explain to us what's going on? Because if he should have been excluded, that's not apparently what Dr. Chahal says.”
Waye answered. “Well, again, in the scientific method, if you compare the two profiles, you're either going to accept your hypothesis that the DNA originated from Mr. Grant or you're going to reject your hypothesis. In three locations here, you're not amplifying up alleles that are present in Mr. Grant's, so when you look at that data, you're left with rejecting your initial hypothesis or reinvestigating your, your hypothesis, doing further experiments to try to explain why those alleles are missing.”
“…but what did Dr. Chahal do?”
“He excluded all of the data from those three loci.”
It took me awhile to understand this. Apparently they were talking about three markers that had been overlooked deliberately. Chahal and Hildebrandt had found thirteen – ten which matched Grant, three which didn’t.
Simmonds moved the flip chart to the side, still in view of the jury but not quite as prominent in the room. He stood close to the witness, still looking at the jury. “Perhaps we can go on, then, Doctor, because from our perspective, somehow in the Chahal evaluation, he, having the same findings, then goes to a theory about 250 base pairs.”
“Correct.”
“Now, is there any foundation in science for his position on 250 base pairs?” Simmonds repeated the question a few times and in different ways. “Are you aware of any theory that would allow someone to just reject everything at 250 base pairs?”
“No. If you knew for a fact that your DNA fragments fall below that level, you're not going to amplify things that are over 250. Then you wouldn't need an imposed cut-off like Chahal did, because you wouldn't get results for over 250.”
At this point, Dr. Waye began to cough and reach for his glass of water.
Simmonds turned to the Judge. “Dr. Waye left his flu bed to be here.”
We were given a break. We walked out into the hallway, all very quiet. This was new information for us. It unsettled us – mainly because we had no idea what was going on. We were all disappointed that we could not see the flip chart. And we were all very curious about these three markers.
When we came back, Simmonds picked up the question.
“I thought either there was a scientific rule or there wasn't a scientific rule. Do you have an opinion?” Simmonds asked.
“Yes. You set that rule at the beginning, not at the end. If you wanted to say at the beginning everything over 250 base pairs is unreliable, you have to actually take those tests out of your kit because once you're faced with data you have to deal with it, you can't ignore it.”
“He's ignored data, Doctor?”
“He's got data on three of those loci that he is ignoring.”
They then went back to that flip chart and began to talk about alleles.
Referring to the chart, he asked Waye, “If alleles were going to drop off, which ones would normally drop off first?”
“The largest alleles - so if you're looking at locus D7, larger alleles would be 11 and 12. So if you’re looking at what was amplified, those would be much – those would be more likely to drop out before 8.”
And off they went into a world I could neither see nor understand. It was a hum of numbers for me.
Finally, Waye came to his conclusion. “So you've got two things happening there. The minor source where you do get an allele amplifying, and the major source you get a larger allele amplifying. So again, allele dropout I would consider unlikely, given these results.”
Simmonds put the next question clearly. “So now we have three different locations with unlikely positions that are being posed by Dr. Chahal where, from our - from your perspective, he should have been excluded. What do we do with that, then? Should have been excluded in three locations; Dr. Chahal just seems to throw that evidence out. Is that scientific from your perspective, Doctor?”
“Well, based on, on how we know these, these systems behave with degraded DNA, the idea of allele dropout here sort of flies in the face of how we know they behave. So it's counter to your expectation.”
Bell was on his feet. “Objection. That’s a leading question, My Lord, in a critical area.”
But the Judge overruled the objection. “He's seeking an opinion from an expert. For the purposes of clarity, I will let Mr. Simmonds pose the question as posed.”
Simmonds conceded. “Okay, I'll put it in a different - way ….”
Simmonds moved on. “I thought either there was a scientific rule or there wasn't a scientific rule. Do you have an opinion?”
Waye responded. “Yes, you set that rule at the beginning, not at the end.”
And they went through different angles of the scientific process again and again.
Simmonds asked, “When you approached all of this material, Doctor, and reviewed it all, from your, from your perspective, is this a scientific process that you have seen here?” And they went through it again.
Finally, Bell was on his feet in frustration. “With respect, My Lord, that's been covered about seven times during the course of my learned friend’s direct.”
Simmonds hesitated. “I’ll put it in a different way.”
There was another pause as Simmonds studied his notes, then looked up.
“What is reproducibility, Dr. Waye, the concept of reproducibility?”
The minute Simmonds said that, I knew we were being reminded again of the “last shot.”
What had seemed like one of the most courageous moves by Bradbury and Lutz, our champion investigators, was now under fire.
Waye sounded like an impatient science teacher at this point as he described the importance of reproducibility. “It is another hallmark in science - one of the things that’s demanded. One needs to replicate it especially if it's a result that's unexpected or may be controversial. So you either replicate it in your lab or have someone else replicate it for sure.”
The Judge leaned over… “Just”…. then thought better of it. “Sorry, go ahead. Go ahead.”
Waye continued. “In a case like this, if you have a theory that allele dropout has occurred, one of the things your lab would need to do is replicate it.”
“But in a case like this when in fact there is none left to try and run this again, how do you feel about these results under those circumstances, Dr. Waye?”
Waye shook his head. “That's unfortunate. Again, you're stuck with an untestable hypothesis. Unless… you do a validation study?”
“What are validation studies?”
“Validation studies are when a lab tries to simulate what they will see in a forensic case, using samples that they know what the results should be. So you’ll take a sample that you know exactly what the pattern is. If everything works fine, and then you do things to that sample. You'll subject it to insults; you'll expose to sunshine, bleach; you'll, you’ll intentionally add enzymes that will degrade the DNA. And you’ll see how all of that affects how your tests work. So you try to mimic what, what's going to happen in the, in the forensic world in the real world, but you know what the answer should be if it works.”
“Do you do validation studies in your lab, Doctor?”
“On every test before it's used on patients, it gets validated.”
“According to Dr. Chahal, there were no validation studies here. How does that affect the results we've been told about?”
“A lab shouldn't be doing a test unless you do your own internal validation studies,” concluded Waye.
And then Simmonds went back to the question of scientific processes.
“From your perspective, can you give an opinion as to the scientific validity of what you're seeing, then?”
“No validity whatsoever,” Waye answered.
“Because, of course, we've now seen the data and you've looked at the data that deals with exclusion. Now would you look at the final report that he produces?”
“The final report again at those three loci - D7, D16, and D18 - takes out all the results. So it takes out 19 what you did see in the actual profile, it takes out what in the previous profile he assumed was there and had 1 dropped out, and it just leaves it a blank slate and says that that data's unreliable because it's over 250. So now he’s come up with a rule that excludes that data.”
“Again, from your perspective, is that defensible in any way as a scientist?”
“No.”
“Those are my questions. Thank you, Doctor.”
Bell stood up.
It was now time for the cross-examination.
I wondered if Bell would change into a “Defence lawyer” mode, but he didn’t. He was very precise in his language, pointed and brief.
“Well, with respect to your position on scientific hypothesis, I noted Dr. Chahal was asked that question on direct and he said that what he was really doing was a test. Would you disagree with his interpretation of what he did?”
“There's a difference between a test and an experiment. A test - if I analyze my DNA and develop my profile, that's a test. It doesn't answer a question; all it does is tell you what my profile is. And you can test my height as well; it doesn't test the hypothesis. An experiment, which is exactly what's done here, is a test that’s done under controlled conditions to test a hypothesis, to test the validity of a hypothesis.”
“Well, when you testified in direct this morning, Doctor, you - your hypothesis seemed to be that Mr. Grant fit the profile. Now, if you're commencing, would you agree or disagree with that?”
“That his ...?” Waye was stunned.
“That appeared to be your hypothesis prior to testing.”
“Parts of the data, he fits. Parts of the data, he doesn't fit.” Waye said.
Bell continued. “No, no, maybe you misunderstand my question. You told Mr. Simmonds, I believe – and correct me if I'm wrong – that your hypothesis for this example was that Mr. Grant fit the profile, and then you said that you would conduct an experiment and then if the results didn't come out the way that you anticipated, you would amend or change your hypothesis. Was that what you told Mr. Simmonds this morning?”
Waye was horrified. “Not in this lifetime. No, that's not the way I do science. You start off asking a question: Could Mr. Grant have contributed to the sample or if it’s somebody other than Mr. Grant. You do the, the test.”
Bell persisted. “I'm not suggesting for a second, Doctor that I’m trying to get you to include Mr. Grant in anything here. All I'm saying is when I took the notes with respect to what you said about hypothesis, it struck me that your hypothesis seemed to be, at the outset, and Mr. Grant fits the profile.”
“No, it's something that you're going to test. It's, it's a hypothesis.”
“Yes?” Bell was attentive.
“It may be right, it may be wrong.”
“I understand that, but during your example this morning was that not your hypothesis? I'm not saying it was - ended up being….”
“It's the only question on the table: Could Mr. Grant be a possible contributor to this?”
“Okay.”
“That's, that's different than saying my hypothesis is that Mr. Grant is a contributor to this.”
Judge Joyal said, “Okay, you're answering my question now. That was the same question on the table for Dr. Chahal as well.”
Waye answered, “Correct.”
“So it wasn't his hypothesis, it was the question on the table, correct?” said the Judge.
“Right.”
“Okay, good,” nodded the Judge.
The penny dropped. It was a test – not an experiment, and a question rather than a hypothesis. It wasn’t bad science – just a different science.
Pure science is used to develop information to predict and perhaps explain phenomena in the natural world. But applied science uses existing scientific knowledge to develop more practical applications, such as technology or inventions.
Again, I didn’t understand all of the conversations regarding the markers on the flip chart that was turned away from us, but it seemed as if the bravado exuded by Dr. Waye in the beginning was very carefully being punctured by the Crown lawyer, now a fencing expert wielding a foil so quick that it was hard to see – except every once in a while, it seemed a light would go on when the foil found its mark.
We all knew that the “last shot” by the investigators was controversial. This also included “boosting” using a stronger primer – which was also controversial.
When Bell asked Waye if the use of the primer was legitimate, Waye answered, “That's one, one thing you can do. You can load more product onto the machine which will amplify the small peaks become larger peak heights.”
“Okay….” Bell nodded.
“It's called boosting.”
“Then you'd agree with me that this isn’t simply a situation where you load the machine, press a button, and you get the results. Sometimes you have to have a look and see what you get from the first test?”
“Yes,” answered Waye.
“And you have to make some adjustments with respect to the amount of PCR product that you might introduce into the machine.”
“That's one manipulation you can do, yes,” answered Waye.
Bell was getting his answers. “Okay. So you said that if you don't put enough PCR products in, you might get a small peak, but on the second test, if you put more PCR product in, you might get a bigger peak?”
“You will get a bigger peak.”
“Okay. Well, let's say you put a whole bunch of PCR product in for the third test. What might happen there?”
“Well, as, as you increase the peaks that, that you’d like to see, you also increase your background.”
“Okay.”
Waye continued. “So if you've got a peak that's below threshold, adding twice as much material will bring it over threshold, you may, you may actually create some other peaks.”
In other words, it was possible to push a sample so far it begins to pick up unrelated data.
Bell continued. “I think it was pretty clear both today and last week that Molecular World has sort of a minimum of 100 - base pair? Would you agree that's an acceptable minimum?”
Waye did not deny it but said, “In their lab, if that’s what they decide to do, it should be something that they've established as a minimum.”
“Fair enough.”
“It varies from lab to lab,” Waye said.
“Then you said if you put too much PCR product in, it might go over the top, and where would the top number be?”
“Again, that's something that each lab sets.”
“Okay. So you'd agree with me, though, there's some fine-tuning involved in an effort to get the, the best possible data.”
“Yes.”
With one deft stroke of his foil he had just confronted Waye’s assertions that one must always have a predetermined 250 base pair limit before doing a test.
Sometimes one experiments.
Bell continued. “And sometimes - although it's not encouraged - sometimes in an effort to get the best possible data, you have to use up all the DNA you have.”
Waye nodded. “Yeah, that's the reality of this business. You're starting off with a certain amount of material at the beginning and you have only so many chances to analyze it. And when you're done, you're done.”
“When you're done, you're done,” Bell repeated.
This took care of the validation and reproducibility aspect. In a perfect world, one does take those precautions, but this wasn’t a perfect world – the DNA was old.
“Now, with respect to calculations, Dr. Waye,” continued Bell. “I understand that something commonly used is called the Hardy-Weinberg Theory.”
This theory is a crucial concept in population genetics. The Hardy-Weinberg principle states that in a large, randomly breeding population, allelic frequencies will remain the same from generation to generation assuming that there is no mutation, gene migration, selection or genetic drift. This principle is important because it gives biologists a standard from which to measure changes in allele frequency in a population.
They went into a lengthy explanation of what this would look like given the DNA findings on the chart.
“You would look at the frequency that you would expect to see those results at those places, and then you would multiply all those results together. Is that correct?” asked Bell.
“Yeah,” Waye said. “So if you use the Hardy-Weinberg Theory, you take the frequency of that in the population, so say that's ten percent have a 13 allele.”
“Okay. So what you do is you take that number, you look it up in the database.”
“Right.”
“That gives you the frequency as to how many times you expect to see that in the general population,” suggested Bell.
“Right.” Waye was into it now. “If you assume that there could be something in there that I didn't detect and you want to be very conservative, you could say that there could be a 13 and any other allele.”
“I thought we were keeping it simple.”
“Nothing's simple in genetics.”
“Yeah, I get that.” Bell smiled.
“And the way to calculate that, again using the Hardy-Weinberg equation, is you take the frequency of the 13, multiply it by one minus the 13 which gives you the frequency of all the other alleles, times two. So it’s two - 2P times one minus P times two. And then you add those two together, and that’ll give you the proportion of the population who could have contributed a 13 allele.”
“You'll get a number from those calculations.”
“Right.”
Bell continued. “And then you take the number from, say, D3, TH01, et cetera, and multiply all those numbers together.”
“That's why the technology is so powerful, is that you take things that aren't individually very discriminating, one in ten, one in eight, but the combined process, going through all of the markers, you multiply them and you end up with one in a trillion.”
One in a trillion!!! I was overwhelmed with the numbers.
“So can you tell us, then - or would you agree with me that Dr. Chahal was looking for 15 markers when he did the tests that he did?” Bell’s voice was soft.
“That's what's included in the test, yes.”
“Okay. And that comes from the kit; is that correct?”
“Correct.” Waye seemed very cooperative.
“And if you got - if he had indications on all 15 markers using the formula you described, you'd multiply all those numbers over and over and over, basically 15 times.”
“Correct.”
“So you could conceivably get a number in the trillions. That fair?” asked Bell.
“If it was a single source sample, you would get something that is - the denominator's very, very large.”
“Crazy numbers,” said Bell.
“Yes.”
“Okay. And that's with 15 markers or loci?” Bell asked.
“Yes.”
Bell stepped back. “Now, when you worked for the RCMP you said that you sort of got them on track to do DNA work back in the late 1980s. Was that with PCR technology or RFLP? What was that?”
“RFLP,” answered Waye.
“How many markers were you looking at back then?”
“Four to six.”
Bell seemed surprised. “Four to six. And I take it once the RCMP got into PCR technology, certainly for a while at least, a long time, they were looking at nine markers?”
“Yes,” nodded Waye.
“So you'd agree with me, then, that the more markers you look at, the more discriminating it is.”
“Correct.”
“So looking at nine markers is better than, say, seven markers?”
Waye nodded. “It'll give you a more discriminating profile.”
“Okay. And obviously ten markers is better than nine markers?”
Waye had a hard time admitting it, but he couldn’t deny it either. He was quiet.
Bell consulted his notes. “Dr. Waye, just one last question. If you assume for a moment that when Dr. Chahal, with respect to the nuclear gene, the STR, made the cut-off at 250 base pair, if you assume that he's correct in doing that, is it true that the known profile for Mark Edward Grant would be the major contributor to the mix profile on the twine?”
Waye’s voice was strong. “If you blindly assume that he is correct in that assumption, then your conclusion that Grant cannot be excluded is valid, and the numbers that he gets for the nuclear DNA, one in 33,000 is valid and multiplying it by the Adam gene Y-STR is valid.”
“Correct.”
“Thank you, sir. No further questions.”
It was 12:30. The direct examination and cross-examination had taken all morning.
Here is another blog – some random thoughts.
*****
A string of miracles: February 2011
Over these last three weeks we have seen the Crown very carefully lay down the outline of the story of what happened November 30, 1984.
We now know that the accused frequented a party house of drugs and drinking behind the little convenience store next to the school that Candace attended.
We know that his girlfriend came forward with lies that could have sabotaged the six-week search for Candace by lying to the police. We know they were both running from something at the time and temporarily lived in an abandoned hole in the ground near railway tracks.
We know that Grant is a convicted, registered sex offender who raped another girl under the Nairn overpass on the same path where Candace was taken by force.
Whether everything we are learning will come to light, I have no idea and it no longer matters. It is enough for me.
I consider it all a miracle, actually a string of miracles, starting with the police investigation 26 years ago that ended in this court case.
It doesn't have to be conclusive for me to be satisfied.
I don’t think any of us will ever be the same again. We have the front row seats for the most profound drama of life.
There are very few human beings who receive the truth, complete and staggering, by instant illumination. Most of them acquire it fragment by fragment, on a small scale, by successive developments, cellularly, like a laborious mosaic. Anais Nin
RSS Feed
