The Analyst In Her Own Words

Start with the premise. Ms. Taylor agreed that finding one person’s DNA on another person’s clothing does not mean a sexual assault happened. Then she described the examination she actually designed: a search of the fly for skin cells Ms. Ligons might have left there while “steadying herself.” Taylor Dep. 17:21–18:1, 18:16–25.

No one told her that happened. No one has ever described steadying herself by taking hold of the fly of a man’s trousers, and Ms. Taylor admitted she did not know “if women do that or not.” She made what she called “a really big assumption,” and then built the State’s only forensic examination around it. Taylor Dep. 18:16–25, 55:2–6.

Then look at where she swabbed. She described her target again and again as the “front of his pants.” What she sampled was the fly, the strip of fabric that holds the zipper, and she said it was the only place she meant to test: she “wasn’t going to check for DNA on his lower pants.” A woman steadying herself against a man would be more likely to touch his thighs, front pockets, knees, or hips. None was swabbed. The Velcro belt came to her threaded through the belt loops as part of the same package, logged as Item 18, and it was never swabbed either. His uniform shirt was never collected. Her hypothesis was hands on the front of the pants. She tested the one part of the front the State would later tie to a sexual act. Taylor Dep. 18:16–25, 25:16–23, 31:17–24, 33:14–18, 38:17–22, 54:19–55:2, 149:9–13; Forensic Examination Report SD-14-273 (Nov. 12, 2014), at 2, 6–9; Bench Notes, SD-14-273, at 3–4, 18; Scott Adams 04341–42 (property receipt), 04831–32, 04834 (interview transcript).

And suppose she had found Ms. Ligons’s skin cells on the fly. What would that have proved? By his own account that afternoon, Daniel searched her car with her consent: he looked under the driver’s seat, went through her purse “like she said,” checked the name on her pill bottles, and sniffed her cup of Kool-Aid for alcohol. She also sat in the back of his patrol car. Scott Adams 00036, 00079, 00081, 00084. Her cells on his clothing would have been the expected result of a traffic stop that nobody disputed took place. The test could not tell an assault from a stop. As it turned out, none of Ms. Ligons’s DNA was found on the fly.

Finally, look at what she was searching for: touch DNA, left by ordinary contact. The State’s witness endorsement, filed with the court before trial, described that kind of transfer as “not reasonable and perhaps not even possible.” Shown it, she said: “these are Gayland’s words, not mine.” The State’s analyst designed her examination around a mechanism the State’s prosecutor would call nearly impossible. State’s Endorsement of Witnesses, item 124, O.R. 240 (Ex. 11 to Taylor Dep.); Taylor Dep. 132:24–25.

Her principle has real support. Current federal guidance calls for keeping analysts away from task-irrelevant information, such as a suspect’s record or investigators’ theories, because it can bias a forensic judgment. But the same guidance says the system must also ensure that all task-relevant information reaches the analyst. What was alleged is task-relevant. For an analyst examining the fly of a man’s trousers, the difference between alleged oral contact and alleged intercourse decides what to look for: saliva, semen, vaginal fluid, skin cells from ordinary touch, or a mixture from more than one route. Those are not prejudicial details. They define the biological question.

That information never reached her. The first request, in 2014, told her “this was an oral sodomy case,” and she built her examination on it. The form requesting the interior swabs in 2015, nearly a year after Adaira Gardner and others alleged vaginal rape, still described the offense as “Forcible Oral Sodomy,” the Ligons allegation, while asking her to look for another woman’s DNA. The system meant to shield her from irrelevant context withheld the relevant context too. It is as if a hospital laboratory, properly shielded from gossip about a patient, received a requisition naming the wrong condition, chose its tests for that condition, and then had its results read as proof of another. Taylor Dep. 20:5–10; Scott Adams 03602; National Commission on Forensic Science, Ensuring That Forensic Analysis Is Based Upon Task-Relevant Information (2015); NIST IR 8503, Forensic DNA Interpretation and Human Factors (2024), § 3.2.1 and Glossary.

This matters because of the question her results were later used to answer. “Whose DNA is on this swab?” is a question about source; an analyst can answer it knowing nothing about the allegation. “How did Ms. Gardner’s DNA get onto the fly?” is a question about activity, and it cannot be answered without knowing what activity is alleged. The federal guidance notes that an investigation may begin with the first question and turn into the second. Here it did, and no one told the analyst. NIST IR 8503, Forensic DNA Interpretation and Human Factors (2024).

If Ms. Taylor believed every allegation was oral sodomy, she never designed or evaluated her examination around the possibility that Ms. Gardner’s DNA was vaginal material deposited during intercourse. Yet that is exactly what the State told the jury the DNA showed. The question is not only whether her DNA typing was right. It is that the State’s interpretation of what the DNA meant went far beyond the question her examination was built to answer.

The gap was not only the system’s. By 2015 she knew two other women had come forward and that she was swabbing the pants for their DNA; she could not later recall their names, and nothing shows she asked what they alleged. At trial she said she thought the allegations “were all oral sodomy,” and in the same testimony, asked whether transfer was “much more likely” in a liquid “such as vaginal fluid,” she called it “a very good possibility.” She had no basis for that opinion either way. Not knowing the allegation meant she never designed an examination to address it. Knowing it would have given her a hypothesis to test, not an answer, and no test for vaginal fluid was ever run; her laboratory had none. Taylor Dep. 59:15–16, 66:23–67:4; Trial Tr. 4064:4–5, 4073:20–24.

This is a basic question of forensic serology: if a man’s pants were in contact with a woman’s vaginal area for ten minutes, would you expect to see a stain? The State alleged two vaginal rapes through the same fly within a day of the pants being collected.

She paused for about seven seconds, sighed, and answered a different question, about semen and about the Ligons investigation, neither of which Ms. Zellner had mentioned. Then she summed up her own examination better than anyone else has: “looking for seminal fluid, no; looking for skin cells, yes.” At trial she had told the jury she examined the fly under bright light and magnification and found “nothing on there suspicious.” Taylor Dep. 64:20–65:8 (video 1:36:51–1:36:57); Trial Tr. 4084:2–7.

This is the center of the case, and it is worth considering carefully. The DNA on the fly was offered to the jury as independent, scientific corroboration of Ms. Gardner’s account. Asked where that DNA came from, the State’s analyst said she did not know. It “could have come from her mouth,” her skin, “her vaginal vault,” or, in her words, “her sneezing on him.” She could not say to any degree of scientific certainty that it came from vaginal fluid. Taylor Dep. 141:19–142:16.

Yet she would not dispute the prosecutor’s claim that it was vaginal fluid, and her reason was the accuser: “She knew what happened to her, I didn’t.” “Why would I dispute the woman’s testimony… because I do not have a test?” Taylor Dep. 140:17–19, 142:2–5.

Follow the logic. The State told the jury the DNA proved Ms. Gardner was telling the truth. The analyst’s basis for the most important part of that proof, that the DNA arrived in vaginal fluid, was that Ms. Gardner was telling the truth. The accusation supplied the science, and the science was then offered to prove the accusation. Forensic evidence exists to test a witness’s account. Here the witness’s account was used to interpret the evidence, and the evidence was presented to the jury as independent confirmation of the witness.

“Not suitable for comparison” means the data can’t include anyone and can’t exclude anyone. Her written reports said the minor components were not suitable for comparison purposes, due to insufficient genetic material. On the stand she said it too, and then excluded Daniel anyway. She told the jury faint results could be used to exclude but not to include. Her own reports had already found these too thin for either. Gill et al. (2017), p. 26 (quoting O.R. 182, 253).

Both exclusions fail on their own terms. The first rested on a missing number. When there’s very little of a person’s DNA in a sample, some of his types simply fail to show up. Analysts call it drop-out, and her laboratory’s own interpretation manual calls for an inconclusive result when it’s in play. A 16 that isn’t there is what drop-out looks like. The second rested on a number that wasn’t his. An 11 that doesn’t belong to Daniel shows that someone else contributed DNA. It says nothing about whether he did too. The 2017 panel that included Peter Gill found that Daniel couldn’t be excluded on her data: most of his types (22 of 32 on that flap swab, 19 and 21 of 32 on the two inside swabs) were present in the mixtures, many of them at low levels. Gill et al. (2017), pp. 23–26.

Set her answers side by side. One location where Daniel’s type appeared was not enough to include him. One location where his type didn’t appear was enough to exclude him. That exclusion is what the prosecutor built his next questions on.

By 2019 the record contradicted her trial testimony on several points: her report did not permit the exclusion she gave, her own data measured the male DNA she said was not there, and the vaginal-fluid opinion rested on no test. Given the chance to revisit any of it, she refused. “I will not change a single word.” Asked whether she would correct a mistake: “I don’t believe I made any mistakes.” Shown the prosecutor’s summary of her testimony, which she had just disowned: “if there’s something wrong, so be it.” Trial Tr. 4058:18–4059:10, 4071:11–13, 4072:19–25; Supp. Forensic Exam. Rep. SD-14-273 (Oct. 7, 2015); Ex. 15 at 15-058, 15-060; Taylor Dep. 96:4–6, 124:21–22, 132:24–25, 135:4–9.

The male DNA is the plainest case. Earlier in this same deposition, asked whether the laboratory had measured male DNA on the two swabs from inside the fly, she read out the figures herself: it was quanted at .0102 and .0117, a male-to-female ratio of 1 to 20 and 1 to 21. Forty-four pages later, asked whether male DNA was found on the fly swabs, she answered Nope, and said the two inside swabs only show to be female DNA. A measurable quantity of male DNA is not a profile anyone can compare, and the laboratory never obtained one. But it is male DNA, and at trial she had told the jury there was none. Taylor Dep. 99:20–25, 143:22–144:4; Trial Tr. 4072:14–25.

Science corrects itself when the data require it. She treated her testimony as a position to defend, and said so on video.

Her supervisor was in the courtroom. Mr. Ruddock watched one of her two days on the stand, and before trial he had given her a review article on contact DNA. His word for her testimony, again and again, was “confused.” That is the testimony that gave the jury the exclusion and the missing male DNA, and neither was ever corrected. He said analysts should “restrict ourselves to DNA unless we have data.” What he did at the time was talk with her back at the lab, with no written review. He described that talk as general advice about contact DNA, not about her testimony in this case, and could not recall raising the lubrication testimony with her. Nothing in his deposition indicates he told the prosecutor or the court. Ruddock Dep. 15:6–15, 19:8–10, 20:14–15, 26:21–23, 28:7–10, 28:22–29:6, 31:15–17, 41:13–20, 42:12–14.