May-June 2007

Entertainment or expectation? How ‘CSI’ affects today’s juries

John Ernest Boundy

Assistant District Attorney in Nacogdoches County

My life is a devilish dichotomy. At night I join millions of Americans in the delicious pursuit of crime-solving with “CSI” and its successful siblings in Miami and New York. But by day, I am a Texas prosecutor who doesn’t have Horatio Caine or Gil Grissom to woo a jury. Oh, to make a case in 45 minutes, without commercial break, with such a stunning array of technology and science as to leave the jury with no doubt of a defendant’s guilt!

We all know real criminal trials don’t work that way. I recently had a case that ended in mistrial because of jurors’ demand for fancy science and indisputable evidence, even though the law required neither. I hope this article helps other investigators and prosecutors with future cases.

The case

It was Friday, August 4, 2006, and it had the makings of a great day. Twenty-six-year-old Julia Thomas was celebrating her fifth wedding anniversary. Her aunt, uncle, and cousin had driven three hours from Louisiana to attend the party planned for that night. The rest of the weekend would be spent visiting family in the area.

Her uncle, Kenneth DeSelle, a truck driver, dropped in that morning as he sometimes did on his way back to Louisiana from a haul. A little past noon, the others left to go visiting. DeSelle, after taking a shower, went to his truck and spent 20 to 30 minutes in the cab. (Though he was never tested and it can never be proven, I believe he was likely using methamphetamine; Julia says his demeanor changed completely after he returned from the truck, which is consistent with meth use.)

Around 12:30 p.m., Julia was in the kitchen making lunch for her 3-year-old daughter, Susan, and younger son. She heard a sound, and as she turned around, DeSelle charged her from the living room and demanded sex. He shoved Julia against the wall, and the two struggled. As she fought, DeSelle grabbed Julia around the throat and threatened to kill her and the kids if she didn’t stop screaming.

DeSelle then stuffed a baggie of white powder into her mouth and again demanded sex. He continued to threaten to kill if she did not comply.

Spitting powder, Julia sobbed, begging for an explanation. “Why are you doing this? You’re my uncle. That’s gross!” DeSelle responded by ripping the buttons off her pants, forcing them down. Julia started screaming again. Hearing the commotion, young Susan ran in. Crying and screaming, she began kicking and slapping DeSelle. “Please don’t kill my momma. Please don’t hurt her!” DeSelle elbowed Susan, knocking her backwards several feet into a cabinet. Turning back to her mother, he forced his right fist into Julia’s genital area before she found the strength to push him off.

Julia rounded up Susan and began throwing pots, dishes, and anything she could get her hands on. She screamed she’d call the law. As he left, DeSelle said, “Go ahead. The cops don’t scare me.”

Hysterical, Julia called Rachel Gray, her friend and cousin-in-law. Rachel in turn called 911 and immediately drove to Julia’s house, as did her other family members after hearing about the incident.

Patrol Deputy Danny Kitchens responded to the 911 call. He saw the family crowded around Julia on the front porch as he arrived. Separating her from the rest, Kitchens tried to interview Julia, but she was hysterical. He called for investigators to respond to the scene.

EMS arrived. Julia refused to be touched by the male paramedic but agreed to be transported to the hospital only if a family member could ride along. Sheriff investigators Bill Ball and Larry Murphy arrived and photographed the home and those present. They returned to the sheriff’s office and contacted DeSelle’s employer. Using on-board GPS, they were able to track DeSelle’s truck and his whereabouts. After securing a warrant, DeSelle was arrested later that day in Louisiana.

Meanwhile, Deputy Kitchens went to the emergency room and again tried to interview Julia. While calmer, she still couldn’t relate the full events of the assault. They agreed Julia would come to the sheriff’s office the next day and provide a written statement, which she did.

Sandra Williams, a sexual assault nurse examiner (SANE), collected evidence and performed her examination. The evidence was sent to the lab for analysis. Because of the nature of the assault, no forensic evidence (i.e., semen) was present. Nothing scientific would help in prosecuting the assault.

When this case first hit my desk, it seemed, sadly, routine. No red flags went up for a need of additional forensic evidence or evaluation. I had a credible and very believable victim. The SANE nurse was competent and confident in the facts. The deputy was strong and would make a good witness. Finally, there was adorable and precocious Susan, the 3-year-old heroine who tried to help her momma during a brutal assault.

The trial

Julia’s testimony was as expected. She walked through a horrifying account of the attack. Nurse Williams ran through the exam and what she did as a sexual assault nurse examiner. Then Deputy Kitchens took the jury over his involvement from initial response to obtaining Julia’s statement the next day. Out of court, young Susan was playful and talkative; however, the thought of talking to 12 strangers made her withdraw. Her grandmother took the stand instead to explain my efforts to get Susan to testify, then brought Susan in and introduced her to the jury.

Because the sexual assault examination didn’t produce any usable evidence, no criminologist was called. Both investigators were tied up on other cases at the time of trial. Because their only actual involvement had been taking photos, they were not called either. And because the predicate for a photo is satisfied by any witness who has personal knowledge of the scene and recognizes the picture as an accurate depiction of the scene or event it purports to portray,1 the pictures of the scene and Susan’s injuries were introduced through her mother, Julia.

The defense presented a two-pronged attack. First, because Julia had told everyone from the paramedic to the police that the defendant had “shoved his whole fist in me,” defense counsel contested the physical findings as inconsistent: There were no vaginal tears or lacerations. Secondly, meth was found in Julia’s urine, and she had “lesions” on her face and body, so the defense alleged she must be a chronic meth user and was lying about the assault.

Because she was living in Louisiana after the assault, I talked with Julia many times by phone in the weeks leading up to trial. While admitting meth use after the assault, she adamantly denied it in the period beforehand. As for the “lesions” documented in the medical records, she explained she had acne and reactions to Clearasil. In my pretrial interviews with the SANE nurse, she confirmed Julia’s account of her acne and said she saw nothing consistent with meth use.

On cross-exam of Julia, the defense alluded to an alleged argument with the defendant the morning of the attack about drug paraphernalia she supposedly left on the kitchen table. Julia’s response was not only that there was no argument, but also that there was no kitchen table. She pointed to a picture of the kitchen that had been introduced showing no table present. That line of questioning was quickly abandoned. When defense counsel attempted to question Julia regarding meth use prior to the assault, the court sustained our objection on grounds of relevance. The defense then went through the EMS and medical records introduced by agreement, focusing on Julia’s recitation of the incident that the defendant had inserted his entire fist into her. Julia re-iterated her explanation from direct testimony that the defendant had rammed his fist into her genital area and that the intensity of the pain felt like the defendant’s entire fist was inside her.

With the SANE nurse, the defense pointed to the emergency room doctor’s reporting of “lesions” on Julia’s face and back. Inquiry about the nurse’s experience with “crank bugs” was allowed by the court. Then the defense meticulously went through the physical examination and findings. Counsel concentrated on the lack of vaginal tearing or laceration. On re-direct, Nurse Williams testified there were no “crank bugs” and that the “lesions” were acne. Regarding the physical findings, Nurse Williams re-affirmed her direct testimony that the bruising of the victim’s inner thighs she saw and documented, as well as a lack of tearing or laceration, were consistent with the assault history. Although I had interviewed her twice prior to trial, on cross-examination, I learned for the first time Nurse Williams had taken photographs of Julia. She had not told law enforcement or me of their existence. They were safe—and useless—locked away in a cabinet at the hospital.

For their case, the defense called three witnesses: Dr. Brown, the emergency room doctor; Rachel Gray, the cousin-in-law Julia called after the assault; and Rhonda Lowe, Julia’s sister-in-law. Through Dr. Brown, defense counsel established the presence of meth in Julia’s urine and “lesions” on her face and back. On cross, the doctor was happy to state he had examined Julia prior to taking the stand and the “lesions” on her face at trial were acne similar to what he had seen the day of the assault. Further, the presence of meth in Julia’s urine was consistent with the history of having ingested white powder believed to be meth during the commission of the assault (when her uncle shoved a plastic bag of white powder into Julia’s mouth to muffle her screams). Finally, the lack of vaginal tearing or laceration was also consistent with the stated history.

Witnesses Gray and Lowe were presented under TRE 608 to provide testimony of Julia’s reputation of truthfulness. Both stated they were familiar with her reputation of being untruthful. On cross, Ms. Gray admitted that in response to Julia’s call, she called 911 and frantically drove to Julia’s home. Ms. Lowe conceded that her brother, Michael, is going through a rocky divorce with Julia. She denied threatening, just the night before, to take Julia’s children in violation of a court order to prevent Julia from having custody.

For the injury to a child charge, Julia testified 3-year-old Susan cried after being hit and complained of pain in her right shoulder where she hit the kitchen cabinet. Pictures taken on the front porch following the assault showed redness and slight swelling to the child’s shoulder. The defense attacked that Susan had not been taken to the emergency room that day and that Julia did not have any medical record or bills for a subsequent doctor’s visit she testified taking Susan to days after the assault.

After closings, the jury began deliberations. The judge requested I begin working on a punishment charge because he felt a quick guilty verdict was coming.

Two hours later, the first note came out requesting information on when and where the defendant had been arrested. There was an immediate sinking feeling in the pit of my stomach. The jury was looking beyond the evidence and conducting an independent investigation. I could read their thoughts: They were looking for trace evidence. They had heard from Julia that she struck and tried to scratch the defendant. Then the deputy told them the defendant was arrested in Louisiana later the day of the assault. I knew they wanted to know if he had been forensically examined at the time of his arrest.

My fears were realized hours later when the coup-de-gras note entitled “Discrepancies that prevent a decision” came out. As I read it over, my head throbbed.

I didn’t have the pictures taken by the SANE nurse, a major source of irritation and self-flagellation. But Nurse Williams had testified to and the records documented Julia’s injuries. As far as police reports, I couldn’t explain that those are inadmissible hearsay.2 The jury had heard from the victim and had seen photos of the scene and the child victim’s injury. They listened to Nurse Williams correlate her findings with the assault. Yet they disregarded the court’s instruction to consider only the evidence before it. Instead, jurors conducted their own investigation in the jury room and decided what evidence would be necessary to reach a verdict. They even disregarded the defense’s theories in lieu of what they expected to see—save for the injury to a child charge regarding young Susan. There, the jury bit on the defense notion that there was no injury—even though the charge had standard language that bodily injury includes pain, illness, or any impairment. Although they had the mother’s testimony of Susan’s complaint of pain and the photographs, the jury wanted more proof.


The jury hung, and a mistrial was declared. Because of the logistics involved with a new trial, we reached a plea bargain with DeSelle. The defendant pled guilty and accepted a 10-year deferred probation and lifelong sex offender registration.


We all have faced the “‘CSI’ effect,” which I now refer to as “techiness” in honor of Stephen Colbert. Because of “CSI” and similar shows, the public has a misguided notion that police can go to a crime scene, pick up a single piece of evidence, feed it into a computer, and in a matter of seconds, know who the bad guy is. While television creates the illusion of unequivocal black-and-white crime solving, it just doesn’t work that way in the real world. The public—our jury pool—has been fascinated by “CSI” and its successful spin-offs, but the backlash is that they have crossed over from entertainment to unrealistic juror expectations. That fictional combination of smooth police work and science, techiness, can sink a case if it’s not neutralized early.

To combat techiness, I start at the outset. During my investigation, I look at what forensic testing, if any, was done. Then I put on a “defense hat” to anticipate the cross of an investigator:  What more could have been done regarding forensic testing and the general investigation? (Some inevitable questions from the defense include: Did you test for DNA? How about running the evidence through the mass spectrometer?) Obviously, we can’t anticipate everything, nor should we even try to run down every rabbit trail, but “case cleared” doesn’t mean we stop investigating. Because of techiness, remember that juries have an expectation of hard forensic data and are less likely to convict without it. The more we can think of what evidence they blissfully believe we should have, the better armed we are to give them a reality check.

In voir dire, I now ask in every case, “How many people watch ‘CSI’ or related shows?” I explain to the jury pool that there is no “CSI: Nacogdoches” and that Horatio and Gil are not on my witness list. I confirm their understanding of the difference between television fiction and real world investigation. It’s updating the old “Is one witness enough?” question for the techiness age.

Point out the illogic in some of jurors’ expectations for evidence. If you’re trying a burglary case, for example, ask about the number of burglaries in a year’s time in your jurisdiction. How many investigators are there to work that number of cases? What does the standard investigation consist of? What kind of time does a real investigation like that take? Is it realistic to dust every window on a building even if there is no indication the suspect was near that part of the structure?

In that vein, I also talk with my investigators before they take the stand. I explain what the jury is anticipating because of television fiction. I prep them that I’m going to ask, “Did you do this test or that test? Why or why not? Do we have the equipment to test for that? Does Dallas? Or Houston? Does that equipment even exist?” and so on.

If I have a criminologist, instead of just asking about the specific scientific method of whatever examination he performed, I ask more, though how much more depends on the case. I have him explain how many different instruments are used in various types of analysis. I also ask about the cost of such testing and who pays for it. Another reality check is how long it takes to get a result for various tests and how many submissions the lab generally receives in a given period of time. I ask whether techiness has had an effect on labs and if they are inundated with requests and demands to perform faster. Jurors are amazed that instead of the instant gratification of television, patient prosecutors wait months for results on drugs and forensic analysis. More amazing still is that there isn’t a single information database that can provide all the answers as on television.

And though I am generally loathe to introduce uncertainty about the system, jurors need to know that while TV tests are instantaneous and accurate beyond contestation, real-life testing is done by humans. There is always a chance of error in performing the tests; analyses are done by highly trained people, but people nonetheless. As such, their results are often open to interpretation.

The “sorry man”

In a hotly contested case where I anticipate forensics to be a centerpiece of the defense, I now consider calling a negative expert. It’s akin to a “sorry man” I employed when I was a kid; this “sorry man” (my best friend) would accompany me to tell my mother when I had done something wrong, the idea being: 1) he could help explain why I had done what I’d done, or in the alternative, 2) my punishment would not be as bad in front of a witness.

During a trial, someone with the expertise to explain why there is no forensic evidence in a particular area acts as a “sorry man” for the jury and assures them that nothing is out of the ordinary in the case before them. Unlike on television, real-life criminals aren’t always as accommodating in leaving behind a smoking-gun epithelial cell.

Using as much demonstrative evidence as you can muster—pictures, diagrams, models, summaries, charts, and event chronologies—can only bolster your case. Anything that brings a jury closer to the trial of their expectations is helpful.

To close things out in final argument, I rely on the standard charge language our jurisdiction has:  “During your deliberations in this case, you must not consider, discuss, nor relate any matters not in evidence before you …” It’s pretty standard and only one sentence long, but it’s what I hang my hat on to pound home the point: Rely on the evidence before you, not what techiness you think should be there.

For those of you who want more science, technology, and law than any reasonable person should ever need, it’s just a mouse click away at, the home of the National Clearinghouse of Science Technology and Law at Stetson University. The site is funded by the National Institute of Justice and provides a searchable database of legal, forensic, and technology resources and a reference collection of law, science, and technology material. A list of available topics is below.

NCSTL topics

•    Arson/fire debris*
•    Biometrics (body scans, retinal scans,  facial recognition)
•    Bioterrorism*
•    Bloodstain pattern analysis
•    Crime laboratories
•    Cyber crime*
•    Digital evidence
•    Digital image enhancement
•    DNA analysis
•    Entomology
•    Expert witness malpractice
•    Explosives*
•    Fingerprints
•    Firearms
•    Forensic accounting
•    Forensic engineering*
•    Forensic linguistics
•    Forensic nursing*
•    Forensic odontology (bite marks)
•    Forensic pathology
•    Forensic psychology
•    Law enforcement technology (communications and interoperability, vehicles and personal equipment, computer software and hardware)*
•    Locating, selecting, and evaluating experts
•    Neuro-psychology
•    Questioned documents
•    Smart Cards
•    Thermal imaging
•    Toolmarks
•    Toxicology
•    Trace evidence (hair analysis, fiber evidence, glass, paint)
•    Voice analysis

* Topic under research and soon to be added to the database


While juries have become increasingly demanding and expectant due to techiness, that does not make our burden insurmountable. With a little preparation and education, we can help jurors come to a just and right decision based on the evidence before them. Or as a crime fighter of yesteryear would say, “Just the facts ma’am.” Thanks, Joe.


1 Drone v. State, 906 S.W.2d 608 (Tex.App.—Austin, 1995 pet. ref’d).
2 See TRE R803(8); Cole v. State, 839 S.W.2d 798 (Tex.Crim.App. 2000)(op. on reh’g).