In the Court of Appeals Second Appellate District of Texas at Fort Worth ___________________________
No. 02-19-00395-CR ___________________________
DAKOTA ZACHARY WRIGHT, Appellant
V.
THE STATE OF TEXAS
On Appeal from the 372nd District Court Tarrant County, Texas Trial Court No. 1504386D
Before Birdwell, Womack, and Wallach, JJ. Opinion by Justice Birdwell OPINION
On July 3, 2017, Jose Angel Soria was sitting in his car outside his house in east
Fort Worth when Appellant Dakota Zachary Wright approached him and drew a gun.
Wright demanded Soria’s wallet and phone, and Soria complied. For this, Wright was
convicted of aggravated robbery and sentenced to fifty years’ confinement. See Tex.
Penal Code Ann. § 29.03(a)(2).
During trial, the State sought to introduce data that was extracted from Wright’s
phone, including text messages to loved ones in which Wright expressed his belief that
law enforcement was closing in on him and photos that he took of himself and his
firearms. On appeal, Wright challenges the admission of these text messages and
photos.
As a means to that end, Wright contests the reliability of the expert testimony
through which the State sponsored the messages and photos. Wright asserts that the
State failed to establish the reliability of the technique that the State’s expert, Detective
Rex Kiser, used to extract the data from Wright’s phone: forensic extraction using a
computer program called Cellebrite. At trial, Detective Kiser explained that to extract
the data, he simply connected the phone to his computer and initiated the Cellebrite
program, which copied the data to the detective’s memory stick and made it readable
as text messages, call logs, images, etc. But Wright contends that pursuant to Kelly v.
State, the State was required to establish the reliability of the scientific theories
underlying Detective Kiser’s extraction technique before the messages and photos
2 could be introduced. 824 S.W.2d 568, 573 (Tex. Crim. App. 1992). Wright contends
that the State’s reliability predicate for this extracted evidence was fatally inadequate,
and the trial court therefore reversibly erred in admitting it.
As we explain, we disagree with Wright’s central premise: for such a basic
technique, the State was not required to establish a Kelly-style reliability predicate through
expert testimony. It is therefore immaterial whether the State’s effort to lay such a
predicate was deficient. We overrule Wright’s sole issue and affirm.
We review a trial court’s decision to admit evidence for an abuse of discretion.
Colone v. State, 573 S.W.3d 249, 263–64 (Tex. Crim. App. 2019). Under this standard,
the trial court’s decision to admit evidence will be upheld as long as it was within the
zone of reasonable disagreement. Beham v. State, 559 S.W.3d 474, 478 (Tex. Crim. App.
2018). If the trial court’s ruling on admissibility is correct under any theory of law, the
trial court’s decision should not be disturbed, even if the trial court gives the wrong
reason for its ruling. Johnson v. State, 490 S.W.3d 895, 908 (Tex. Crim. App. 2016).
An expert witness may offer an opinion if he is qualified to do so by his
knowledge, skill, experience, training, or education and if scientific, technical, or other
specialized knowledge will assist the trier of fact in understanding the evidence or
determining a fact in issue. Rhomer v. State, 569 S.W.3d 664, 669 (Tex. Crim. App. 2019)
(citing Tex. R. Evid. 702). “Witnesses who are not experts may testify about their
opinions or inferences when those opinions or inferences are rationally based on the
3 perception of the witnesses and helpful to a clear understanding of the witnesses’
testimony or the determination of a fact in issue.” Id. (citing Tex. R. Evid. 701).
“[A]s a general rule, observations which do not require significant expertise to
interpret and which are not based on a scientific theory can be admitted as lay opinions
if the requirements of Rule 701 are met.” Osbourn v. State, 92 S.W.3d 531, 537 (Tex.
Crim. App. 2002). “[E]ven events not normally encountered by most people in
everyday life do not necessarily require the testimony of an expert.” Id. “The personal
experience and knowledge of a lay witness may establish that he or she is capable,
without qualification as an expert, of expressing an opinion on a subject outside the
realm of common knowledge.” Id.; see id. at 536 (collecting cases).
“Because of its nature, scientific evidence will almost always be offered through
the testimony of experts.” Everitt v. State, 407 S.W.3d 259, 263 (Tex. Crim. App. 2013)
(cleaned up) (quoting Kelly, 824 S.W.2d at 572). Under Rule 702, the proponent of
scientific evidence must show by clear and convincing proof that the proffered evidence
is sufficiently reliable and relevant to assist the jury in accurately understanding other
evidence or in determining a fact in issue. Wells v. State, 611 S.W.3d 396, 426 (Tex.
Crim. App. 2020). We evaluate reliability under one of two tests, depending on whether
it involves “hard” science or “soft” science. Rhomer, 569 S.W.3d at 671. “When an
expert’s testimony is based on a hard science involving precise calculations and the
scientific method, the expert must satisfy the test set forth in Kelly, 824 S.W.2d at 573.”
Rhomer, 569 S.W.3d at 671. “The Kelly test for reliability of evidence derived from a
4 scientific theory requires that: (1) the underlying scientific theory must be valid, (2) the
technique applying the theory must be valid, and (3) the technique must have been
properly applied on the occasion in question.” Id. To aid its determination of reliability
under Kelly, the trial court may refer to seven non-exclusive factors: (1) the extent to
which the underlying scientific theory and technique are accepted as valid by the
relevant scientific community; (2) the qualifications of any expert testifying; (3) the
existence of literature supporting or rejecting the underlying scientific theory and
technique; (4) the potential rate of error of the technique; (5) the availability of other
experts to test and to evaluate the technique; (6) the clarity with which the underlying
scientific theory and technique can be explained to the court; and (7) the experience
and skill of any person who applied the technique on the occasion in question. Kelly,
824 S.W.2d at 573.
For expert testimony in fields of study outside the hard sciences, i.e., soft science,
we apply the framework set forth in Nenno. Rhomer, 569 S.W.3d at 671 (citing Nenno v.
State, 970 S.W.2d 549, 561 (Tex. Crim. App. 1998), overruled on other grounds by State v.
Terrazas, 4 S.W.3d 720 (Tex. Crim. App. 1999)). Soft sciences consist of areas
“involving technical or other specialized knowledge.”1 Nenno, 970 S.W.2d at 560. The
Nenno test asks whether (1) the field of expertise is a legitimate one, (2) the subject
Examples include “the social sciences of psychology, sociology, or 1
criminology.” Stephenson v. State, 226 S.W.3d 622, 625 n.2 (Tex. App.—Amarillo 2007, no pet.).
5 matter of the expert’s testimony is within the scope of that field, and (3) the expert’s
testimony properly relies upon and/or utilizes the principles involved in the field. Id.
at 561.
One Texas court has held that the admissibility of expert testimony concerning
information copied from a defendant’s cell phone should be governed by the Nenno test
for soft sciences because the exercise of copying the data “is more technical than
scientific.” Rand v. State, No. 14-16-00409-CR, 2017 WL 4273177, at *8 (Tex. App.—
Houston [14th Dist.] Sept. 26, 2017, pet. ref’d) (mem. op., not designated for
publication). Other courts have analyzed the reliability of data-extraction testimony
using the Kelly test. See Krause v. State, 243 S.W.3d 95, 108–10 (Tex. App.—Houston
[1st Dist.] 2007, pet. ref’d); Sanders v. State, 191 S.W.3d 272, 277–78 (Tex. App.—Waco
2006, pet. ref’d); Williford v. State, 127 S.W.3d 309, 312–13 (Tex. App.—Eastland 2004,
pet. ref’d).
In this case, the parties instead elected to litigate the reliability of the technique
pursuant to the test applied in federal courts under Daubert v. Merrell Dow Pharmaceuticals,
Inc., 509 U.S. 579, 591–94, 113 S. Ct. 2786, 2796–97 (1993). In Daubert, the Supreme
Court enumerated four factors to aid trial judges in determining whether scientific
evidence is reliable under federal Rule 702, and those factors overlap with the seven
factors outlined in Kelly: (1) whether a theory or technique can be and has been tested;
(2) whether the theory or technique has been subjected to peer review and publication;
(3) the technique’s known or potential rate of error; and (4) the general acceptance of
6 the theory or technique by the relevant scientific community. E.I. du Pont de Nemours
& Co. v. Robinson, 923 S.W.2d 549, 555 (Tex. 1995) (summarizing Daubert); see Hartman
v. State, 946 S.W.2d 60, 62 (Tex. Crim. App. 1997) (describing the Daubert standard as
“virtually identical” to the Kelly standard).
At the hearing to determine the admissibility of the data that Detective Kiser
extracted from Wright’s cell phone, Wright took Detective Kiser on voir dire and asked
him questions concerning Daubert’s factors for reliability. To these questions, Detective
Kiser gave minimal answers: Wright asked whether his technique had been tested
(“Yes, sir.”), whether it had been subject to peer review and publication (“Peer review,
yes. Publication, I’m not sure.”), whether there were standards controlling its operation
(“Yes, sir.”), whether there was a potential rate of error (“I would not have those
numbers.”), and whether it was accepted in the relevant community (“Yes, sir, it has.”).
When Wright concluded his voir dire, the State did not put on any further evidence to
address the Kelly factors or to otherwise shore up this testimony.
On appeal, Wright contends that the State failed to establish the reliability of the
scientific principles underlying Detective Kiser’s technique under the first prong of
Kelly.
If Kelly governed, then there could be some merit to this contention, because
there is no evidence pertaining to the Kelly factors beyond Detective Kiser’s bare
conclusions. “An expert’s simple ipse dixit is insufficient to establish a matter; rather,
the expert must explain the basis of his statements to link his conclusions to the facts.”
7 Coble v. State, 330 S.W.3d 253, 277 n.62 (Tex. Crim. App. 2010) (quoting Earle v. Ratliff,
998 S.W.2d 882, 890 (Tex. 1999)). “Without more than credentials and a subjective
opinion, an expert’s testimony that ‘it is so’ is not admissible.” Vela v. State, 209 S.W.3d
128, 134 (Tex. Crim. App. 2006) (quoting Viterbo v. Dow Chem. Co., 826 F.2d 420, 424
(5th Cir. 1987)).
We need not decide whether the State laid an adequate reliability predicate under
Kelly, though, because we are convinced that Kelly does not govern here. Neither do
Nenno or Daubert.
Rather, this court’s reasoning in Washington v. State suggests that Detective Kiser’s
evidence qualifies as lay testimony of the sort that does not require a reliability predicate
at all. See No. 02-13-00526-CR, 2015 WL 505172, at *2–5 (Tex. App.—Fort Worth
Feb. 5, 2015, pet. ref’d) (mem. op., not designated for publication). In Washington, a
detective used “UFED Cellebrite” to extract data from a cell phone and copy it to
another location. Id. at *2. The detective explained that the process was simple: he
connected a cable to the phone, selected the model of phone in the software, and
initiated the process. Id. He further explained that he knew the process was successful
because the program confirmed it and because the program copied a report of the
extracted data onto a memory stick. Id. The detective admitted that he did not know
the “technological details” of how the program worked, whether the program had been
tested for accuracy, or even what the letters “UFED” stood for. Id. Nonetheless, the
trial court admitted the exhibit over a reliability objection because there were other
8 witnesses to corroborate that the extracted text messages truly and accurately
represented the content of their text exchanges with the defendant. Id. at *3.
On appeal, we affirmed the trial court’s ruling. Id. at *4. We reasoned that unlike
other technologies, a simple data transfer was “amenable to precise independent
corroboration or confirmation by untrained lay witnesses [and] proper application by
an untrained judge or jury without supporting expert testimony.” Id. And because the
witnesses’ corroboration “empirically showed that the extractions accurately copied and
displayed text-message exchanges,” we held that no expert testimony was required to
demonstrate whatever reliability was necessary. Id.
Of course, this reasoning effectively skipped the reliability inquiry altogether,
especially the prong that Wright challenges here: whether the underlying scientific
theory is valid. We did not give a second thought to whether the principles of
electrodynamics and computer science were worthy of credence. See Morris v. State, 361
S.W.3d 649, 655 n.28 (Tex. Crim. App. 2011) (citing Hernandez v. State, 116 S.W.3d 26,
34–35, 37 (Tex. Crim. App. 2003) (Keller, P.J., concurring)) (acknowledging that
“matters of common knowledge can be recognized without a prior determination of
reliability”).
Our holding suggests that no Kelly-style reliability predicate is required for this
specific process. One purpose served by the reliability requirement is to ensure that
complex or unfamiliar systems and ideas—those that “require significant expertise to
interpret” or that are “based on scientific theory”—do not lead the lay fact-finder astray.
9 See Osbourn, 92 S.W.3d at 537. Rather than advancing those aims, the drift of our
reasoning in Washington was nearly the opposite: that this sort of data-transfer evidence
was so simple and familiar to any computer user, so plainly verifiable and free of
scientific variability, that a lay witness’s corroboration of the results based on personal
knowledge would sufficiently guard against misleading the fact-finder, even without an
expert or a reliability predicate. And a lay witness’s corroboration that a process yielded
accurate results usually serves the goal of authentication, not reliability; Rule 901 provides
that evidence about a process or system may be authenticated by “describing a process
or system and showing that it produces an accurate result.” See Tex. R. Evid. 901(b)(9).
Thus, our opinion in Washington suggests that so long as the data extracted from the
cellphone is properly authenticated by lay testimony, no reliability predicate or expert
testimony is required.
In this case, Detective Kiser offered that lay authentication. He employed the
same software as in Washington, Cellebrite. He explained that the extraction process
consisted entirely of plugging a cable into the phone and starting the extraction program
on his computer. See Rand, 2017 WL 4273177, at *8 (emphasizing the “simplicity” of
data extraction). His testimony before the jury rested almost solely on his personal
experience. To the extent that his testimony featured any opinions at all, they were
based upon his perceptions and his training in using the software, not on anything that
would require significant expertise or scientific acumen. See Davis v. State, 313 S.W.3d
317, 349 (Tex. Crim. App. 2010) (holding that a detective did not cross into the territory
10 of experts because the detective “testified based upon his first-hand observation of the
wounds themselves. His observations did not require significant expertise to interpret
and were not based on scientific theory.”). No “specialized knowledge” was required
to see that the extraction produced intact photos of Wright and text messages addressed
to Wright’s loved ones (as opposed to the corrupted files, error messages, and
unreadable mess that would have likely resulted from a failed or tainted transfer). 2 See
Osbourn, 92 S.W.3d at 537. And Detective Kiser properly authenticated all the text
messages and photos that the State presented. Thus, we believe that this case calls for
a similar result as Washington: it is enough that the extracted data was authenticated by
lay testimony, even without a reliability predicate properly laid by an expert. 3
Our conclusion finds support in several analogous federal cases.4 In United States
v. Marsh, the Second Circuit held that a federal agent’s testimony concerning Cellebrite
data extraction did not cross the line into expert testimony, for which a reliability
2 Wright was free to put on evidence that the transfer or the resulting data were in fact tainted, manipulated, or fabricated. 3 See Harrell v. State, No. 10-16-00342-CR, 2019 WL 2385682, at *5–6 (Tex. App.—Waco June 5, 2019, no pet.) (mem. op., not designated for publication) (assuming that the State failed to properly qualify its witness as an expert, but nonetheless upholding the ruling because the same testimony could be brought in as lay opinion testimony). 4 Cf. Somers v. State, 368 S.W.3d 528, 536–37 (Tex. Crim. App. 2012) (“When evaluating a trial judge’s gatekeeping decision, appellate courts may take judicial notice of other appellate opinions concerning a specific . . . technique.”); Morris, 361 S.W.3d at 659 (relying on federal decisions to reaffirm a conclusion concerning reliability).
11 predicate would be required, because the agent “did not purport to render an opinion
based on the application of specialized knowledge to a particular set of facts; nor did
his testimony turn on or require a technical understanding of the programming or
internal mechanics of the technology.” 568 F. App’x 15, 17 (2d Cir. 2014) (order). Just
as we did in Washington, the Second Circuit emphasized that the agent simply “described
how he used Cellebrite to” extract the data and “confirmed the results by checking the
messages on the phone.” Id. Applying similar thinking, the Ninth Circuit excused a
detective’s testimony concerning Cellebrite data extraction from Rule 702’s expert-
testimony requirements. United States v. McLeod, 755 F. App’x 670, 673 (9th Cir.) (mem.
op. on reh’g), cert. denied, 139 S. Ct. 1641 (2019); see United States v. Seugasala, 702 F. App’x
572, 575 (9th Cir. 2017) (mem. op.) (“The officers who followed the software prompts
from Cellebrite . . . to obtain data from electronic devices did not present testimony
that was based on technical or specialized knowledge that would require expert
testimony.”). In the First Circuit’s treatment of the same issue, the court concluded
that an officer’s testimony concerning cellphone data extraction better resembled a lay
opinion than an expert one, and thus the court was not required to “vet the opinion to
ensure it’s ‘reliable’”:
These days, most anyone with a cellphone knows they store information about text messages, including the sender, recipient, and content. You don’t need to be a software engineer to pick up a cellphone, open a messaging application, and interpret the words in the bubbles as messages sent and received. In doing so, ordinary people rely on a “process of reasoning familiar in everyday life,” not any expert knowledge about software coding or cellphone circuitry. If Officer Pérez had opened
12 Montijo’s phone and taken screenshots of his conversations with Meléndez, no one suggests she’d need any “scientific, technical, or specialized knowledge” to identify them as text messages. United States v. Montijo-Maysonet, 974 F.3d 34, 47 (1st Cir. 2020). Likewise, according to
the Fourth Circuit, a Department of Homeland Security intern did not render an expert
opinion, such that a reliability predicate would be required, because “[a]t most, he
offered the opinion that Cellebrite copies data from a cellphone, which he derived from
his personal experience using the software.” United States v. Chavez-Lopez, 767 F. App’x
431, 434 (4th Cir. 2019). The Fourth Circuit relied on another court’s “analogy that
expert testimony is necessary to describe specialized medical tests but not to describe
reading a thermometer”; the DHS intern’s “testimony that he copied information from
a phone to a hard drive is, in our view, closer to the latter.” Id. at 435 (citing United States
v. Ganier, 468 F.3d 920, 926 (6th Cir. 2006)); accord Montijo-Maysonet, 974 F.3d at 48.
We agree with these courts that copying data using Cellebrite is like the common
exercise of using a thermometer,5 though it is perhaps even more like bagging up
evidence at a crime scene. To be sure, a given expert might offer sophisticated insights
into the process of bagging evidence, with topics ranging from the physics through
which the plastic bag keeps the evidence inside, to the anatomy of the officer doing the
bagging, or to the chemistry of the fluid in the marker that the officer uses to write his
initials on the bag. But ordinarily, we do not require expert testimony to establish the
5 Or, as the State argued in one Texas case, like a “photocopier for computers.” Williford, 127 S.W.3d at 312.
13 reliability of these aspects of the process before there may be simple testimony that an
officer took evidence, preserved it in a bag, and initialed it for authentication purposes.
In the same way, Detective Kiser effectively used a digital tool to “bag up” digital
evidence. While we are confident that the right experts could shed light on the many
complex dimensions involved in that process, we do not believe that an expert was
required to demonstrate the reliability of each of those dimensions before Detective
Kiser could testify that he simply copied data from one location to another. It is enough
that Detective Kiser authenticated the photos and text messages that were obtained
through the extraction, even without a proper Kelly predicate.6
We therefore overrule Wright’s sole issue, in which he contends that the trial
court reversibly erred by admitting the text messages and photos without an adequate
predicate under Kelly. We affirm the trial court’s judgment.
/s/ Wade Birdwell Wade Birdwell Justice
6 Even assuming that a reliability predicate was generally required for this type of evidence, we believe that the reliability of the scientific theory underlying Cellebrite has been adequately established by the previous cases that have assessed this sort of technology. See Rand, 2017 WL 4273177, at *8; Krause, 243 S.W.3d at 108–10; Sanders, 191 S.W.3d at 277–78; Williford, 127 S.W.3d at 312–13. “[C]ourts do not necessarily have to relitigate what is valid science in every case: ‘It is only at the dawn of judicial consideration of a particular type of forensic scientific evidence that trial courts must conduct full-blown “gatekeeping” hearings under Kelly.’” Morris, 361 S.W.3d at 655 (quoting Hernandez, 116 S.W.3d at 29). “Trial courts are not required to re-invent the scientific wheel in every trial.” Id. (quoting Hernandez, 116 S.W.3d at 29). “[C]ourts can take judicial notice of the reliability (or unreliability) of the type of evidence at issue.” Id. Thus, to the extent that Kelly applies, we take judicial notice of the cases holding that this type of technology satisfies Kelly’s mandate for reliable scientific underpinnings.
14 Publish
Delivered: February 4, 2021