People v. Lund ( 2021 )

Filed 6/1/21
CERTIFIED FOR PUBLICATION
IN THE COURT OF APPEAL OF THE STATE OF CALIFORNIA
FIRST APPELLATE DISTRICT
DIVISION FOUR
THE PEOPLE,
Plaintiff and Respondent,
A157205
v.
ERIC CURTIS LUND, (Solano County
Super. Ct. No. FCR310878)
Defendant and Appellant.
A jury convicted Eric Lund of one count of possession of
more than 600 images of child pornography, at least 10 of which
involved a prepubescent minor or a minor under 12 years old, in
violation of Penal Code section 311.11, subdivision (c)(1). The
trial court sentenced Lund to five years in prison.
Lund contends the trial court committed four errors. First,
he argues the trial court should have excluded some of the data
produced by a computer program because the data was case-
specific, testimonial hearsay under People v. Sanchez (2016)
63 Cal.4th 665 (Sanchez). Second, he argues the prosecution
failed to establish that the computer program was reliable and
generally accepted in the scientific community under People v.
Kelly (1976) 17 Cal.3d 24 (Kelly) and Sargon Enterprises, Inc. v.
University of Southern California (2012) 55 Cal.4th 747 (Sargon).
Third, Lund urges that his conviction should be reversed because
the prosecutor committed repeated, pervasive misconduct.
1
Finally, he argues that the trial court abused its discretion under
Evidence Code section 352 in allowing the prosecution to play for
the jury a number of child pornography videos. We reject each of
these arguments and therefore affirm the judgment.
I. BACKGROUND
A. Peer-to-peer networks
Peer-to-peer networks allow sharing of files, including child
pornography, over the internet. To access each different peer-to-
peer network, users must download and install software that
uses the programming protocol specific for that network.
eDonkey is one example of a peer-to-peer network commonly used
to share and download child pornography. eMule is a program
people commonly use to get onto the eDonkey network.
When a user installs peer-to-peer networking software, the
software randomly generates a globally unique identifier (GUID),
which is used to specifically identify the instance of the software
being used. The software also designates a five-digit port
number, which is necessary for the software to communicate with
the network. When a user sends out a search query, the request
goes to one or more other “peer” computers in the network, which
in turn propagate the request to other peers, and so on. This
process exponentially increases the number of computers
effectively receiving the search request. Each peer receiving the
query will respond to the original user with a list of files
matching the query that the peer has available for download.
Despite the exponential spread of a search query, a user’s query
will not typically reach all other peers on the peer-to-peer
2
network and a user will not see every file from every computer on
the network matching the query. When a computer connects to a
peer-to-peer network, it will automatically start receiving queries
from other users and returning a list of files that the computer
has available. Peer-to-peer networks use hash values to identify
each file being shared. A hash value is like a DNA signature for
a digital file; it is statistically unique and never changes, so it
provides a way to authenticate that two digital files are identical,
even if the names are different.
B. CPS Software
In August 2014, Vacaville police detective Jeffrey Datzman
was investigating child pornography cases over peer-to-peer
networks. One of the tools Datzman used was privately-
developed software called the Child Protection System (CPS).
CPS is the web interface for viewing results from a suite of
several software tools that each search for child pornography on a
specific peer-to-peer network.1 It is used around the world in 84
countries by over 10,000 users, all of whom are law enforcement
personnel.
The CPS software suite automates the process of searching
peer to peer networks. Previously, law enforcement officers
would have to manually input keyword search terms to discover
computers that were hosting suspected child pornography and
then further investigate those GUIDs. By contrast, CPS sends
1 Some of the CPS components include Peer Spectre, Nordic
Mule, Gnew Watch, and GT Logger. For simplicity, we use CPS
to refer both to the web interface and the underlying tools.
3
out search terms continuously. CPS also compares the files listed
in response to the keyword searches against CPS’s database of
hash values, which contains the hash values of files that law
enforcement officers somewhere in the world have previously
tagged as being child pornography. If there is a match between
the hash values for the files listed in response to the search and
the hash values in the CPS database, CPS logs the details of the
event in a CPS database for police officers to follow up on later.
CPS logs the filenames and hash numbers of the suspected child
pornography files being offered; the GUIDs, IP addresses, port
number, and, in most cases, software used to offer the files; and
the dates and times CPS detected the GUID with the files. Police
officers obtain records from internet service providers to
determine the physical location of the computer associated with
the GUIDs, IP addresses, and port numbers logged by CPS.
A match between the hash number of a particular file being
offered and a hash number in CPS’s database suggests the file is
likely child pornography. However, because child pornography
laws can differ from one jurisdiction to another, CPS users are
trained to always view a file personally in order to determine
conclusively whether the file constitutes child pornography under
applicable law. To assist with this, CPS also helps law
enforcement users create their own separate, local databases of
hash values called a media library. Where the CPS database
contains only hash values and not the child pornography files,
law enforcement users’ media libraries contain both the hash
values and the corresponding files. Users can use their media
4
libraries when they cannot download a file from the offering
computer directly to view it. In such cases, users can compare
the hash value of the file being offered to the hash value of a file
in the media library and then use the media library file to
confirm that the file is child pornography under applicable law.
C. Investigation of target GUID
When Datzman signed on to CPS in August 2014, he
noticed that there was one user, identified by a specific GUID,
who possessed several suspected child pornography files.
Datzman downloaded a few files from the target GUID and
confirmed that the files were in fact child pornography under
California law. This GUID moved between different IP addresses
but kept returning to a few addresses. This was unique, because
GUIDs that moved from one IP address to another usually did
not return to any of the IP addresses. After analyzing the target
GUID’s behavior, Datzman noticed that the GUID only showed
activity overnight on Wednesday, Thursday, Friday, and
Saturday nights. Because law enforcement officers often work
overnight shifts four nights a week from Wednesday evening
through Sunday morning, Datzman suspected that the target
GUID user was a security guard, law enforcement officer, or
someone else working such a shift.
Datzman obtained the physical addresses for the IP
addresses the target GUID was using. Because the target GUID
was active in the middle of the night when the businesses were
closed, Datzman did not consider the owners of any of the IP
addresses to be suspects. The most frequently recurring IP
5
address in Vacaville belonged to a business called the Yogurt
Beach Shack, which was owned by two former law enforcement
officers Datzman knew. Datzman confirmed that the Yogurt
Beach Shack’s wireless internet (wifi) router was “open,” meaning
it did not require a password, and could be accessed from outside
the building. Datzman therefore conducted overnight
surveillance at the Yogurt Beach Shack in early October 2014.
During the surveillance, Datzman connected to the Yogurt
Beach Shack’s router so that he could observe whether any
devices connected to the router and see such devices’ “mac ID,”
which is a unique specific identifier for a device. On one night, at
around 1:00 a.m., Datzman saw a device connect to the router,
and he recorded the mac ID. Datzman then drove around the
outside of the building to see who was nearby that could be using
the device. Datzman noticed a California Highway Patrol (CHP)
vehicle parked near the business. Lund was the sole occupant of
the vehicle, seated in the driver’s seat and looking down and to
his right at a lighted object. Datzman then contacted Sergeant
Jason Johnson in the Vacaville Police Department. Johnson
agreed to contact Lund using a ruse to determine his name. The
ruse succeeded and Lund told Johnson his name. After Johnson
spoke to Lund, Lund drove away, and Datzman noticed that the
mac ID of the device that was using the Yogurt Beach Shack
router dropped off at the same time. No other devices connected
to the router that night. Datzman later checked CPS to see if the
target GUID had been detected at the time that Lund was seen
at the Yogurt Beach Shack. CPS had no record of it at that IP
6
address at that time, but it detected the GUID at a different
router later that night.
Datzman contacted the CHP and learned from Lund’s
commanding officer that Lund was a sergeant assigned to the
Fairfield CHP office, worked alone, and worked a schedule that
matched the target GUID’s pattern of activity. Lund lived in
Chico, and stayed in a hotel in Vacaville during the days he was
working.
Every CHP patrol vehicle has a computer with software
installed that, when an officer logs into it, logs activity and also
activates a global positioning system (GPS) location tracker.
However, Lund had not logged into the software between June
and October 2014, so there was no GPS data for him.
Additionally, the dispatcher had recorded activity for him only
three times during that span. This paucity of records was
unusual and surprising. It was common knowledge among CHP
officers that logging into the vehicle computer would transmit
location data.
Datzman arranged with the commanding officer to put a
GPS tracker on the two patrol cars assigned to CHP sergeants in
the Fairfield office. The first night after the GPS trackers were
installed, the GPS tracker showed that the car that Lund had
been observed driving was stopped for over two hours at a
location in Cordelia Park near a house with open wifi. CPS
detected the target GUID with child pornography that night at
that same location for about two hours.
7
D. Searches of devices in Lund’s desk, car, and locker
Pursuant to search warrants, Vacaville police officers then
searched Lund’s desk at work and his personal car. Police found
a flash drive in the center console of Lund’s car. This flash drive
did not contain child pornography. In the trunk of the car were
Lund’s CHP uniforms, his citation book, his old cell phone in a
box, a USB wifi adaptor, and a tan backpack. Inside the
backpack were two long range USB wifi adaptors with a panel
antenna that could be used to pick up a wifi signal from greater
distances, a laptop, two external hard drives, and three flash
drives. One of the USB wifi adaptors had a mac ID identical to
the one Datzman recorded from the router at the Yogurt Beach
Shack.
The cell phone and one of the flash drives from the trunk of
the car did not contain child pornography. This flash drive had
been connected to a computer in Lund’s home. The other two
flash drives contained deleted child pornography videos and
pictures that Datzman forensically recovered.
The external hard drives together contained over 10,000
files that Datzman suspected to be child pornography, based on
their hash values’ matches to the CPS database. Datzman
reviewed a sample of 73 videos from the hard drives and
confirmed that they were child pornography, with almost all of
them containing at least one prepubescent minor. One of the
hard drives also contained the same version of eMule that the
target GUID used and the software necessary to use the panel
antenna. Datzman concluded that one of the hard drives had
8
been used to store child pornography since 2012 because one of
the folders on it indicated that eMule had been used with the
hard drive since that date.
The laptop contained a copy of the eMule software with the
same version number, GUID, and port number that CPS had
detected. eMule had been used to download over 3,000 complete
files whose names suggested they were child pornography. Like
the flash drive from his desk and one of the hard drives, the
laptop had the software necessary to use the panel wifi antenna.
The laptop showed it had connected to the router at the Yogurt
Beach Shack and had run eMule on the night Datzman and
Johnson observed Lund there, but the program had crashed. The
laptop also showed it had connected to the router where the IP
address in Cordelia Park was located. The laptop had been used
to access email and Facebook accounts, but the lack of activity in
those accounts or information available about the named users
indicated that the accounts did not belong to real people.
The laptop’s last wifi connection was to the network at the
Fairfield Inn in Vacaville an hour before Lund’s arrest. This
hotel was across the street from the hotel where Lund had a
reservation and where his car was seen during the day of his
arrest, before he came to the office. With a panel range antenna,
the laptop could have accessed the Fairfield Inn’s internet from
Lund’s room. The laptop’s user viewed child pornography files
throughout that day, and CPS detected the target GUID as being
active throughout that day.
9
Lund’s desk, which was known to other officers to be
unlockable, contained Lund’s active cell phone and three flash
drives. All three flash drives showed they had been used with
the computer on Lund’s desk. None of the devices contained
pornography, but one flash drive contained a copy of the eMule
program and the software for the long range wifi adaptor found
in the trunk of Lund’s car.
About a week after Vacaville police searched Lund’s desk
and personal car, CHP officers searched Lund’s locker at the
Fairfield CHP building. They found hotel breakfast cards and a
Diskgo flash drive with the first three digits of Lund’s CHP badge
number written on it. CHP Officer Ryan Duplissey took the
Diskgo flash drive for analysis. In his report, Duplissey
originally stated that the flash drive was found connected to
Lund’s work computer, but he later corrected his report to reflect
that it was found in his locker. Datzman later acquired the drive
and performed his own analysis. Both analyses showed the flash
drive contained documents associated with Lund as well as 10
child pornography files that had been marked for deletion but
could be forensically recovered. CHP officers also searched
electronic devices found in Lund’s home but did not find any child
pornography on them.
E. Procedural history
After Lund was arrested and charged, pretrial litigation
relating to the constitutionality of the searches stretched over the
course of several years and involved two writ petitions to this
court. A jury trial in the summer of 2018 resulted in a mistrial.
10
At the second trial in October 2018, the prosecution played for
the jury brief portions of some of the child pornography files
found on each device that Datzman had confirmed were child
pornography.
Unlike the first trial, Lund testified in his own defense. He
denied ever possessing or downloading child pornography. Lund
said he was sitting in the Yogurt Beach Shack parking lot to eat
and denied having a laptop in the car or connecting to the
internet. He said the CHP computer in his vehicle, which was to
the driver’s right, would glow at night. He denied driving the
patrol car that was recorded by GPS in Cordelia Park on the
same night that CPS detected activity there and said he was in
the office the whole night. On cross-examination, however, he
admitted that he sent his wife a text that night saying he was
going to go out and drive for fresh air. Lund also denied that the
electronic devices found in his car, desk, or locker were his. He
claimed other officers used his desk when he was not there. Lund
testified that he never locked his locker because he had once
forgotten the key at home in Chico and been unable to get his
uniform, but he admitted on cross-examination this was not
general knowledge. He said he kept hotel breakfast cards in his
desk, not his locker, so someone at CHP must have moved them
into the locker.
Lund explained he did not use the CHP computer in the
patrol car because he thought it was unsafe. Lund said he
instead used his radio to make requests through dispatch. He
claimed the records of his radio activity were not obtained from
11
the correct office, and the proper records would have shown more
activity. Lund admitted he had been suspended in 1996 for using
a screensaver on his work computer that displayed adult
pornography. Lund claimed that Datzman planted the evidence
against him but had no theory for why.
In rebuttal, the prosecution called retired CHP Sergeant
Steven Lott, who testified that officers used to tell him they could
not reach Lund on the radio during his shifts. Lott also recalled
that Lund would leave the office in the middle of the night
complaining about the office being too hot or needing to stay
awake, but it seemed like an excuse to leave the office.
The jury found Lund guilty and the trial court sentenced
him to five years in prison.
II. DISCUSSION
A. Sanchez error
Lund first challenges the trial court’s admission of the hash
value information from the CPS database. He contends William
Wiltse’s and Officer Datzman’s testimony about the CPS hash
values corresponding to suspected child pornography files was
inadmissible hearsay because the hash value database consists of
out of court statements made by unidentified officers across the
country and around the world.2 He further argues the admission
of this testimony violated his Sixth Amendment right to confront
and cross-examine the witnesses against him.
2 Wiltse oversees the development of the CPS software, and has
himself written certain components of the software.
12
Relevant legal principles and standard of review
“[A] hearsay statement is one in which a person makes a
factual assertion out of court and the proponent seeks to rely on
the statement to prove that assertion is true. Hearsay is
generally inadmissible unless it falls under an exception.”
(Sanchez, supra, 63 Cal.4th at p. 674.) “Documents like letters,
reports, and memoranda are often hearsay because they are
prepared by a person outside the courtroom and are usually
offered to prove the truth of the information they contain.” (Id. at
pp. 674.) However, “ ‘[o]nly people can make hearsay statements;
machines cannot.’ ” (Id. at p. 690, fn. 16.)
The use of hearsay potentially conflicts with defendants’
rights under the Sixth Amendment to the U.S. Constitution to
confront witnesses against them. (Sanchez, supra, 63 Cal.4th at
pp. 679–680.) As the California Supreme Court summarized, “In
light of our hearsay rules and Crawford [v. Washington (2004)
541 U.S. 36 (Crawford)], a court addressing the admissibility of
out-of-court statements must engage in a two-step analysis. The
first step is a traditional hearsay inquiry: Is the statement one
made out of court; is it offered to prove the truth of the facts it
asserts; and does it fall under a hearsay exception? If a hearsay
statement is being offered by the prosecution in a criminal case,
and the Crawford limitations of unavailability, as well as cross-
examination or forfeiture, are not satisfied, a second analytical
step is required. Admission of such a statement violates the right
to confrontation if the statement is testimonial hearsay, as the
high court defines that term.” (Id. at p. 680, italics omitted.)
13
Sanchez applied both steps of this inquiry to expert
testimony. (Sanchez, supra, 63 Cal.4th at pp. 680, 687.) Experts
had long been allowed to rely on hearsay when offering their
opinions. (Id. at p. 676.) This rule arose because experts
frequently acquire general knowledge in their field of expertise
from third parties, but the rule was extended to apply to case-
specific facts as well. (Id. at pp. 676, 678–679.) Case-specific
facts are “those relating to the particular events and participants
alleged to have been involved in the case being tried.” (Id. at
p. 676.) Courts had recognized the tension between experts’ need
to consider extrajudicial matters with defendants’ interest in
avoiding the substantive use of unreliable hearsay. (Id. at
p. 679.) They had tried to balance these concerns by generally
allowing experts to explain the bases for their opinions, even
when those bases were general or case-specific hearsay, subject to
an instruction that juries should only consider such hearsay as
the basis for the expert’s opinion and not for its truth. (Ibid.)
Sanchez concluded this was a mistake because a jury
cannot avoid considering the truth of case-specific hearsay
underlying an expert’s testimony. (Sanchez, supra, 63 Cal.4th at
pp. 679, 684.) Sanchez explained, “When an expert is not
testifying in the form of a proper hypothetical question and no
other evidence of the case-specific facts presented has or will be
admitted, there is no denying that such facts are being
considered by the expert, and offered to the jury, as true.” (Id. at
p. 684.) The court therefore adopted “the following rule: When
any expert relates to the jury case-specific out-of-court
14
statements, and treats the content of those statements as true
and accurate to support the expert’s opinion, the statements are
hearsay. It cannot logically be maintained that the statements
are not being admitted for their truth. If the case is one in which
a prosecution expert seeks to relate testimonial hearsay, there is
a confrontation clause violation unless (1) there is a showing of
unavailability and (2) the defendant had a prior opportunity for
cross-examination, or forfeited that right by wrongdoing.” (Id. at
p. 686, fn. omitted.) Sanchez noted, however, that experts can
still “rely on information within their personal knowledge, and
they can give an opinion based on a hypothetical including case-
specific facts that are properly proven. They may also rely on
nontestimonial hearsay properly admitted under a statutory
hearsay exception.” (Id. at p. 685.)
As to the second step of the hearsay inquiry, Sanchez
reviewed several Supreme Court decisions dealing with
statements made to police officers and summarized, “Testimonial
statements are those made primarily to memorialize facts
relating to past criminal activity, which could be used like trial
testimony. Nontestimonial statements are those whose primary
purpose is to deal with an ongoing emergency or some other
purpose unrelated to preserving facts for later use at trial.”
(Sanchez, supra, 63 Cal.4th at p. 689.)
Sanchez did not address the standard of review for
determining whether a statement is case-specific, testimonial
hearsay. As the first step of the analysis is a “traditional hearsay
inquiry” into whether a statement was made out of court and is
15
offered for its truth, we apply the abuse of discretion standard at
this step. (Sanchez, supra, 63 Cal.4th at p. 680.) “A trial court’s
decision to admit or exclude evidence is reviewed for abuse of
discretion, and it will not be disturbed unless there is a showing
that the trial court acted in an arbitrary, capricious, or absurd
manner resulting in a miscarriage of justice.” (People v. Wall
(2017) 3 Cal.5th 1048, 1069.) By contrast, Sanchez’s second step,
concerning whether a statement is testimonial, implicates the
constitutional right of confrontation, so we independently review
that issue. (People v. Nelson (2010) 190 Cal.App.4th 1453, 1466.)
Analysis
Lund contends testimony from both Wiltse and Datzman
violated Sanchez because both witnesses relied on CPS’s
database of hash values corresponding to previously identified
child pornography. As to Wiltse, Lund asserts that Wiltse
testified that CPS uses its hash value database to search peer-to-
peer networks and that officers can determine whether a file
offered by a suspect is child pornography, even if they cannot
download the file, by comparing the file’s hash value to CPS’s
hash value database. Lund also contends that both Wiltse and
Datzman testified that CPS showed the target GUID downloaded
a file that CPS had tagged as child pornography. As to Datzman,
Lund asserts that Datzman opined that the hash value of a file is
very important to determining whether the file is child
pornography. Lund also cites Datzman’s testimony that he used
a computer program to compare the hash values of files on the
hard drives found in Lund’s trunk to the CPS hash value
16
database and Datzman’s opinion that the hard drive had been
using eMule to obtain child pornography as early as 2012.
Lund argues this testimony violates Sanchez because it is
case-specific, testimonial hearsay. He contends it is case-specific
hearsay because the entirety of the case against him was based
on the CPS data and CPS cannot work without the assumption
that whoever put a hash value in the CPS database correctly
tagged it as child pornography. He argues the CPS hash value
data is testimonial because it represents the fruits of previous
law enforcement investigations.
We conclude the trial court did not abuse its discretion in
determining the CPS hash values were not hearsay in this case
because they were not admitted for their truth, so we do not
reach the question of whether the hash values are testimonial.3
Preliminarily, Lund’s Sanchez arguments regarding the
CPS hash values rest on a misinterpretation of the record,
stemming from the fact that the CPS software relies on hash
values in two separate databases. First, CPS maintains a
database of hash values on its own servers, without associated
files. It uses this database as part of its searches of peer-to-peer
networks. CPS uses keywords to search the peer-to-peer
networks, then matches the hash values of the files that are
listed in response to those keyword searches against the CPS
3 While we apply the abuse of discretion standard in this
case, as explained above, we would reach the same conclusion
even if we were to independently review the trial court’s
conclusion that the CPS hash values are not hearsay.
17
database of hash values of suspected child pornography. In cases
where the hash values match, the software logs the file being
offered on the peer to peer network as suspected child
pornography and saves the record for law enforcement officials to
review later.
Second, CPS assists law enforcement officers with
maintaining a media library, which is a local database of child
pornography files and associated hash values. Police officers use
this media library, not the CPS hash value database, to
determine whether a file that they cannot download from a
suspect is in fact child pornography. The police officer does so by
matching the hash value of the suspect’s file to the hash value of
a file in the library. Using the copy of the file in the media
library, the officer can opine conclusively that the file is child
pornography.
Wiltse and Datzman explained carefully and repeatedly
that the only way to determine whether a file is child
pornography is for an officer to personally view a file or have
personally viewed an identical copy of the file (meaning a
matching hash value) in the past. This is because the police
officer who input a hash value in the CPS database could have
erred, and because the legal definitions of child pornography
differ between jurisdictions. Wiltse testified consistently that the
hash value database exists to create criminal leads about
suspected but unconfirmed child pornography, not to provide
definitive proof that any file is child pornography. Moreover,
Datzman’s conclusion that the hard drive had been used to store
18
downloaded child pornography since 2012 was not based on the
CPS hash values, but rather on the fact that there were folders
indicating that eMule had been used with the hard drive since
then.
This proper understanding of the record regarding CPS
demonstrates why Lund’s Sanchez argument fails at the first
step. “ ‘Out-of-court statements that are not offered for their
truth are not hearsay under California law [citations], nor do
they run afoul of the confrontation clause.’ ” (People v. Bell (2019)
7 Cal.5th 70, 100.) Neither Wiltse nor Datzman relied on the
CPS hash values for their truth to opine that any file was child
pornography. The prosecution used Wiltse’s and Datzman’s
testimony about the hash values only to explain Datzman’s
course of conduct in investigating the GUID and Lund. Lund’s
defense that Datzman or someone else planted the evidence in
his car, desk, and locker made it fair for the prosecution to show
that Datzman followed reasonable leads. This is an example of
the principle that “ ‘ “ ‘[e]vidence of a declarant’s statement that
is offered to prove that the statement imparted certain
information to the hearer and that the hearer, believing such
information to be true, acted in conformity with that belief . . . is
not hearsay, since it is the hearer’s reaction to the statement that
is the relevant fact sought to be proved, not the truth of the
matter asserted in the statement.’ ” ’ ” (Ibid.) Even if every entry
in the CPS hash value database were wrong, such hypothetical
errors would not undermine the prosecution’s proof of the
elements of the charge against Lund. The prosecution proved the
19
files found on devices associated with Lund were pornography
either by showing them to the jury or by having Datzman testify
that he personally viewed them and verified that they met the
legal description.4
Even if Lund were correct that the CPS hash value data
constituted hearsay that was also testimonial, its admission was
harmless beyond a reasonable doubt. (See Sanchez, supra,
63 Cal.4th at p. 698 [applying federal standard of harmlessness
beyond a reasonable doubt to violation of Confrontation Clause
through admission of testimonial hearsay].) At its core, the
prosecution proved its case by establishing that the police found
child pornography on electronic devices found in Lund’s locker
and in the trunk of his car. These devices were found in locations
within his control, particularly his car. This evidence is
circumstantial but strongly persuasive, especially because Lund
4 The facts of this case are therefore distinguishable from
cases like U.S. v. Juhic (8th Cir. 2020) 954 F.3d 1084, 1088–1089,
and U.S. v. Bates (11th Cir. 2016) 665 Fed.Appx. 810, 814–815,
which held that CPS reports were inadmissible testimonial
hearsay. In those cases, it appears the prosecution used the
reports’ notation that certain files had been previously identified
as child pornography as evidence that files were indeed child
pornography. (Juhic, at p. 1089 [CPS reports were “out-of-court
statements offered for the truth of the matter asserted: that the
videos and images were child pornography”]; Bates, at p. 815
[“The record shows that the government used the reports to
demonstrate the steps of [the officer’s] investigation and to prove
that the files [the defendant] downloaded were child
pornography”].) There is no indication in these decisions that
any witnesses identified any files as child pornography after
viewing them, like Datzman did here.
20
does not explain how someone would have gained access to his
car.
Lund notes that the cell phone and the flash drive found in
his trunk, which had been connected to a computer in his home,
contained no child pornography. He therefore asserts that none
of the digital devices containing child pornography could be
linked to him. This argument ignores the critical fact of the
devices found in the trunk of his car, as well as the flash drive
found in his locker. That flash drive, which contained deleted
child pornography, could be tied to Lund both by its location and
by the presence of Lund’s files on it. It was not common
knowledge that Lund’s locker was unlocked. Lund contended the
devices were planted, but he offered the jury no reason why
Datzman or any of the officers who performed the searches of his
car, desk, and locker would have wanted to frame him or had the
opportunity to do so.
Lund also takes the position that the theory that he used
eMule to download child pornography was essential to the
prosecution’s case. He observes that the CPS evidence helped
refute Lund’s defense that the electronic devices were planted by
showing Lund was at the same locations as the devices
containing the pornography. While the CPS evidence was not
legally essential, it was certainly helpful for rebutting Lund’s
defense. But the CPS data that rebutted the defense was the
date, time, and IP address information, since these types of data
helped establish the connections with Lund’s movements. The
GUID, eMule version number, and port also helped connect Lund
21
to the laptop found in his trunk. The CPS hash values were not
helpful to this rebuttal. For example, if CPS had logged the
target GUID as offering some other type of file, like copyrighted
movies or music, the CPS location-related and laptop-related
data would have been just as powerful for the prosecution’s effort
to undermine Lund’s defense.
Lund does not argue CPS’s date, time, IP address, GUID,
version number, or port data are hearsay, nor could he, because
they were all generated automatically by the CPS software.
(Sanchez, supra, 63 Cal.4th at p. 690, fn. 16 [“ ‘Only people can
make hearsay statements; machines cannot’ ”].) Lund instead
appears to assume that if the CPS hash values were hearsay, the
rest of the CPS data would have been excluded. This assumption
is unfounded, as the hash values could easily have been omitted
from the relevant exhibits and testimony. Lund also maintains
the CPS hash value data helped establish that Datzman
conducted a thorough investigation, in response to Lund’s
attempts to show the investigation was flawed. This is true, but
Lund’s attacks on the investigation’s minor flaws were not so
strong that the CPS hash value evidence was necessary to refute
them, and thus the admission of that evidence—even if
erroneous—did not cause prejudice.
B. Kelly and Sargon error
In his second argument, Lund argues the trial court should
have granted his motion in limine to exclude evidence of CPS
entirely because the prosecution failed to establish that CPS
satisfied the standard for admission of new scientific evidence
22
under Kelly, supra, 17 Cal.3d 24, and because the trial court
failed to perform its gatekeeping function under Sargon, supra,
55 Cal.4th 747. He argues the admission of CPS evidence
rendered his trial fundamentally unfair in violation of the 14th
Amendment and was not harmless under the federal or state
standards.
Relevant legal principles and standard of review
a. Kelly
“Kelly was the genesis of a rule, previously called the
‘Kelly/Frye rule,’ [5] that governs the admissibility of evidence
derived from new scientific techniques. ‘Under Kelly, the
proponent of evidence derived from a new scientific technique
must establish that (1) the reliability of the new technique has
gained general acceptance in the relevant scientific community,
(2) the expert testifying to that effect is qualified to give an
opinion on the subject, and (3) the correct scientific procedures
were used.’ ” (People v. Jones (2013) 57 Cal.4th 899, 936.) “The
purpose of these threshold requirements—commonly referred to
as the Kelly test—is to protect against the risk of credulous juries
attributing to evidence cloaked in scientific terminology an aura
of infallibility.” (People v. Peterson (2020) 10 Cal.5th 409, 444
(Peterson).)
“Not every subject of expert testimony needs to satisfy the
Kelly test. Courts determining whether Kelly applies must
5“See Frye v. U.S. (D.C.Cir.1923) 293 F. 1013, 1014,
superseded by statute as explained in Daubert v. Merrell Dow
Pharmaceuticals, Inc. (1993) 509 U.S. 579 [(Daubert)].”
23
consider, first, whether the technique at issue is novel, because
Kelly ‘ “only applies to that limited class of expert testimony
which is based, in whole or part, on a technique, process, or
theory which is new to science and, even more so, the law.” ’
[Citation.] Second, courts should consider whether the technique
is one whose reliability would be difficult for laypersons to
evaluate. A ‘Kelly hearing may be warranted when “the
unproven technique or procedure appears in both name and
description to provide some definitive truth which the expert
need only accurately recognize and relay to the jury.” ’ [Citation.]
Conversely, no Kelly hearing is needed when ‘[j]urors are capable
of understanding and evaluating’ the reliability of expert
testimony based in whole or in part on the novel technique.”
(Peterson, supra, 10 Cal.5th at p. 444.)
“Appellate courts review de novo the determination that a
technique is subject to Kelly.” (People v. Jackson (2016) 1 Cal.5th
269, 316.)
b. Sargon
Regardless of whether expert evidence relates to a new
scientific technique under Kelly, a trial court must ensure that
expert testimony has a sufficient basis to merit the jury’s
consideration. (Sargon, supra, 55 Cal.4th at p. 770.) “[U]nder
Evidence Code sections 801, subdivision (b), and 802, the trial
court acts as a gatekeeper to exclude expert opinion testimony
that is (1) based on matter of a type on which an expert may not
reasonably rely, (2) based on reasons unsupported by the
material on which the expert relies, or (3) speculative.” (Id. at
24
pp. 771–772.) “The trial court’s preliminary determination
whether the expert opinion is founded on sound logic is not a
decision on its persuasiveness. The court must not weigh an
opinion’s probative value or substitute its own opinion for the
expert’s opinion. Rather, the court must simply determine
whether the matter relied on can provide a reasonable basis for
the opinion or whether that opinion is based on a leap of logic or
conjecture. The court does not resolve scientific controversies.
Rather, it conducts a ‘circumscribed inquiry’ to ‘determine
whether, as a matter of logic, the studies and other information
cited by experts adequately support the conclusion that the
expert’s general theory or technique is valid.’ [Citation.] The
goal of trial court gatekeeping is simply to exclude ‘clearly invalid
and unreliable’ expert opinion. [Citation.] In short, the
gatekeeper’s role ‘is to make certain that an expert, whether
basing testimony upon professional studies or personal
experience, employs in the courtroom the same level of
intellectual rigor that characterizes the practice of an expert in
the relevant field.’ ” (Id. at p. 772.)
“Except to the extent the trial court bases its ruling on a
conclusion of law (which we review de novo), we review its ruling
excluding or admitting expert testimony for abuse of discretion.”
(Sargon, supra, 55 Cal.4th at p. 773.)
Analysis
a. Kelly
Lund argues the trial court should have excluded all
evidence regarding CPS under Kelly because CPS is an unproven,
25
largely untested, and inherently unreliable computer program.
He contends it had the appearance of reliability but there was no
evidence that CPS was widely accepted or that CPS followed the
correct procedures.
Lund’s argument bypasses the threshold Kelly question of
whether the testimony about CPS is part of “ ‘ “that limited class
of expert testimony which is based, in whole or part, on a
technique, process, or theory which is new to science and, even
more so, the law,” ’ ” whether it “ ‘appears in both name and
description to provide some definitive truth which the expert
need only accurately recognize and relay to the jury,’ ” and
whether jurors “ ‘are capable of understanding and evaluating’
the reliability of expert testimony based in whole or in part” on
CPS. (Peterson, supra, 10 Cal.5th at p. 444.) Instead, he simply
assumes both that CPS itself is a scientific process or technique,
rather than a program using such a technique, and that it was
novel. These assumptions are misplaced. CPS is not a technique
or process; it is a program that deploys a technique or executes a
process. (Cf. People v. Nolan (2002) 95 Cal.App.4th 1210, 1215
[“a Kelly/Frye hearing is not required for new devices; it applies
to new methodologies”]; People v. Lazarus (2015) 238 Cal.App.4th
734, 782–786 [distinguishing between new DNA test kit and the
existing technique it used].) And the technique or process that
CPS uses is not novel, nor is it so persuasive that it threatens to
beguile a jury with misleading scientific certainty.
Before CPS, a police investigator investigating peer-to-peer
networks had to manually enter search terms in a software
26
program, record the results, and follow up on leads by obtaining
records from internet service providers to connect IP addresses to
physical locations. The only novelty in CPS lies in its approach of
automating the pre-existing search process. Instead of officers
needing to perform manual searches during working hours and
noting the pertinent information regarding any leads, CPS runs
searches around the clock and has the computer log the relevant
details regarding leads. (U.S. v. Thomas (2d Cir. 2015) 788 F.3d
345, 348 [describing CPS’s operation].) CPS’s process or
technique, then, is simply to perform the same searches that law
enforcement officers used to do, but in larger volume and at all
times of the day, and record the results for later perusal.
CPS’s ability to generate a larger volume of search results
over a longer period of time undoubtedly makes it a useful, time-
saving device. It also unlocks the possibility of using the search
results for different purposes, such as the prosecution’s approach
here of using CPS results to map the behavior and location of the
target GUID over time. But CPS’s results themselves do not
purport to offer any more scientific or technical certainty
regarding the data they contain than the manual searches CPS
replaced. Computers are now commonplace, so the general public
can be expected to be generally familiar with the notions that
software can repeatedly perform simple tasks that previously
would have taken extensive human labor to complete in the same
quantity and that the resulting quantity of data can be analyzed
and used in different ways that were not possible before. We do
not see how CPS’s addition of automation to the routine police
27
work of finding and recording evidence is so mysterious or
seemingly authoritative that it would be difficult for laypersons
to understand or evaluate it. As a federal district court remarked
when considering the admissibility of evidence from software like
CPS, “Computer programming is not a scientific theory or
technique, it is not new or novel, and it does not implicate the
[c]ourt’s responsibility to keep ‘junk science’ out of the courtroom.
Any doubts about whether [the software] operates in the manner
that [its creator] represents go to the weight, and not the
admissibility, of his testimony.” (United States v. Blouin (W.D.
Wash., Aug. 15, 2017, No. CR16-307 TSZ) 2017 WL 3485736, *7.)6
The prosecution’s use of CPS can be analogized to the
computerized fingerprint matching program challenged in People
v. Farnam (2002) 28 Cal.4th 107. In that case, the police had
used a computerized database for fingerprint matching to
produce a list of candidates, including the defendant, whose
fingerprints resembled those at the crime scene. (Id. at p. 159.)
Fingerprint analysts then compared the defendant’s fingerprint
to a fingerprint found at the crime scene and concluded they
matched. (Ibid.) The California Supreme Court held that the
evidence regarding the computerized fingerprint matching
program did not implicate Kelly. (Id. at p. 160.) The expert who
6 Though the federal Daubert standard for admission of
scientific evidence differs somewhat from the Kelly standard,
appellate decisions affirming the admission of scientific evidence
under that standard are relevant to the Kelly analysis. (People v.
Venegas (1998) 18 Cal.4th 47, 88 [relying in part on state court
decisions using the Daubert standard]; People v. Buell (2017)
16 Cal.App.5th 682, 690–691 [same].)
28
relied on the database never claimed that the database positively
identified the defendant’s print. (Ibid.) The jurors could
determine the reliability of the database by comparing the
defendant’s fingerprint to that found at the crime scene. (Ibid.)
And no opinion regarding the fingerprint identification was based
on the computer results. (Ibid.) The mere use of the computer
system to narrow the range of potential candidates was
insignificant because the prosecution relied on the long-
established method of expert fingerprint comparison to show the
defendant’s prints matched. (Ibid.)
People v. Johnson (2006) 139 Cal.App.4th 1135 followed
People v. Farnam in the context of a DNA database. There, a lab
technician obtained a DNA profile obtained from analysis of a
sexual assault examination kit taken from the victim. (Johnson,
at p. 1143.) The technician submitted the sample to a
computerized DNA database and found a match to the defendant.
(Ibid.) A blood sample was then taken from the defendant, and
the technician matched the DNA profiles of the rape kit sample
and the defendant’s sample. (Ibid.) The court affirmed the
admission of this DNA evidence, in part because the DNA
analysis techniques at issue had been generally accepted for some
time. (Id. at p. 1149.) But the court further rejected the
defendant’s argument that the expert erred in calculating the
probability of a match between a random person and the rape kit
DNA profile. (Ibid.) The court stated, “the use of database
searches as a means of identifying potential suspects is not new
or novel,” because DNA databases and data bank statutes have
29
been enacted in all 50 states and by the federal government.
(Ibid.) The court made clear that its “core point” was that “the
database search merely provides law enforcement with an
investigative tool, not evidence of guilt.” (Id. at p. 1150.) In the
court’s view, “the means by which a particular person comes to be
suspected of a crime—the reason law enforcement’s investigation
focuses on him—is irrelevant to the issue to be decided at trial,
i.e., that person’s guilt or innocence, except insofar as it provides
independent evidence of guilt or innocence.” (Ibid.)
Like the fingerprint database in People v. Farnam, supra,28 Cal.4th at page 159 or the DNA database in People v. Johnson,
supra, 139 Cal.App.4th at page 1149, CPS automated the process
of searching for computers suspected of containing child
pornography. But as in those cases, a police officer, Datzman—
not CPS—made the ultimate conclusion that the files on the
devices linked to Lund were child pornography. The
prosecution’s use of CPS went slightly beyond the fingerprint or
DNA databases in those cases, since the prosecution relied on
CPS to match the target GUID’s patterns and locations to Lund’s
schedule and movements. In this regard, the CPS information
could be said to provide, as the court put it in Johnson, at page
1150, evidence of guilt independent of the child pornography
found on the devices. Nonetheless, we do not view this fact as
significant because, as in People v. Farnam, the reliability of this
information was readily apparent to the jury.
The CPS data had no aura of authority on its own because
CPS had no records mentioning Lund specifically or matching to
30
an address that could be tied to him. The CPS evidence was only
persuasive to the extent that the target GUID’s behavior and
locations matched Lund’s schedule and movements, and the jury
was well-equipped to evaluate whether the two matched. The
prosecution’s evidence on that point was also strong. At precisely
the time that Datzman observed a device access the router at the
Yogurt Beach Shack and then drop off, Datzman and Johnson
observed Lund in his police car outside the business and then
drive away. The long range antenna found in Lund’s trunk
matched the mac ID Datzman observed access the router. The
antenna was used with the laptop found in the trunk. The laptop
had accessed the router at the Yogurt Beach Shack at the same
time Lund was there. The laptop also had a copy of eMule with
the same GUID and port number that CPS detected. The laptop
accessed the router near Cordelia Park, at the same time that
CPS detected the GUID there and the police car Lund was
driving was located by GPS there. CPS detected the GUID using
the Fairfield Inn internet at the same time Lund’s car was
observed at the hotel nearby, where he also had a reservation.
The only point on which CPS did not match Lund’s movements
was that CPS did not detect the GUID at the Yogurt Beach Shack
when Lund was there. But Datzman’s later analysis of the laptop
suggested it was because the eMule program crashed when Lund
was trying to use it there.
The jury could evaluate the credibility of the individual
witnesses’ testimony that the prosecution used to establish
Lund’s movements, and then determine for itself whether that
31
testimony matched the data from CPS. CPS did not have any
particular heightened power to dazzle the jury, so there was no
need to hold a Kelly hearing to evaluate it.
b. Sargon
Even though Kelly did not apply to the CPS evidence, the
trial court was still obligated under Sargon to exclude the CPS
testimony if it was “(1) based on matter of a type on which an
expert may not reasonably rely, (2) based on reasons unsupported
by the material on which the expert relies, or (3) speculative.”
(Sargon, supra, 55 Cal.4th at pp. 771–772.) Lund contends
Wiltse’s and Datzman’s expert testimony based on CPS failed
under Sargon because there was insufficient evidence to show
that CPS reliably worked. We disagree.
The trial court did not abuse its discretion in finding Wiltse
provided sufficient information to demonstrate CPS was reliable
enough to be presented to the jury. Wiltse opined directly, as an
expert, that the software is reliable and widely used in 84
countries by over 10,000 licensed users. Wiltse wrote the code for
many of the tools in the CPS, and he oversaw the development of
others as a supervisor. Wiltse had never had a complaint from
users that CPS’s leads were unsubstantiated. CPS itself had
never been hacked or corrupted by an external source. Wiltse
tested the program extensively in a closed environment to ensure
its accuracy before using it on the wider internet. Wiltse
explained this initial testing was sufficient because there was no
way to test it outside the closed environment and until the peer-
to-peer network protocol changed, there was no need to change
32
the CPS software. Wiltse had testified before as an expert in
Oregon, Illinois, Utah, New Mexico, Vermont, and Florida, and
had twice participated in court-ordered testing of the program for
federal courts in California. Wiltse was not aware of any
convictions ever being reversed as a result of him being allowed
to testify about CPS.
Lund offers several theories as to how Wiltse failed to prove
CPS’s reliability, such as the contentions that Wiltse improperly
assumed that the peer-to-peer protocols did not change, failed to
show the software remained reliable after its release, relied on
the absence of anecdotal reports of error to conclude the software
was reliable, and did not subject the software to third-party
testing. These arguments are speculative. Lund cross-examined
Wiltse at the pre-trial hearing and at trial about any flaws in
CPS or need for third party testing, but he did not succeed in
casting any doubt on its operations. Nor has Lund presented in
this court any specific reasons to discount Wiltse’s testimony. To
the contrary, the trial court could have reasonably concluded that
the pre-release testing and track record of success demonstrated
that CPS was sufficiently reliable to provide the basis for Wiltse’s
and Datzman’s testimony.
Lund contends Wiltse should have provided more detail
about how CPS worked, such as its source code and how it
determined what items to search for and where to search for
them. He also argues there was no evidence about how CPS
stored information, how it chose what information to store, and
how CPS took information from the peer-to-peer network and
33
saved it within the CPS program. But Wiltse explained that CPS
searches using terms commonly associated with child
pornography files, and its different component programs searched
different peer-to-peer networks that commonly offered child
pornography. He talked about how CPS populated its database
with information other computers provided in response to queries
of the peer-to-peer networks or, in the case of hash values, by
CPS users worldwide. He also detailed how the CPS component
used here, Nordic Mule, logs its own dates and times that files
are being offered and makes direct connections to computers
hosting suspected child pornography to verify those IP addresses,
so that those fields in its database are absolutely accurate. Lund
does not cite authority for or explain how technical detail about
the source code or algorithms was required (or even helpful) to
establish CPS’s reliability.
Finally, we note that many courts have concluded CPS is
sufficiently reliable to provide probable cause for a search under
the Fourth Amendment. (E.g., U.S. v. Thomas, supra, 788 F.3d
at p. 353 [“we discern no error—much less, clear error—in the
District Court’s finding that CPS was a reliable tool that could
serve as the basis of a search warrant affidavit”]; U.S.A. v.
McKinion (C.D. Cal., July 21, 2017, No. 2:14-CR-00124-CAS-1)
2017 WL 3137574, at *4, fn. 7; U.S. v. Collins (S.D. Iowa 2009)
753 F.Supp.2d 804, 809 [Peer Spectre, one of the CPS component
tools, “is routinely and widely used by law enforcement officers to
conduct [peer-to-peer] investigations, with wide-ranging
acceptance for reliability,” notwithstanding the defendant’s
34
expert’s claims to the contrary]; U.S. v. Naylor (S.D. W.Va. 2015)
99 F.Supp.3d 638, 643 [finding, based in part on police officer’s
experience of 100% reliability of CPS in 50 cases, that “CPS
software appears to be a reliable investigative tool for law
enforcement”]; People v. Worrell (N.Y. Sup. Ct. 2018) 71 N.Y.S.3d
839, 854, affd. (N.Y. App. Div. 2019) 170 A.D.3d 1048 [noting
how numerous state and federal courts had rejected Fourth
Amendment challenges to the use of CPS in part because courts
“have repeatedly found CPS to be . . . a reliable investigative
tool”].) Lund has cited and our research has discovered no case
that has determined in any context that evidence from CPS or
programs like it is unsupported or unreliable. The apparently
35
uniform acceptance and reliance on CPS evidence supports the
trial court’s decision to admit the CPS evidence here.7
C. Prosecutorial misconduct
Next, Lund argues the prosecutor engaged in repeated,
pervasive misconduct that violated state law and rendered his
trial fundamentally unfair, denying him his federal right to due
process. Lund bases these contentions on four areas of alleged
misconduct by the prosecutor: asking objectionable questions,
engaging in repeated argumentative questioning, asking Lund to
comment on the testimony of other witnesses, and testifying as a
witness. He further asserts the pervasive nature of the
7 Our holdings that the prosecution’s CPS evidence did not
need to undergo Kelly review and was admissible under Sargon
should not be taken to preclude inquiry, through cross-
examination or discovery, into the possible fallibility of the
software. Lund was still entitled to attack the weight of that
evidence, which he might have done by, for example, examining
the source code and pointing out any flaws in its operation. (See,
e.g., U.S. v. Budziak (9th Cir. 2012) 697 F.3d 1105, 1111–1113
[district court abused its discretion in denying discovery into
software like CPS]; U.S. v. Hartman (C.D. Cal., Nov. 24, 2015,
No. SACR 15-00063-JLS) 2015 U.S. Dist. Lexis 197382 at *30–
*41 [allowing discovery into software used to investigate peer-to-
peer networks, including two components of CPS].) Lund
initially sought the CPS source code from the prosecution in
discovery, and the prosecution did not dispute that such evidence
was relevant. The trial court denied his request because it
concluded CPS’s creators were not part of the prosecution team
for the purposes of criminal discovery, but it did so without
prejudice to Lund seeking such evidence via subpoena. There is
no indication in the record that Lund sought the source code via
subpoena.
36
misconduct was prejudicial under both state and federal
standards for reversal. We are not persuaded.
Relevant legal principles
“ ‘ “A prosecutor who uses deceptive or reprehensible
methods to persuade the jury commits misconduct, and such
actions require reversal under the federal Constitution when they
infect the trial with such ‘ “unfairness as to make the resulting
conviction a denial of due process.” ’ ” ’ [Citation.] ‘ “Under state
law, a prosecutor who uses such methods commits misconduct
even when those actions do not result in a fundamentally unfair
trial.” ’ [Citation.] . . . Prosecutorial misconduct can result in
reversal under state law if there was a ‘reasonable likelihood of a
more favorable verdict in the absence of the challenged conduct’
and under federal law if the misconduct was not ‘harmless
beyond a reasonable doubt.’ ” (People v. Rivera (2019) 7 Cal.5th
306, 333–334.) A showing of bad faith is not required to establish
prosecutorial misconduct; “the term prosecutorial ‘misconduct’ is
somewhat of a misnomer to the extent that it suggests a
prosecutor must act with a culpable state of mind. A more apt
description of the transgression is prosecutorial error.” (People v.
Hill (1998) 17 Cal.4th 800, 823, fn. 1.)
“ ‘As a general rule a defendant may not complain on
appeal of prosecutorial misconduct unless in a timely fashion—
and on the same ground—the defendant made an assignment of
misconduct and requested that the jury be admonished to
disregard the impropriety.’ ” (People v. Prieto (2003) 30 Cal.4th
226, 259.) “ ‘[O]nly if an admonition would not have cured the
37
harm is the claim of misconduct preserved for review.’ ” (People
v. Friend (2009) 47 Cal.4th 1, 29.)
Analysis
a. Repeated objectionable questions
We begin with Lund’s argument that the prosecutor
committed misconduct by repeatedly asking him objectionable
questions during cross-examination. Lund contends that it is
misconduct for a prosecutor to purposefully try to elicit
inadmissible testimony, especially after defense counsel has
objected or a trial court has already ruled. He claims the
prosecutor violated this principle by (a) posing five questions that
had previously been asked and answered, (b) asking several
questions that infringed on the attorney-client privilege, and (c)
asking him to discuss other evidence introduced at trial that was
outside his personal knowledge, even though the trial court
repeatedly sustained his objections to all of the questions.
Lund did not raise in the trial court his argument that the
pattern of the prosecutor’s behavior constituted misconduct.
However, he did raise the issue of misconduct with respect to the
questions that he contends intruded on the attorney-client
privilege, and the trial court sua sponte admonished the
prosecutor with regard to some of her questions asking Lund
about other evidence at trial. We therefore conclude the issue
has been sufficiently preserved for review on appeal.
The prosecutor’s questions do not rise to the level of
misconduct. Lund cites cases holding that a prosecutor may
commit misconduct by intentionally seeking to admit
38
inadmissible evidence. (People v. Smithey (1999) 20 Cal.4th 936,
960; People v. Bell (1989) 49 Cal. 3d 502, 532; People v. Johnson
(1978) 77 Cal.App.3d 866, 873–874.) But those cases involved
situations in which prosecutors tried to introduce evidence that
was inadmissible in its entirety, particularly when the trial court
had already so ruled. The principle illustrated by those decisions
is not on point here, where the prosecutor asked Lund the same
questions he had already answered or asked him questions
designed to contrast his testimony with evidence that had
already been admitted. To be sure, a prosecutor may engage in
misconduct by repeatedly asking these types of questions, but
such repetition would have to be far more extensive than the few
questions Lund highlights. (People v. Armstrong (2019) 6 Cal.5th
735, 795–796 [hostile, repetitive, and argumentative cross-
examination of capital murder defendant was not misconduct].)
Lund’s argument based on the attorney-client privilege is
unpersuasive, but for a different reason. The prosecutor asked
Lund whether a computer expert gave Lund any reason to doubt
Datzman’s testimony, whether Lund called that expert at trial,
whether Lund would have called the expert if the expert could
dispute Datzman’s actions, and whether Lund hired the expert
but decided not to call him. Lund objected to these questions on
bases other than attorney-client privilege, and the trial court
sustained the objections though not necessarily for the reasons
Lund asserted. Lund later argued outside the jury’s presence
that the questions were misconduct because they intruded on the
attorney-client privilege. The trial court indicated it had
39
sustained some of Lund’s objections because of the privilege and
said that if the issue recurred it would admonish the jury to
disregard questions to which the court sustained an objection.
But the prosecutor said she was done with that area of
questioning, so the issue did not recur.
Reasonable minds might differ, but the trial court did not
abuse its discretion in sustaining objections to the prosecutor’s
questions, since they perhaps could have called for discussions of
trial strategy. But we find no merit in Lund’s contention that the
questions constituted misconduct by allowing the jury to infer
that he did not call the expert because the expert’s testimony
would not have helped his defense. “[P]rosecutorial comment
upon a defendant’s failure ‘to introduce material evidence or to
call logical witnesses’ is not improper.” (People v. Wash (1993)
6 Cal.4th 215, 263.) Lund points out that there were innocuous,
irrelevant reasons why he might not have called the expert and
that it was his counsel’s ultimate decision whether to call the
expert. This may be true, but it does not negate the prosecutor’s
right to comment on his failure to call the expert or convert any
such comment into an intrusion on attorney-client confidences. If
Lund were correct, then prosecutors would be barred from
commenting on a defendant’s failure to call any witness, which
would give defendants an unfair advantage at trial. As the trial
court indicated, the prosecutor should have waited to make such
commentary in her closing argument rather than introducing the
point through her questioning. But because the failure to call a
defense expert was a proper subject for prosecutorial comment,
40
the improper introduction of the concept in Lund’s cross-
examination was not prejudicial misconduct.
b. Argumentative questioning
Lund’s second misconduct argument is more accurately
characterized as a subset of his first. Lund contends that the
prosecutor engaged in misconduct by continuing to ask
argumentative questions even though the trial court repeatedly
sustained his objections. “An argumentative question is a speech
to the jury masquerading as a question. The questioner is not
seeking to elicit relevant testimony. Often it is apparent that the
questioner does not even expect an answer. The question may,
indeed, be unanswerable.” (People v. Chatman (2006) 38 Cal.4th
344, 384 (Chatman).)
Lund directs us to eight questions and comments he claims
are argumentative. Lund did not object to all of these questions,
did not always raise the issue of argumentative questioning when
he did object, and never raised the issue of misconduct or asked
the trial court to admonish the prosecutor to refrain from such
questions. However, the trial court stopped the questions even
when Lund did not object and twice admonished the prosecutor
sua sponte. We will therefore proceed to examine Lund’s
argument on the merits.
First, after the trial court had sustained his objection to a
question and no question was pending, Lund interjected that he
couldn’t answer the prosecutor’s questions because she would not
let him read the exhibits. The prosecutor responded, “Do you
think you just get to talk when you want to?” Lund did not
41
object, but the trial court sua sponte admonished the prosecutor,
“Don’t do that.”
Second, the prosecutor asked Lund whether he was aware
she had a rebuttal witness coming to contradict his testimony.
Lund objected that the question was argumentative, and the trial
court sustained the objection, on the basis that the question
called for speculation.
Third, after Lund stated that he believed it was unsafe to
have computers in patrol cars, despite the CHP policy mandating
such computers, the prosecutor asked, “So you know better than
all the people that decided that these should go in the CHP patrol
cars?” The trial court sustained Lund’s objection that the
question was argumentative.
Fourth, shortly after asking Lund about a text message he
sent in which he described himself as a “neurosurgeon” in
comparison to his fellow CHP officers, whom he described as
“idiots,” the prosecutor asked Lund, concerning a CHP policy he
said he was not familiar with, “So you’re the neurosurgeon but
you never read the policy?” The court sustained the objection
that the question was argumentative.
Fifth, after Lund testified that other officers used his desk
when he was not in the office, the prosecutor asked, apparently
referring to Lund’s lack of witnesses corroborating his claims,
“Let me guess, you have a witness coming in to say that?” The
trial court sustained Lund’s objection that the question was
argumentative.
42
Sixth, when Lund told the prosecutor that the pornography
in his locker was not his but he did not know who put it there,
the prosecutor rejoined, “Okay. You have no answers?” The trial
court sua sponte ruled that the question was argumentative.
Seventh, when Lund testified that he could not find a
witness to corroborate his claim that another sergeant had a key
to his locker because “[i]t’s been five years ago,” the prosecutor
responded, “But it’s not been five years since day one.” The trial
court sua sponte admonished the prosecutor, “We don’t need to
keep repeating that.”
Finally, after Lund admitted that he had been disciplined
earlier in his career for viewing pornography on his work
computer, the prosecutor asked how many other CHP officers had
gotten in trouble for that. Lund objected that the question was
argumentative, and the trial court sustained the objection on the
grounds of relevance. The prosecutor then asked whether Lund
had a “golden” career, referring to a text message in which he had
said that. The trial court sustained Lund’s objection that the
question had been asked and answered.
Lund contends these questions were argumentative and
demonstrated the prosecutor’s attempt to agitate and belittle
Lund to make the jury not like him. Assuming all of the
questions were argumentative, only the first question, in which
the prosecutor asked whether Lund thought he could talk
whenever he wanted, constituted misconduct. That question was
not designed to elicit information, or even make an argument to
the jury cloaked as a question, but rather aimed to belittle Lund.
43
However, on its own this single instance of misconduct is de
minimis. (People v. Collins (2010) 49 Cal.4th 175, 208.) None of
the other questions, singly or together, rises to the level of
“ ‘deceptive or reprehensible methods’ ” that the doctrine of
prosecutorial misconduct prohibits. (People v. Rivera, supra,
7 Cal.5th at p. 333; People v. Armstrong, supra, 6 Cal.5th at
p. 796 [“Effective and legitimate cross-examination may involve
assertive and even harsh questioning”].) Despite the questions’
sarcastic or biting tone, Lund “identifies no line of questioning,
and the transcript reveals none, that crossed over any boundaries
of fair play or that would have led the jury to decide this case on
anything other than the facts and the law.” (Armstrong, at p. 96.)
Even if these questions did constitute misconduct, we are
satisfied under any standard of prejudice that they did not affect
the outcome of the trial. The trial court sustained Lund’s
objections, dispelling any prejudice. (People v. Fuiava (2012)
53 Cal.4th 622, 687.) Additionally, as discussed above, the
evidence connecting Lund to the devices containing child
pornography was very strong, and the prosecution buttressed it
with the striking correspondence between the CPS data and
Lund’s schedule and movements.
c. Asking whether witnesses were lying
Lund further contends the prosecutor erred by asking him
whether other witnesses were lying. Lund objected to only one of
the first questions in the prosecutor’s series of questions on this
topic, on the grounds that it called for speculation and was
improper. The trial court overruled the objection. We assume
44
the objection of impropriety was equivalent to a request for an
admonition. Because the court overruled the objection, we
conclude any objections to the rest of the questions would have
been futile. We will therefore examine Lund’s argument on the
merits. (People v. Hill, supra, 17 Cal.4th at p. 820.)
Lund relies on People v. Zambrano (2004) 124 Cal.App.4th
228, 242, which held that it is misconduct to ask a defendant if
other witnesses are lying where the question serves no
evidentiary purpose and serves only to berate the defendant and
inflame the passions of the jury.8 But as Chatman, supra,38 Cal.4th at pages 377–384, later made clear, such questions
can serve an evidentiary purpose. Chatman explained that a
“party who testifies to a set of facts contrary to the testimony of
others may be asked to clarify what his position is and give, if he
is able, a reason for the jury to accept his testimony as more
reliable.” (Id. at p. 382.) The court noted that Zambrano had
held a prosecutor committed misconduct by asking the defendant
whether other witnesses were lying, when the defendant did not
know the other witnesses, could not testify about their bias,
interest, and motive to be untruthful, and had already
contradicted their testimony with his own. (Id. at p. 381.) In
Chatman, by contrast, the prosecutor’s questions were proper,
because the defendant “was not asked to opine on whether other
8 Lund also cites U.S. v. Sanchez (9th Cir. 1999) 176 F.3d
1214, 1219, which held that asking whether other witnesses are
lying is improper because witness credibility is a question for the
jury. Chatman, supra, 38 Cal.4th at pages 380, 382, rejected this
principle.
45
witnesses should be believed” but instead “asked to clarify his
own position and whether he had any information about whether
other witnesses had a bias, interest, or motive to be untruthful.”
(Id. at p. 383.) The defendant had relevant personal knowledge
and knew the other witnesses, who were his friends or relatives.
(Ibid.)
The prosecutor’s questions here fall squarely within the
ambit of Chatman. The prosecutor did not ask Lund to opine
generally on whether the jury should believe the other witnesses,
but instead asked more specifically whether he had any reason to
believe the other witnesses would make up lies against him.
Because Lund took the stand and claimed the witnesses against
him were wrong in a way that could only result from deception or
bias, it was fair for the prosecution to explore the basis for Lund’s
belief. This is especially true because Lund worked with many of
the witnesses against him and so might have been able to offer
some specific testimony regarding any biases or improper
motives. Accordingly, the prosecutor’s questions about other
witnesses do not constitute misconduct.
d. Prosecutor testifying as a witness
Lund argues the prosecutor committed misconduct by
testifying as a witness about Duplissey’s correction of his report.
Lund objected to only two questions in the exchange he cites; he
objected that the questions called for hearsay or were leading and
called for speculation. Because Lund neither objected to the
prosecutor’s line of questioning on the ground he now raises nor
requested an admonition, he has forfeited this argument. (People
46
v. Prieto, supra, 30 Cal.4th at p. 259.) Even were we to consider
the argument on the merits, we would reject it because the
prosecutor’s questions of Duplissey about interactions she had
with him are not equivalent to her appearing as a witness.
D. Child pornography videos
Lund’s final argument focuses on the child pornography
videos the prosecution played for the jury. Lund asserts the
prosecution played 50 video segments and argues the prejudicial
effect of this evidence substantially outweighed its probative
value under Evidence Code section 352. He also asserts the
court’s decision to admit the evidence without reviewing it in
advance was arbitrary. He further argues the admission of this
evidence deprived him of a fair trial and violated his right to due
process.
“Under Evidence Code section 352, a trial court may
exclude otherwise relevant evidence when its probative value is
substantially outweighed by concerns of undue prejudice,
confusion, or consumption of time. ‘Evidence is substantially
more prejudicial than probative [citation] if, broadly stated, it
poses an intolerable “risk to the fairness of the proceedings or the
reliability of the outcome.” ’ ” (People v. Riggs (2008) 44 Cal.4th
248, 290.) “In applying this statute we evaluate the ‘risk of
“undue” prejudice, that is, “ ‘evidence which uniquely tends to
evoke an emotional bias against the defendant as an individual
and which has very little effect on the issues,’ ” not the prejudice
“that naturally flows from relevant, highly probative evidence.” ’ ”
(People v. Salcido (2008) 44 Cal.4th 93, 148.) We review for
47
abuse of discretion a trial court’s decision based on Evidence
Code section 352. (Riggs, at p. 290.)
Lund’s assertion that the prosecutor played 50 video
segments is misleading. In response to Lund’s motion in limine
to exclude the child pornography images and videos, the trial
court limited the prosecutor to playing five files from each of the
six devices. The prosecutor played even fewer than that,
displaying only 15 files to the jury. The prosecutor also skipped
ahead quickly in the video files she played for the jury rather
than playing entire files. It appears from the record that many
video segments were very brief, sometimes a matter of seconds.
Lund’s approach of counting multiple portions of a single video
file as separate video segments would penalize the prosecutor for
her apparent attempt to minimize the time spent displaying child
pornography. We therefore reject Lund’s attempt to inflate the
number of videos at issue and will analyze Lund’s arguments
using the number of 15 child pornography files.
Lund first challenges the files as lacking probative value.
He points out that he did not dispute that the files constituted
child pornography, merely his knowing possession of them, so he
contends the child pornography itself was not probative of his
guilt. Lund recognizes the prosecution had to prove that he
possessed child pornography, but he maintains the prosecution
could have used Datzman’s testimony about the videos for this
purpose, so the videos were cumulative. We reject these
arguments because the law is clear that “the prosecution was not
required to accept defense concessions as a sanitized alternative
48
to the full presentation of its case.” (People v. Zambrano (2007)
41 Cal.4th 1082, 1149, disapproved on other grounds by People v.
Doolin (2009) 45 Cal.4th 390.) This is especially true here, where
the videos were not merely circumstantial evidence like a crime
scene in a murder case, but rather constituted direct evidence of
one of the elements of the possession of child pornography charge
against Lund. (Pen. Code., § 311.11, subds. (a), (c)(1); People v.
Holford (2012) 203 Cal.App.4th 155, 171 (Holford).) Lund’s
contention that the videos were unfairly prejudicial by their very
nature therefore misses the mark. The relevant question is not
whether the prosecution was allowed to show child pornography
to the jury, but how much.
On that score, we conclude the trial court did not abuse its
discretion in allowing the prosecution to play portions of a few
files from each of the electronic devices found to contain child
pornography. Holford, supra, 203 Cal.App.4th 155 is instructive.
There, the defendant was charged with possessing a single 25-
minute child pornography video. (Id. at pp. 158–159.) Without
watching the video first, the trial court ruled the entire video was
admissible, and the prosecution played all 25 minutes of it for the
jury. (Id. at pp. 165–166.) On appeal, the court found the video
was highly probative because it proved one element of the crime
and its length and the obviously young age of the minor in it
created an inference that the defendant knowingly possessed it.
(Id. at pp. 171–173 & fn. 8.) The court further found that while
the video was disturbing, “the trial court’s determination that the
probative value of establishing defendant’s knowledge was not
49
‘substantially’ outweighed by a ‘substantial danger’ of prejudice
was not arbitrary, capricious nor patently absurd.” (Id. at
p. 174.) The court stated that while it did not condone the
practice of ruling on the admissibility of the video without
watching it first, the trial court was entitled to rely on an offer of
proof and it apparently did so by accepting the parties’ agreement
that the video was graphic. (Id. at pp. 174–175.) The court also
noted that the trial court’s ruling would have been the same
regardless since it reaffirmed its ruling after watching the video.
(Id. at p. 175.)
In comparison to Holford, where there was a single video at
issue and the prosecution played all of it, the prosecution’s
approach here appears to strike a reasonable balance between
the potential undue prejudice and the prosecution’s desire to
present a compelling case. The prosecutor told the court at the
hearing before Lund’s second trial on Lund’s motion to exclude
the child pornography that she would follow the same practice
that she did at the first, which entailed playing brief portions
from up to the first five videos from each device. The trial court
estimated the prosecutor spent about 8 to 10 minutes in total
displaying photos or videos to the jury in the first trial. The
record does not indicate how much time elapsed during the
playing of the videos at the second trial, but from the court
reporter’s transcription of the prosecutor’s remarks as she was
playing the files, this estimate seems fair. Considering the
electronic devices contained thousands of images and videos,
eight to ten minutes does not seem excessive, especially because
50
that total amount of time was broken up into smaller chunks of
time for each device.
Lund contends the material issue in Holford was whether
the defendant knew the video was child pornography, and
contrasts that with his defense, which turned only on how the
electronic devices came to be found in his locker, desk, and car.
He concludes the videos themselves carried little relevance. But
as we have noted, the nature of the videos and images as child
pornography was still an element of the prosecution’s case, and
the prosecution was not required to sanitize its case before the
jury. The question is one of balance, and Lund does not offer any
argument for how much child pornography the prosecution
should have been allowed to show, aside from saying that the
amount played at trial was too much.
Lund points out that Holford suggested its balancing
analysis might have come out differently had the defendant been
alleged to possess multiple pieces of pornography, like Lund was.
(Holford, supra, 203 Cal.App.4th at p. 171, fn. 7.) Holford stated,
“[I]n a case involving multiple pieces of child pornography, the
probative value of admitting the entirety of a defendant’s
collection may not be any higher than admitting only a few pieces
unless there are other circumstances. Moreover, depending on
the depictions in the collection and other circumstances in the
case, the danger of prejudice resulting from the admission of an
entire collection could substantially outweigh the probative
value, particularly since admitting the extra pieces could have
51
very little effect on the issues given the charging rules for
possession of child pornography in California.” (Ibid.)
These statements are obviously dicta, but they also do not
reflect the current state of the law. When Holford was decided in
2012, as the court stated, the possession of multiple pieces of
child pornography was chargeable as only a single criminal
offense. (Holford, supra, 203 Cal.App.4th at p. 171, fn. 7.) In
2013, the Legislature amended Penal Code section 311.11 to
create the separate offense of possession of over 600 child
pornography images, more than 10 of which involve a minor
under the age of 12, with one video equal to 50 images.
(Stats. 2013, ch. 777, § 3.) Now, when a prosecutor is trying to
prove such an offense, the display of multiple child pornography
images or videos has significant probative value to show a
defendant possessed 600 images or 12 videos. Even so, Evidence
Code section 352 still plays a role in such cases. The statute does
not set any minimum length for a video to qualify as 50 images,
so the new offense does not mean a prosecutor can necessarily
play the entirety of 12 videos like the 25-minute video at issue in
Holford. Conversely, a trial court does not necessarily abuse its
discretion in allowing a prosecutor to play more than 12 videos,
like the prosecutor here, even when the nature of the videos as
child pornography is not disputed. The balancing of probative
value against the risk of undue prejudice under Evidence Code
section 352 remans an issue for the trial court’s sound discretion.
And under the circumstances here, where the prosecutor
displayed one image and portions of 15 videos for what appears to
52
be a total of approximately eight to ten minutes, we cannot find
that the court abused its discretion in striking the balance as it
did.
Lund finally argues the trial court abdicated its role as the
gatekeeper of evidence by not previewing any of the videos that
the prosecution played for the jury, just like the trial court in
Holford, supra, 203 Cal.App.4th at pages 174–175. He contends
the trial court’s admission of the videos could only have been
arbitrary because the prosecution did not even make an offer of
proof as to the number of videos, how long the videos were, or the
file names of the videos. This argument misconstrues the record.
As noted above, the prosecution told the trial court at a pretrial
hearing before the second trial that she would limit herself to the
first five files from each device, just as she had in the first trial.
The record shows the prosecutor did just that. The same judge
presided over both trials, and the trials took place only a few
months apart, so the trial court was well aware of the nature of
the child pornography before the prosecutor played it. Lund
quibbles over whether this procedure technically qualified as an
offer of proof as to the content of the files, but it amounts to the
same thing. The trial court specifically noted that the
prosecutor’s approach in the first trial was not excessive and she
ultimately displayed fewer files to the jury than the trial court
had authorized. The second trial was similar, with the
prosecution displaying 15 videos and one image to the jury, not
the full 30 files authorized by the trial court’s ruling. Because
the trial court had already seen the videos that would be played,
53
its decision was “an informed one and not ‘ “a shot in the dark.” ’ ”
(Id. at p. 174.) The trial court did not abdicate its role.9
III. DISPOSITION
The judgment is affirmed.
BROWN, J.
WE CONCUR:
POLLAK, P. J.
STREETER, J.
People v. Lund (A157205)
9The only error that Lund has shown in his trial was the
de minimis misconduct of one improper statement by the
prosecutor. As a result, we need not consider Lund’s contention
that the cumulative effects of multiple errors in his trial made it
fundamentally unfair in violation of due process.
54
Trial Court: Solano County Superior Court
Trial Judge: Hon. Daniel J. Healy
Counsel:
Law Offices of Beles & Beles, Robert J. Beles, Joseph L.Ryan, for
Defendant and Appellant.
Xavier Becerra, Attorney General, Lance E. Winters, Chief
Assistant Attorney General, Julie L. Garland, Assistant Attorney
General, A. Natasha Cortina, Annie Featherman Fraser, Deputy
Attorneys General, for Plaintiff and Respondent.
55