Colwell v. Sig Sauer, Inc.

24-2724(L)
Colwell v. Sig Sauer, Inc.
UNITED STATES COURT OF APPEALS
FOR THE SECOND CIRCUIT
August Term, 2025
(Argued: October 22, 2025 Decided: May 28, 2026)
Docket Nos. 24-2724(L), 24-3047(CON)
MICHAEL COLWELL AND JULIA COLWELL,
Plaintiffs-Appellants,
— v. —
SIG SAUER, INC.,
Defendant-Appellee.*
B e f o r e:
LEVAL, LYNCH, and SULLIVAN, Circuit Judges.
__________________
* The Clerk of Court is respectfully directed to amend the official caption in this
case to conform with the caption above.

-- 1 of 52 --

Michael and Julia Colwell appeal a summary judgment decision for
defendant on their strict products liability and negligence claims. The couple
sued firearm manufacturer Sig Sauer after Michael Colwell, a police sergeant,
was injured when his P320 pistol inadvertently discharged during a training
exercise. The Colwells sought to introduce testimony from two experts that the
P320 was defectively designed because, among other flaws, it lacked any external
safety, and that the lack of an external safety caused Colwell’s injury. The district
court (Brenda K. Sannes, C.J.) excluded the experts’ causation opinions because
they could not explain how the accident happened and therefore could not show
how an external safety would have prevented Colwell’s injury. The district court
then granted summary judgment for Sig Sauer, concluding that New York law
required expert testimony to establish proximate causation, as the issue was
otherwise too complex for a jury.
We conclude that the district court was within its discretion to exclude the
experts’ causation opinions. The experts largely failed to address the
circumstances of Colwell’s accident and instead based their opinions as to
causation primarily on the fact that a gun with an external safety is less likely to
discharge inadvertently than one without. They therefore lacked a reliable
foundation for their conclusion as to causation. Nonetheless, we conclude that
the district court erred in granting summary judgment for Sig Sauer. New York
law does not require expert testimony to establish proximate causation where
jurors can use their own judgment to assess the characteristics of the allegedly
defective product and a witness’s account of the accident. Here, the jury could
rely on the experts’ unchallenged design-defect opinions and its own
consideration of the P320 and Colwell’s holster to determine whether an external
safety would have prevented the accident. We also decline to affirm the district
court’s judgment based on New York’s optional equipment doctrine, as Sig Sauer
failed to show that there exist normal circumstances of use for the P320 that do
not implicate its allegedly defective design. We therefore VACATE the district
court’s judgment and REMAND for further proceedings.
Judge Sullivan dissents in a separate opinion.
__________________
ROBERT W. ZIMMERMAN, Saltz Mongeluzzi Bendesky, P.C.,
Philadelphia, PA for Plaintiffs-Appellants.
2

-- 2 of 52 --

ERIC G. LASKER, Hollingsworth LLP, Washington, D.C. (Brett F.
Clements, Hollingsworth LLP, Washington, D.C., Robert L.
Joyce, Littleton Joyce Ughetta & Kelly LLP, Purchase, NY, on
the brief) for Defendant-Appellee.
Lee Mickus, Evans Fears Schuttert McNulty Mickus, Littleton, CO,
for Amicus Curiae Lawyers for Civil Justice, in support of
Defendant-Appellee.
GERARD E. LYNCH, Circuit Judge:
Michael and Julia Colwell sued firearm manufacturer Sig Sauer, asserting
strict products liability and negligence claims, after Michael Colwell (“Colwell”),
a police sergeant, was injured when his P320 pistol inadvertently discharged
during a training exercise. While Colwell does not know what caused the trigger
to actuate (i.e., what forced it to move), he testified that the gun had been
holstered at the time it discharged. The Colwells sought to introduce expert
testimony from James Tertin, a gunsmith, and Dr. William Vigilante, a human
factors and ergonomics expert, that the P320 was defectively designed because,
among other flaws, it lacked an external safety, and that the lack of an external
safety caused Colwell’s injury. The district court (Brenda K. Sannes, C.J.)
excluded the experts’ opinions, insofar as they addressed causation, because the
3

-- 3 of 52 --

experts could not explain how the accident happened and thus could not show
how an external safety would have prevented Colwell’s injury. It then granted
summary judgment for Sig Sauer, concluding that New York law required expert
testimony to establish proximate causation, as the issue was otherwise too
complex for a jury.
We conclude that the district court was within its discretion to exclude the
passages in the experts’ opinions dealing with causation. The experts largely
failed to consider the circumstances of Colwell’s accident and instead based their
causation opinions primarily on the fact that a gun with an external safety is less
likely to discharge inadvertently than one without. They therefore lacked a
reliable foundation for their conclusions that the absence of such a safety
contributed to causing Colwell’s injury. Nonetheless, we find that the district
court erred in granting summary judgment for Sig Sauer. New York law does not
require expert testimony to establish proximate causation in all circumstances,
including in cases where jurors can use their own judgment to assess the
characteristics of the defective product (as described in the evidence, including
expert testimony about the product and its design flaws) and a witness’s
testimony about the circumstances of the accident. Here, the jury could rely on
4

-- 4 of 52 --

the experts’ unchallenged design-defect opinions, Colwell’s account of the
circumstances, and its own examination of the P320 and Colwell’s holster to
determine whether an external safety would have prevented the accident. We
separately decline to affirm the district court’s judgment based on New York’s
optional equipment doctrine, as Sig Sauer failed to show that there exist normal
circumstances of use for the P320 that do not implicate its allegedly defective
design. We therefore vacate the district court’s judgment and remand for further
proceedings.
BACKGROUND
Sergeant Michael Colwell, a veteran officer with the Troy Police
Department, was injured when his department-issued Sig Sauer P320 pistol
inadvertently discharged during a training exercise, sending a bullet into his
right leg. Colwell does not know what caused the trigger to actuate. He testified
that the pistol went off after he had holstered the gun on his right side and as he
reached across his body to retrieve a simulation taser holstered on his left. In
other words, his hand was not on the gun, let alone the trigger, when the gun
discharged. Colwell’s recollection was challenged by the emergency responders’
report and other documentation of the accident that state that the pistol
5

-- 5 of 52 --

discharged as Colwell was still in the process of holstering it.
Colwell and his wife Julia sued Sig Sauer for, among other claims, strict
products liability and negligence, alleging that the P320 was defectively designed
and that this design allowed an unknown foreign object to actuate the P320’s
trigger in its holster and cause it to discharge.1 In support of that theory, the
Colwells sought to introduce expert testimony from Tertin and Dr. Vigilante,
who have served as experts in numerous products liability cases involving the
P320 across various jurisdictions.
The experts opined that the P320 was defectively designed because its
short and lightweight trigger, insufficient internal safeties, and lack of any
external safety created an unreasonable risk of unintentional discharge. They also
opined that this defective design caused Colwell’s injury and that an external
safety would have prevented it. Sig Sauer introduced its own expert, Derek
Watkins, who contended that the P320 was sufficiently safe without any external
safeties and that Colwell’s accident was caused by his own mishandling of the
pistol.
1 The Colwells and their experts do not claim that the gun spontaneously fired.
There is no dispute that the P320 cannot discharge without its trigger being
depressed.
6

-- 6 of 52 --

Sig Sauer moved to exclude both Tertin’s and Dr. Vigilante’s causation
opinions and for summary judgment. The district court granted all three motions.
We summarize the expert reports and the district court’s decision below.
A. Tertin’s Opinion
Tertin’s report sought to “analyze the adequacy and sufficiency” of the
P320’s safety mechanisms. App’x 190. To do so, it first provided an overview of
the P320’s design as a single-action striker-fired pistol. Pistols are typically
categorized as “striker-fired” or “hammer-fired,” and as “single-action” or
“double-action.” A striker-fired pistol employs a tensioned “striker” (as opposed
to a hammer) as its firing mechanism. Id. at 191. By pulling the trigger, a user
releases a tensioned firing pin so that it can impact the cartridge (i.e., the
ammunition).
Whether a striker pistol is “single-action” or “double-action” depends on
whether its trigger serves the “single” function of releasing an already tensioned
firing pin or the “double” function of cocking and releasing it. Cocking a gun
readies it for firing by placing the striker pin under tension and positioning a
cartridge in the chamber. The P320 is initially cocked manually, but after the user
7

-- 7 of 52 --

fires the first cartridge, the gun’s recoil automatically cocks it for the next round.2
Because the trigger on a single-action gun does less work than one on a
double-action gun, it tends to be more sensitive, requiring a shorter and less
forceful pull. For example, Tertin found that the P320’s trigger pull is four times
shorter than that of a Glock, a double-action pistol that is one of Sig Sauer’s
primary competitors in the striker-fired pistol market. It is therefore no surprise
that Sig Sauer’s expert described the P320 as having a “short displacement trigger
system.” Id. at 193.
Tertin then explained that to avoid inadvertent discharges, guns rely on
external and internal safety mechanisms. Colwell’s P320 did not have any
external safeties and instead employed a system of two internal safeties. The first
internal safety is the striker-sear engagement, a connection that holds back a
cocked striker pin until the trigger is pulled. But the connection is very small,
“only .040” of an inch, “less than the thickness of a dime.” Id. at 197. Accordingly,
the sear can disengage by moving only “one-twenty-fifth of one inch,” which can
happen by “vibration, jostling, or contact with the P320.” Id. As a second internal
safety, the P320 has a striker safety lock. The lock serves to “block the striker pin
2 Colwell testified that he was required to carry his P320 cocked at all times.
8

-- 8 of 52 --

from moving forward” if “the striker-sear engagement is lost.” Id. at 198. But the
P320 does not require a full trigger pull to disengage the safety lock. Tertin found
that the lock can be disengaged with a trigger pull of less than one-tenth of an
inch under 1.5 pounds of pressure.3 The independent shortcomings of those two
internal safeties, coupled with the lack of any external safety, he concluded,
“makes it far too easy for the trigger to be accidentally actuated.” Id. at 206.
Tertin then identified external safeties that could have been incorporated
into the P320’s design to help prevent inadvertent discharges. An external safety
prevents discharges by blocking the trigger until the user is ready to fire. The Sig
Sauer P320 is unique among single-action guns in that it is sold without any
external safety. Two external safeties are relevant here: a manual thumb safety
and a tabbed trigger safety.4 First, a manual thumb safety is a switch on the side
of the pistol that can be flipped on or off. Only when the safety is flipped off will
the gun fire; the trigger remains locked when the switch is in the on position. Sig
3 For comparison, Sig Sauer’s expert measured the average force required for a
full trigger pull to be 6.425 pounds.
4 While Tertin’s report uses the phrase “manual safety” or “external manual
safety” to describe the two safeties discussed below, see, e.g., App’x 200, we refer
to both simply as “external safeties,” since the tabbed trigger safety operates
passively and does not require any “manual” engagement on the user’s part.
9

-- 9 of 52 --

Sauer offers a model of the P320 with a manual thumb safety, but the Troy Police
Department did not purchase that version.
Second, a trigger safety refers to a small tab placed in the center of the
trigger that must first be depressed before the entire trigger can move. It thus
substantially limits the surface area of the trigger where a contact can effectuate a
discharge. The tabbed trigger is “the most widely used type of safety for striker
fired guns” like the P320. Id. at 202. Tertin recorded videos showing that, unlike
the trigger on a P320, the tabbed trigger on a Glock will move only when
pressure is applied to its center and will not budge when pressed on its edges.
According to Tertin, either a thumb safety or a tabbed trigger could have been
added to the P320’s design with minimal cost or functional impact.5
Only at the conclusion of his report did Tertin address causation. He
claimed that the P320 — whether its trigger was touched by Colwell’s finger or a
foreign object — “most likely would not have discharged” if it were a double-
action pistol or had an external safety and that, as a result, the defective design
was a “proximate cause” of Colwell’s injury. Id. at 206–07. He did not conduct
5 Separately, Tertin’s report also discusses a third type of external safety – a grip
safety – but the Colwells did not devote any discussion to it in their briefing
either before the district court or on appeal.
10

-- 10 of 52 --

any testing specific to the circumstances of Colwell’s accident.
B. Dr. Vigilante’s Opinion
Dr. Vigilante’s report largely follows Tertin’s. He similarly explained that
the P320 functions like a single-action pistol and has a trigger “more sensitive”
than its competitors, making it more “susceptible to unintentional discharge than
other striker-fired guns.” Id. at 120.
He then examined the intended uses and user populations of the P320,
noting that law enforcement was a key customer base and that use by law
enforcement officers involves “dynamic movement” that increases the risk of an
unintentional discharge “due to inadvertent contact with the trigger, jarring, or
dropping of the firearm.” Id. at 123. Dr. Vigilante also described videos
documenting several instances similar to Colwell’s where P320s inadvertently
discharged while holstered and recorded his own videos showing how a side-
pull could actuate the P320’s trigger but not a tabbed trigger. He therefore
concluded that it was “reasonably foreseeable to Sig Sauer” that there was a “risk
of unintended discharge associated with an object inadvertently and/or
unknowingly contacting the trigger” and that an external safety like a tabbed
trigger or thumb safety would have reduced that risk. Id. at 125.
11

-- 11 of 52 --

Like Tertin, Dr. Vigilante considered causation only at the very end of his
report, conclusorily stating that “Sig Sauer’s failure to integrate” an external
safety into the P320 “was a cause” of Colwell’s accident and that the accidental
discharge “most likely” would not have occurred had the P320 been equipped
with a tabbed trigger. Id. at 137.
C. The District Court’s Opinion
The district court granted Sig Sauer’s motion to exclude Tertin’s and
Dr. Vigilante’s causation opinions under Federal Rule of Evidence 702. The court
appeared to accept their design-defect opinions as unchallenged and relied on
Tertin’s report to describe the P320’s design. But it found that their causation
opinions were not based on sufficient facts or reliable methods. It noted that both
experts had limited knowledge of Colwell’s accident and failed to point to any
facts specific to the event in justifying their opinions. Neither, for example, had
any idea what foreign object may have made contact with the P320’s trigger or
where on the trigger it would have made that contact. Instead, their theory of
causation was simply predicated on the fact that a gun with an external safety is
generally safer and less likely to discharge inadvertently than one without.
Having excluded the experts’ causation opinions, the district court granted
12

-- 12 of 52 --

summary judgment for Sig Sauer on the strict products liability and negligence
claims. It held that, under New York law, the Colwells could not establish
causation without expert testimony because the issue of causation here was too
complex and not “within the understanding of an ordinary juror.” Id. at 18. The
parties stipulated to a dismissal of the Colwells’ other claims, and the Colwells
filed this appeal.
DISCUSSION
I. Standard of Review
“When a party challenges the district court‘s evidentiary rulings
underlying a grant of summary judgment, we undertake a two-step inquiry.
First, we review the trial court's evidentiary rulings, which define the summary
judgment record. . . . Second, after we have defined the record,” we address
summary judgment. Bustamante v. KIND LLC, 100 F.4th 419, 425 (2d Cir. 2024)
(internal quotation marks omitted).
We review a decision to exclude expert testimony under a “highly
deferential abuse of discretion standard.” In re Mirena IUS Levonorgestrel-Related
Products Liability Litigation (No. II), 982 F.3d 113, 122 (2d Cir. 2020) (internal
quotation marks omitted). Under Federal Rule of Evidence 702, for expert
13

-- 13 of 52 --

testimony to be admissible, the proponent must demonstrate by a preponderance
of the evidence that the testimony is relevant and reliable. See id. at 122–23; United
States v. Williams, 506 F.3d 151, 160 (2d Cir. 2007). In assessing reliability, a court
considers whether the testimony is “based on sufficient facts or data”; “is the
product of reliable principles and methods”; and “reflects a reliable application
of the principles and methods to the facts of the case.” Fed. R. Evid. 702(b)–(d).
While the method for determining reliability “depends upon the particular
circumstances of the particular case at issue,” Kumho Tire Co. v. Carmichael, 526
U.S. 137, 150 (1999), the abuse-of-discretion standard “applies as much to the trial
court’s decisions about how to determine reliability as to its ultimate conclusion,”
Amorgianos v. National Railroad Passenger Corp., 303 F.3d 256, 265 (2d Cir. 2002)
(emphasis in original).
We review a grant of summary judgment de novo, resolving all ambiguities
and drawing all permissible inferences in favor of the nonmoving party.
Bustamante, 100 F.4th at 432. Summary judgment is appropriate where the
“movant shows that there is no genuine dispute as to any material fact and the
movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). “[W]hen
the burden of proof at trial would fall on the nonmoving party, it ordinarily is
14

-- 14 of 52 --

sufficient for the movant to point to an absence of evidence on an essential
element of the nonmovant’s claim.” Bustamante, 100 F.4th at 432 (internal
quotation marks omitted).
II. Admissibility of Causation Opinions
The district court was well within its discretion in excluding Tertin’s and
Dr. Vigilante’s causation opinions.6 Rather than assess the particular
circumstances of Colwell’s accident, they equated the purported existence of a
design defect to causation. Correcting that defect by adding an external safety
would have made the gun generally safer; a safer gun would have been less
likely to discharge inadvertently; and it follows, according to Tertin and Dr.
Vigilante, that the defect must have caused Colwell’s injury. The district court
6 For utmost clarity, we understand the excluded “causation opinions” to
encompass the following portions of the experts’ opinions. Tertin’s “causation
opinion” encompasses his conclusions that “[t]he defective design of the P320
was a proximate cause of Plaintiff’s accident” and that “[i]f Sig Sauer had not
made the only single-action pistol on the market without any manual safeties,
Plaintiff’s accident most likely would not have occurred.” App’x 207. Dr.
Vigilante’s “causation opinion” refers to his statements that “[h]ad Sig Sauer
integrated a tab trigger safety into the design of the Sig P320, the subject
unintentional discharge would most likely not have occurred and Michael
Colwell would not have been injured,” App’x 136; see also id. at 137 (same), and
that “Sig Sauer’s failure to integrate an external manual safety into the design of
the P320 . . . was a cause of the subject unintentional discharge and Michael
Colwell’s injury,” id at 137; see also id. at 136 (similar).
15

-- 15 of 52 --

did not abuse its discretion in finding that this conclusion, insufficiently tethered
to the facts of Colwell’s accident, did not qualify as a reliable opinion as to
causation.
Both experts made limited efforts to understand the facts of Colwell’s
accident. Neither, for example, inspected Colwell’s holster. Nor did they have
any opinion on what foreign object caused Colwell’s pistol to discharge, where
the object contacted the trigger, or what amount of force the object applied. All
Dr. Vigilante could say was that, based solely on Colwell’s testimony (the
credibility of which he was of course in no position to evaluate), “there was not a
full placement of the finger on the front face of the trigger to pull back.” App’x
503 at 20:22–24. Tertin did not even get that far; he had no opinion as to whether
Colwell’s hand was on the pistol when it went off; he “just kn[e]w that [the
pistol] discharged.” Id. at 421 at 11:15. Their reports provide no insight into how
the accident occurred.
Lacking actual knowledge of Colwell’s accident, they resorted to
probability. Mr. Tertin, for instance, cited the “practical function” of a tabbed
trigger to argue that the P320’s lack of such a mechanism caused Colwell’s injury,
noting that the tabbed trigger is “one more step for safety.” Id. at 423 at 18:22–23,
16

-- 16 of 52 --

425 at 26:7. But he offered nothing to show that such an additional step would
have played a preventative role in these circumstances. Nor did he have any basis
to conclude that a manual thumb safety would have prevented the accident here
because, as he testified at his deposition, he did not know whether Colwell
would have engaged a thumb safety had his pistol been equipped with one. Dr.
Vigilante similarly could not answer whether a tabbed trigger would have
prevented Colwell’s pistol from discharging; he could only say that that was a
tabbed trigger’s “intent and purpose.” Id. at 509 at 42:17. In short, as the Third
Circuit aptly described similar testimony from Tertin and Dr. Vigilante, they can
only show that “the P320’s design could have caused an accident, but not whether
it did cause this accident.” Slatowski v. Sig Sauer, Inc., 148 F.4th 132, 138 (3d Cir.
2025) (emphasis in original).
Tertin and Dr. Vigilante were not without options to test their theory. They
could have tried to recreate the circumstances under which Colwell was injured.
For example, they could have assessed the impact of Colwell’s asserted cross-
body motion on a holstered P320 or analyzed the accessibility of a holstered
P320’s trigger to a foreign object. But they made no such inquiries. Even if such
investigations might not have proven definitively how Colwell’s accident
17

-- 17 of 52 --

occurred, they nonetheless could have laid a stronger factual foundation for
assessing the role an external safety would have played in it. Without such
incident-specific testing, however, the district court was entitled to find that their
opinions on proximate causation rested on “simply too great an analytical gap
between the data and the opinion proffered.” General Electric Co. v. Joiner, 522 U.S.
136, 146 (1997); see also Brooks v. Outboard Marine Corp., 234 F.3d 89, 92 (2d Cir.
2000) (affirming exclusion of expert testimony where the expert ”did not know
precisely what happened” and failed to “reconstruct the accident and test his
theory”).7
The Colwells try to avoid that conclusion by appealing to common sense. It
is, they contend, highly unlikely that, had Colwell’s pistol been equipped with a
tabbed trigger, “[a] foreign object could have entered the trigger guard, turned 90
7 We do not intend to impose a requirement that an expert must always
reconstruct an accident in order to offer a causation opinion, and we need not
probe the boundaries of when such testing is necessary. We merely point out
that, because in this case Tertin and Dr. Vigilante have a very limited
understanding of how the accident happened, incident-specific testing was one
way they might have developed knowledge that could support a reliable
causation opinion. In any event, we do not rule that, even in this case, the
opinions in question were inadmissible as a matter of law; we hold only that the
district court’s decision was within the range of reasonable conclusions and thus
was not an abuse of discretion.
18

-- 18 of 52 --

degrees flush across the face of the trigger, and pulled the trigger rearward.”
Appellants’ Br. 51. As we explain below, a jury may find that to be the case, but
an opinion based on common sense does not qualify as expert testimony under
Rule 702. As various courts have recognized, a jury does not need the testimony
of experts where its own common sense will suffice. See, e.g., In re Mirena IUD
Products Liability Litigation, 169 F. Supp. 3d 396, 484 (S.D.N.Y. 2016); United States
v. Maxwell, No. 20-CR-330, 2021 WL 5283951, at *4 (S.D.N.Y. Nov. 11, 2021);
Variblend Dual Dispensing Systems LLC v. Crystal International (Grp.) Inc., No. 18-
CV-10758, 2022 WL 17156550, at *13 (S.D.N.Y. Nov. 22, 2022).
We therefore conclude that the district court did not abuse its discretion in
excluding Tertin and Dr. Vigilante’s causation opinions.
III. Summary Judgment
After excluding Tertin’s and Dr. Vigilante’s causation opinions, the district
court granted summary judgment for Sig Sauer because it found that expert
testimony was necessary to address proximate causation. We disagree. Once
equipped with an understanding of the P320, its alleged design defects, and
potentially safer alternative designs — all portions of the experts’ testimony that
19

-- 19 of 52 --

Sig Sauer did not challenge8 — a reasonable jury could find by a preponderance
8 In its briefing before the district court, Sig Sauer explicitly disavowed a
challenge to the experts’ design-defect opinions. As it wrote in its reply briefs in
support of its motions to exclude: “Plaintiffs spend the overwhelming majority of
their opposition discussing why they believe the design of the P320 model pistol
is defective. But that is not the aspect of [the expert] opinions that Sig Sauer
challenged in its Rule 702 motion, which focused narrowly on the unreliability of
Vigilante’s and Tertin’s causation opinions.” Dist. Ct. Dkt. 54 at 4; see also Dist. Ct.
Dkt. 55 at 4 (same). On appeal, Sig Sauer responds to this point by claiming to
have challenged the experts’ “general causation” opinions, a term that appears in
neither expert’s report nor in Sig Sauer’s briefing before the district court.
General causation asks whether an allegedly defective “product is capable
of causing the type of injuries from which plaintiffs claim to suffer.” In re Mirena
II, 982 F.3d at 121. It is typically litigated in cases involving products like
pharmaceuticals where it is disputed whether the defendant’s product can even
cause the type of injury the plaintiff suffered. See, e.g., id. But here no one
disputes that the P320 caused (let alone was capable of causing) Colwell’s injury.
Sig Sauer instead argues without relevant citation that general causation here
requires a showing that “P320 pistols — when handled appropriately — are more
likely than pistols equipped with tabbed triggers to accidentally discharge when
their triggers are pulled.” Appellee’s Br. 32. But that argument ignores that “the
availability of a safer design” is a factor used to determine whether a product’s
design is defective. Voss v. Black & Decker Manufacturing Co., 59 N.Y.2d 102, 109
(1983). Sig Sauer cannot revive an issue it waived by now reclassifying it as part
of the causation inquiry.
Finally, any challenge to the admissibility of Tertin’s and Dr. Vigilante’s
design defect opinions, even if not waived, would still fail because the experts
reliably opined that the P320’s trigger is more likely to effect an inadvertent
discharge than a tabbed trigger based on the available surface area for actuation.
See supra pp. 10–12; see also Davis v. Sig Sauer, Inc., 126 F.4th 1213, 1227–28 (6th
Cir. 2025) (finding that similar testimony from Tertin and Dr. Vigilante reliably
demonstrated “how the P320 is more likely than other commercially available
firearms to fire inadvertently from jostling, side-pulls, and grazes”). Sig Sauer, of
course, will have the opportunity at trial to argue, based on cross-examination, its
20

-- 20 of 52 --

of the evidence that whatever caused Colwell’s P320 to discharge would not have
caused it to do so had it been equipped with an external safety.
While a plaintiff must satisfy proximate causation to succeed in a products
liability case, New York law is clear that “[e]xpert testimony with reference to
proximate causation is not always required.”9 Voss, 59 N.Y.2d at 110. Instead, the
necessity of expert testimony turns on the facts of the case. Some categories of
products liability claims, such as those involving toxic exposures or
pharmaceuticals, will almost always require expert testimony as to causation
because those issues are far “beyond the sphere of the ordinary juryman.” Fane v.
Zimmer, Inc., 927 F.2d 124, 131 (2d Cir. 1991). But in many cases, other evidence
can support a theory of causation such that expert testimony, while potentially
helpful, is not mandatory. A jury may be able to assess causation based on a
“consideration of the characteristics” of the allegedly defective product and a
witness’s “description of how the accident happened.” Voss, 59 N.Y.2d at 110–11.
own contrary expert, or any other evidence in the record, that Tertin’s and Dr.
Vigilante’s opinions are wrong and the P320 is reasonably safe as designed.
9 Our analysis here applies equally to the Colwells’ negligence claim. Although
“[p]roximate cause in a products liability case serves a somewhat different role
than in a case sounding in negligence,” Voss, N.Y.2d at 110, those differences do
not implicate whether a jury requires expert testimony here.
21

-- 21 of 52 --

Sometimes expert testimony is not necessary because the causation
question will be obvious to a jury. For example, courts applying New York law
have held that a bus driver claiming that his bus seat was defective due to an
insufficient weight limit did not need expert testimony to show that exceeding
the weight limit could cause injury, see Yaccarino v. Motor Coach Industries, Inc.,
No. 03-CV-4527, 2006 WL 3257220, at *3 (E.D.N.Y. Nov. 9, 2006), and that a
runner who tripped did not need an expert to show that her allegedly defective
shoelaces caused her fall, see Faryniarz v. Nike Inc., No. 00-CV-2623, 2002 WL
530997, at *2 (S.D.N.Y. Apr. 8, 2002).
But even when facing a question of proximate causation that is not obvious
without pinpointed expert testimony, the jury is not helpless. It may still be able
to resolve the issue by applying its own common sense to the evidence before it,
including eyewitness testimony about the incident and expert testimony about
the design and operation of the product in question. Voss illustrates that. There,
the plaintiff was injured while using a circular power saw. 59 N.Y.2d at 104.
When the plaintiff cut into a knot in the wood, the saw launched into the air, and,
as it landed, severed part of the plaintiff’s thumb. Id. He claimed that the saw was
defective because its retractable blade guard left an unnecessarily large amount
22

-- 22 of 52 --

of the blade exposed and submitted expert testimony explaining that the device
could have been designed to be safer by extending the blade guard. Id. at 104–05.
The guard as designed had left 53 degrees of the blade exposed, and the expert
opined that only 45 degrees should have been exposed under the relevant
industry standards. Id. at 105. The expert, however, did not testify specifically as
to whether that eight-degree adjustment in the blade guard would have
prevented the accident, i.e., the proximate causation question. The New York
Court of Appeals held that the trial court should have submitted the issue to the
jury, which could take the design-defect testimony and plaintiff’s account of the
event “to decide whether the defective design was a substantial factor causing
[the] plaintiff’s injury.” Id. at 110.
As in Voss, the Colwells and their experts have presented enough of a
factual basis upon which a reasonable jury could find that the P320’s allegedly
defective design caused it to discharge inadvertently in Colwell’s holster.
Assuming that the jury credits Colwell’s account of the accident and believes that
the gun had been holstered when it discharged (as we must at summary
judgment), a jury could make a series of reasonable determinations that, taken
together, would allow it to find causation satisfied. First, based on Tertin and Dr.
23

-- 23 of 52 --

Vigilante’s admissible design-defect opinions, it could conclude that the P320 has
a sensitive trigger, that its internal safeties can be disengaged easily, and that, as
a result, it is more prone than certain other firearms to discharging inadvertently.
It could then conclude, again based on the experts’ opinions, that a tabbed trigger
reduces the likelihood of inadvertent discharges by limiting the surface area
where the trigger can be actuated to only its center. Finally, based on its own
inspection of Colwell’s holster, a P320, and a gun equipped with a tabbed trigger,
the jury could use its common sense to determine whether any foreign object
shown to be in the vicinity of the incident would have been likely to actuate the
P320’s trigger but not a tabbed one. That is a question that the jury can resolve
based on its own observations and understanding of the facts of what happened
as it finds them to be.10
To be sure, the limited testimony in the present record suggests that the
jury may not be able to determine what foreign object caused the trigger to
actuate, or discern where any such object hit and with what force. And it is
10 A jury could also reject Colwell’s account of the incident while still finding in
his favor on causation. For example, a jury could determine that Colwell’s P320
discharged when his finger brushed against the side of the trigger as he was
holstering it but that the same graze would not have caused a tabbed trigger to
actuate.
24

-- 24 of 52 --

certainly possible that whatever caused the gun to discharge could have done so
even if the gun had a tabbed trigger. But the Colwells need not prove certainty —
they need only show by a preponderance of the evidence that the P320’s
allegedly defective design was a substantial factor in causing Colwell’s injury. On
the record before us, a jury could find just that. While more uncertainty appears
to lie here as to how the accident happened than in Voss, a point that carried
substantial weight in the district court’s analysis, we believe that this uncertainty
does not preclude the Colwells from presenting their case to the jury.11
We limit the above analysis to only the tabbed trigger safety and not the
manual thumb safety. Unlike the tabbed trigger, which operates passively, the
manual thumb safety must be engaged intentionally by the gun user. Despite
discussing the thumb safety in their expert reports and briefing, neither the
Colwells nor their experts provide any evidence that Colwell at the time of the
accident would have had a thumb safety engaged. They point to no expert
testimony as to best practices regarding utilization of the thumb safety or
11 As should go without saying, we hold only that a reasonable jury could reach a
verdict for the Colwells. A reasonable jury could also reject some of the
inferences above, or could find Colwell’s account less than fully credible. But
those are all questions for a jury to decide.
25

-- 25 of 52 --

testimony by Colwell as to his practices or tendencies that would allow the jury
to make such a finding. The Colwells have therefore failed to raise an issue of
material fact as to whether a thumb safety would have prevented the accident.
Sig Sauer’s arguments as to why summary judgment is appropriate fail to
persuade. It first contends, pointing to its expert’s testimony, that a tabbed trigger
is intended only to prevent a gun with an unbalanced trigger mechanism from
discharging when dropped. But the Colwells’ experts have reliably opined
otherwise, creating an issue of material fact for the jury to resolve. It then argues
that the issue of causation is too complex for a jury to decide as the jury would
not be able to understand the amount of force required to actuate a tabbed trigger
versus a non-tabbed trigger. But, as explained above, Tertin and Dr. Vigilante can
testify to the amount of force required to actuate each trigger. Moreover, the
argument largely misreads the Colwells’ theory, which primarily focuses on the
fact that the tabbed trigger lessens the surface area where a trigger can be
actuated and not the force required to do so. On those points as well, we reach no
factual conclusion on the merits of the Colwells’ claims; we rule only that a
reasonable jury could rule in their favor.
Finally, we find the dissent’s reasoning unavailing. It contends that the
26

-- 26 of 52 --

Colwells failed to “offer significantly probative evidence tending to show that
whatever entered the trigger guard was likely to have acted on the trigger in a
way that a tabbed trigger would have prevented.” Dissent Op. at 6–7. But it
adopts an unduly narrow understanding of what could constitute such
“significantly probative” evidence.
First, contrary to the dissent’s argument, the Colwells were not required to
show “the nature of the foreign object” that actuated the trigger. Id. at 7. The
question for the jury is whether it is more likely than not that the accident
occurred in a manner that would have been prevented by a trigger tab safety.
The dissent’s explanation for why the exact nature of any foreign object that hit
the trigger would be dispositive is that a caught “strand of clothing . . . may need
to wrap more completely around the face of the trigger” than a more solid
foreign object. Id. That may be true in theory, but it ignores testimony from
Colwell that limits the universe of potential foreign objects that could have
actuated the trigger, and specifically makes the “strand of clothing” theory less
likely. Colwell testified that the shirt he was wearing the day of the accident was
too small, and that, as a result, it kept “riding up,” leaving his “stomach . . .
hanging out,” App’x 45 at 92:6, and that he was not wearing any other layer on
27

-- 27 of 52 --

top of that shirt. A jury could thus reasonably infer that it was unlikely that some
loose clothing item interfered with Colwell’s holster. And of course, his
testimony, which the jury could credit, was that he did not touch the trigger. The
jury could (though of course it might not, and Sig Sauer would be fully entitled to
argue that it should not) infer that some more solid piece of debris was the most
likely explanation for what hit the trigger.
More fundamentally, the dissent presumes that a preponderance of
evidence can only be satisfied where we know how precisely an accident
occurred. Anything else, in its view, is mere speculation. That is not so. Tort
liability can exist where the precise cause of an accident is not known, as
illustrated by the doctrine of res ipsa loquitur. Courts apply that doctrine to allow
for an inference of a defendant’s negligence “where the accident is more or less a
mystery with no particular cause indicated” but the accident is of the sort that
would typically not occur without “negligence on the part of those in charge.” W.
Page Keeton, et al., Prosser and Keeton on the Law of Torts § 39, at 247–48 (5th
ed. 1984). Though that doctrine itself is not implicated here, it stems from the
more general principle that a “plaintiff is not required to eliminate with certainty
all other possible causes or inferences,” because doing so “would mean that the
28

-- 28 of 52 --

plaintiff must prove a civil case beyond a reasonable doubt.” Id. at 248. Instead,
“[a]ll that is needed is evidence from which reasonable persons can say that on
the whole it is more likely that there was negligence associated with the cause of
the event than that there was not.” Id.
That logic applies with equal force to the issue of causation here. We may
not (and may never) know what actuated Colwell’s trigger, and Sig Sauer is free
to raise to the jury hypotheticals in which the same foreign object could actuate
both a P320’s trigger and a tabbed trigger. But under the circumstances that we
must presume at this stage to be true — that the gun went off in its holster, that
there was no clothing nearby, and that a tabbed trigger substantially reduces the
surface area that can effectuate a discharge — there is enough evidence “from
which reasonable persons can say on the whole” that a tabbed trigger would
have prevented Colwell’s accident. The Colwells need not rule out every
alternative as the dissent would require.
Second, the dissent argues that even if Colwell did know what the foreign
object was, a juror still could not assess the role a tabbed trigger would have
played, because a juror cannot “reliably answer[]” that question “through
processes familiar to the average person in everyday life.” Dissent Op. at 8
29

-- 29 of 52 --

(internal quotation marks omitted). Although the dissent professes that it has not
“underestimate[d] jurors” in concluding that they are ill-equipped to address the
issue of causation without expert testimony, it does precisely that. Id. at 12
(internal quotation marks omitted). Certainly whether a certain foreign object can
actuate a trigger under certain conditions involves questions of “physics,
engineering, and ergonomics.” Id. at 8. But lay people engage with such concepts
as part of their everyday lives, even if they do not use technical scientific terms,
register the precise measurements, or fully comprehend the underlying theories.
From opening stuck jars to testing car brakes on icy roads to using a gun for
hunting or at a shooting range, our lives are rich in experiences that require us to
gauge and calibrate force, motion, and speed; there is no reason why a juror
cannot translate those commonplace experiences to the particular facts of this
case.
Moreover, the dissent ignores the fact that the jury would have the
assistance of expert testimony, which has not been excluded, on the relevant
mechanical operation of the P320’s trigger and of trigger tab safeties. Equipped
with that knowledge and its own assessment of the physical evidence — which
may include examining Colwell’s holster to see what type of foreign object could
30

-- 30 of 52 --

access the trigger, pressing the P320 trigger to assess how much force is required
for it to move, and measuring the degree to which a tabbed trigger lessens the
surface area that can effectuate a discharge — a juror can arrive at a reasonable
conclusion as to whether, by a preponderance of the evidence, the P320’s lack of
a tabbed trigger was a substantial factor in Colwell’s injury.
Our conclusion generally aligns with decisions from the Third and Sixth
Circuits that found similar causation opinions from Tertin and Dr. Vigilante to be
inadmissible while also holding that expert testimony was not necessary to show
causation. While those cases were decided under other states’ products liability
laws, their reasoning accords with ours. In the Third Circuit’s Slatowski decision,
the plaintiff’s P320 discharged when he went to draw it from his holster. 148
F.4th at 135. He theorized that some foreign object had “gotten into the holster,
lightly press[ed] the trigger and so disengag[ed] the safety lock,” and that, as a
result, when he went to draw the gun, “the motion jostled it, dislodging the sear,
releasing the striker, and firing the gun.” Id. at 136. The court found that the
causation issue was “not beyond the average juror,” who “would have the
benefit of the experts’ descriptions of the different safeties” and could examine “a
P320 as well as a gun with a tabbed trigger” to see “whether [the plaintiff’s] story
31

-- 31 of 52 --

really h[e]ld[] up.” Id. at 140.
Similarly in the Sixth Circuit’s Davis decision, the plaintiff testified that the
gun fired while “fully holstered” as he was “getting out of his truck.” 126 F.4th at
1219. As is the case here, official documentation of the incident suggested that the
plaintiff “was ‘attempting to’ but had not fully holstered his gun when he shot
himself.” Id. at 1231. The court found that whether Davis “shot himself by fully
depressing the trigger (which would not have been prevented by any of Davis’s
proposed alternative designs) or whether Davis inadvertently shot himself
through a graze or a side-pull (which the alternative designs would have
prevented) [was] a classic genuine dispute of material fact” for the jury to
resolve. Id. at 1231–32.
We note one circuit decision that came out the other way. In Herman, a
non-precedential decision from the Tenth Circuit, the court affirmed summary
judgment for Sig Sauer in the absence of causation testimony, finding that the
plaintiff failed to “provide[] evidence that the thing that made contact with his
P320 would likely not have squarely depressed a tabbed trigger.” Herman v. Sig
Sauer Inc., No. 23-6136, 2025 WL 1672350, at *7 (10th Cir. June 13, 2025). Key to
the decision was the plaintiff’s failure to identify “any clear evidence in the
32

-- 32 of 52 --

record about whether parts of his body or his holster did or did not come into
contact with the trigger.” Id. at *7 n.4. But here Colwell testified that his finger did
not contact the trigger and that his pistol was fully holstered at the time it
discharged. Herman is thus distinguishable. In any event, while all of these
decisions are instructive, they all deal with the highly particular facts of each case
and the particular laws of different states, and do not bind us. We find the
reasoning of the Third and Sixth Circuits more generally persuasive, but we do
not and need not decide whether we would have reached the same or a different
conclusion on the particular records of each of these cited cases.
IV. Optional Equipment Doctrine
Sig Sauer separately argues that the district court’s exclusion of Tertin’s
and Dr. Vigilante’s opinions and, by extension, summary judgment should be
affirmed on an alternative ground that the district court did not address. It urges
us to apply New York’s optional equipment doctrine, because it sells a model of
the P320 with a manual thumb safety that the Troy Police Department did not
purchase. The optional equipment doctrine, also referred to as the Scarangella
exception, bars a finding of a “design defect” where the plaintiff (or, as is the case
here, his employer) elected not to purchase optional safety equipment for a
33

-- 33 of 52 --

product. Scarangella v. Thomas Built Buses, Inc., 93 N.Y.2d 655, 661 (1999). It
requires that the defendant show “(1) the buyer is thoroughly knowledgeable
regarding the product and its use and is actually aware that the safety feature is
available; (2) there exist normal circumstances of use in which the product is not
unreasonably dangerous without the optional equipment; and (3) the buyer is in
a position, given the range of uses of the product, to balance the benefits and the
risks of not having the safety device in the specifically contemplated
circumstances of the buyer’s use of the product.” Id. (emphasis in original). When
those factors are satisfied, New York law recognizes that the purchaser “is the
party in the best position to exercise an intelligent judgment to make the trade-off
between cost and function, and it is he who should bear the responsibility if the
decision on optional safety equipment presents an unreasonable risk to users.”
Biss v. Tenneco, Inc., 409 N.Y.S.2d 874, 877 (4th Dep’t 1978).12
12 Our conclusion above that the Colwells have not introduced any evidence that
a thumb safety would have prevented Colwell’s injury does not impact our
assessment of whether Sig Sauer can invoke the Scarangella exception based on
the availability of a P320 with a thumb safety. Were we to find that the exception
applies here, it would bar the Colwells’ argument regarding the tabbed trigger.
The doctrine is agnostic as to the particular type of optional safety equipment a
defendant offers, and the Colwells and their experts do not dispute (and, in fact,
affirmatively argue) that a manual thumb safety would have rendered the P320
reasonably safe.
34

-- 34 of 52 --

Sig Sauer first raised the optional equipment doctrine in its reply briefs in
support of its motions to exclude the expert opinions before the district court,
and the district court did not address the issue. Several circuits have held that
“raising an issue for the first time in a reply brief in the district court does not
suffice to preserve the argument for appeal.” Barany-Snyder v. Weiner, 539 F.3d
327, 331–32 (6th Cir. 2008) (internal quotation marks omitted and alterations
incorporated); see also Novosteel SA v. United States, 284 F.3d 1261, 1274 (Fed. Cir.
2002); MBI Group, Inc. v. Credit Foncier du Cameroun, 616 F.3d 568, 574–75 (D.C.
Cir. 2010). Sig Sauer fails to acknowledge that it did not raise this argument in its
summary judgment briefing or in its principal briefs for its motions to exclude
before the district court and offers no reason why we should excuse its failure to
do so.
But even assuming that Sig Sauer properly preserved this argument, we
would conclude that it has failed to satisfy the first two factors of the optional
equipment test, and therefore is not entitled to summary judgment on that
ground. As to the first factor, Sig Sauer presents no evidence that the Troy Police
Department was “actually aware” that the P320 is available with a manual thumb
safety, Scarangella, 93 N.Y.2d at 661, and thus cannot meet the subjective standard
35

-- 35 of 52 --

applied by New York courts, see, e.g., id.; Campbell v. International Truck & Engine
Corp., 822 N.Y.S.2d 188, 189 (4th Dep’t 2006); Mariani v. Guardian Fences of WNY,
Inc., 148 N.Y.S.3d 571, 573 (4th Dep’t 2021). Sig Sauer submits only statements
from its expert that the department “could have purchased SIG P320 pistols with
manual safeties but chose to purchase the SKU of SIG P320 that did not contain a
manual safety.” See Appellee’s Brief 6-7. But the mere availability of a safer
alternative does not establish as a matter of law that the Troy Police Department
made a conscious choice to forego the manual thumb safety add-on. Sig Sauer’s
expert provided no basis for that assertion aside from the general observation
that some police departments prohibit their officers from carrying firearms
“equipped with any traditional manual safety,” and he simply “[p]resum[ed]
[that] the Troy Police Department [fell] into . . . [that] category.” S. App’x 28. We
cannot draw that inference in Sig Sauer’s favor on its motion for summary
judgment.
Sig Sauer similarly fails to meet the second factor of the optional
equipment doctrine, as it makes no effort to show that there are “normal
circumstances of use in which” the P320 “is not unreasonably dangerous
without” the thumb safety. Scarangella, 93 N.Y.2d at 661. All it says on this point
36

-- 36 of 52 --

is that the P320 without a thumb safety “is not defective because it is tailored for
customers whose philosophy of use excludes manual safeties.” Appellee’s Br. 29.
But Sig Sauer cites no legal authority for the proposition that catering to customer
choice is sufficient to satisfy this factor. Instead, the factor as articulated in
Scarangella requires a showing that “there exist buyers who use the product
normally and can forgo the safety feature without unreasonable risk.” Passante v.
Agway Consumer Products, Inc., 12 N.Y.3d 372, 385 (2009) (Smith, J., dissenting)
(emphasis added).13
In Scarangella, for example, the plaintiff was a school bus driver who was
hit by a bus backing up in the bus parking lot. 93 N.Y.2d at 657. She claimed that
the bus was defective because it did not have a back-up alarm. Id. at 661. The
New York Court of Appeals disagreed. Id. Because the buses backed up only in
and around the bus lot, the only “individuals at risk from the absence of back-up
13 In Passante, the majority referred to this factor as requiring that the defendant
show that the product without the optional safety equipment “would normally
be used in circumstances in which the product is not unreasonably dangerous,” a
seemingly more demanding showing than the formulation articulated in
Scarangella. 12 N.Y.3d at 381; see also id. at 385 (Smith, J., dissenting) (explaining
that the majority’s description was a “misstatement, or at best a confusing
paraphrase, of what Scarangella said”). Because Sig Sauer has failed to meet
Scarangella’s less demanding articulation of the second factor, we need not
determine whether Passante in fact narrowed the Scarangella standard.
37

-- 37 of 52 --

alarm equipment” were bus drivers and other lot employees who “were fully
aware of a bus driver’s blind spot in backing up a bus.” Id. at 662. Under such
circumstances, “the risk of harm from the absence of a back-up alarm was not
substantial.” Id. at 661. Here, Sig Sauer offers no description of how a P320
without a thumb safety might be used in a way that would not leave the user
vulnerable to its allegedly defective design.14
Based on the summary judgment record, Sig Sauer has failed to show that,
as a matter of law, it meets the requirements of the optional equipment doctrine.
We therefore decline to affirm the district court’s judgment on that basis.15
14 Because Sig Sauer has not satisfied the first and second Scarangella factors, we
need not consider the third.
15 The Colwells also argue that the district court’s ruling was based on an
incomplete review or misreading of the record. In particular, the Colwells
challenge the district court’s assertion that they “failed to respond to [Sig Sauer's]
statement of undisputed material facts,” as required by the local rule governing
summary judgment oppositions. Appellants’ Br. 43, quoting App’x 388 n.1.
Because we find that the district court erred in granting summary judgment for
Sig Sauer on other grounds, as explained above, we need not reach the Colwells’
arguments regarding the adequacy of the district court’s review of the record. We
note, however, that the Colwells did file a response to Sig Sauer’s statement of
undisputed facts as the lead document for their submission in opposition to Sig
Sauer’s motion for summary judgment. See Dist. Ct. Dkt. 51.
38

-- 38 of 52 --

CONCLUSION
For the reasons above, we VACATE the district court’s grant of summary
judgment and REMAND for further proceedings consistent with this opinion.
39

-- 39 of 52 --

RICHARD J. S ULLIVAN, Circuit Judge, dissenting:
This case turns on a deceptively simple question: would a tabbed trigger
safety have prevented Sergeant Michael Colwell’s Sig Sauer P320 pistol from going
off in its holster? I agree with the majority that the Colwells’ experts – Dr. William
J. Vigilante, Jr. and James Tertin – lacked any basis to opine that it would have. As
the majority explains, a paucity of record evidence regarding the specific
circumstances of Sergeant Colwell’s accident combined with an utter lack of
serious incident-specific testing left Vigilante and Tertin with a “very limited
understanding of how the accident happened.” Maj. Op. at 18 n.7. So, as in
previous cases involving P320-related accidents, Vigilante’s and Tertin’s professed
views that a tabbed trigger would have prevented the accident here reflect nothing
more than their “subjective belief or unsupported speculation.” Slatowski v. Sig
Sauer, Inc., 148 F.4th 132, 138 (3d Cir. 2025) (internal quotation marks omitted).
Unlike the majority, however, I do not think that a lay jury’s understanding
of the accident would be any less limited than Vigilante’s or Tertin’s. Nor would
a jury’s reliance on “common sense” to determine proximate cause in lieu of
scientific testing amount to anything other than unsupported speculation. Maj.

-- 40 of 52 --

2
Op. at 22. I therefore respectfully dissent from the majority’s decision to vacate
the district court’s grant of summary judgment in favor of Sig Sauer.
Under New York law, the plaintiff in a design-defect case bears the burden
of proving proximate cause – i.e., “that the defect was a substantial factor in
causing [his] injury.” Voss v. Black & Decker Mfg. Co., 59 N.Y.2d 102, 110 (1983); see
also Codling v. Paglia, 32 N.Y.2d 330, 342 (1973). To survive summary judgment,
the Colwells were therefore required to offer “significantly probative” evidence
from which a reasonable jury could find that the absence of a tabbed trigger from
Sergeant Colwell’s P320 more likely than not played a substantial role in causing
the gun to go off. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 249 (1986); see Fed.
R. Civ. P. 56(a). Evidence of proximate cause that is “merely colorable,” Anderson,
477 U.S. at 249, or does no more than “show that there is some metaphysical
doubt” as to whether a tabbed trigger would have made a difference, Matsushita
Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 586 (1986), is not enough. Nor
may the Colwells ask the jury to infer proximate cause based on “unsubstantiated
speculation.” Bermudez v. City of New York, 790 F.3d 368, 374 (2d Cir. 2015) (internal
quotation marks omitted); see Hayes v. Dahlke, 976 F.3d 259, 267–68 (2d Cir. 2020)
(“Conclusory allegations, conjecture, and speculation, as well as the existence of a

-- 41 of 52 --

3
mere scintilla of evidence in support of the nonmoving party’s position, are
insufficient to create a genuinely disputed fact.” (alteration adopted and internal
quotation marks and citations omitted)). Under these principles, I disagree with
the majority’s conclusion that the Colwells produced sufficient evidence of
proximate cause to warrant a trial.
As an initial matter, the Colwells’ objective evidence regarding the cause of
the gun’s discharge is wholly inconclusive. Barring a manufacturing defect of the
sort not at issue in this case, there are essentially two ways that a P320 can fire.
First, a P320 will fire if its trigger is fully depressed to the rear – as would happen
during ordinary use of the weapon. Depressing the trigger to the point that the
gun fires normally requires just under 6.5 pounds of force. Second, according to
Tertin’s expert report, a P320 can fire if the trigger is depressed far enough to
deactivate the gun’s internal striker safety lock – which requires only about 1.5
pounds of force and a significantly shorter distance of travel – and the sear1
becomes disengaged from the striker through “vibration, jostling, or contact with
the P320.” App’x at 197.
1 In a striker-fired handgun like the P320, the sear holds back the cocked striker and, when
functioning properly, prevents the striker from falling (and setting off the gun) until the trigger
is pulled.

-- 42 of 52 --

4
According to Sergeant Colwell’s deposition testimony, his gun went off
moments after he holstered it, and only after he had already taken his hand off the
weapon. In Sergeant Colwell’s telling, he was therefore not applying any direct
force to the trigger, nor any downward pressure on the gun itself, at the time it
went off. Nor is there any evidence that, with the gun holstered, a foreign object
somehow applied sufficient force against the trigger to fully depress it.
The record evidence about the second possibility – involving a partial
depression of the trigger to deactivate the gun’s internal striker safety lock,
followed by a disengagement of the sear through vibration, jostling, or contact – is
equally tenuous. Again, no objective evidence in the record establishes that any
foreign object actually touched the trigger of Sergeant Colwell’s P320. And under
this scenario, the gun could have gone off only if the sear disengaged from the
striker through some kind of vibration, jostling, or other contact with the weapon
while the trigger was partially depressed. For instance, if Sergeant Colwell had
dropped the gun or struck it with another object, then, assuming that the trigger
was also partially depressed, the gun could have gone off accidentally. But again,
the record contains no evidence that Sergeant Colwell’s P320, while holstered,
experienced vibration, contact, or jostling of the degree needed to cause the sear

-- 43 of 52 --

5
to disengage. Indeed, the only force applied to the gun was the pressure of
holstering it plus whatever force might have acted on the weapon as Sergeant
Colwell reached across his body for his taser. But neither Vigilante nor Tertin
performed any tests to determine whether those motions could be enough to cause
the sear on a P320 to jostle loose, and, to my mind, that is precisely the sort of
complex engineering and physics question that “should not be resolved by laymen
without expert assistance.” Connors v. Univ. Assocs. in Obstetrics & Gynecology, Inc.,
4 F.3d 123, 128 (2d Cir. 1993) (internal quotation marks omitted).
Given the absence of direct evidence regarding the cause of the accident, the
Colwells must rely on the assertion that something necessarily acted on the P320’s
trigger, causing the gun to fire. The parties agree that Sergeant Colwell’s P320
could not have gone off, at least under the circumstances of this case, absent some
degree of force being applied to the trigger. And because we must assume for
purposes of this appeal that a jury would credit Sergeant Colwell’s testimony that
he was not touching the gun at the time of the accident, we must further accept
that some foreign object – be it a strand of fabric, a pen cap, or even Sergeant
Colwell’s holster itself – necessarily contacted the trigger in such a way as to cause
the gun to fire.

-- 44 of 52 --

6
Although we are constrained to grant the Colwells’ premise that a
reasonable jury would find that some foreign object caused Sergeant Colwell’s gun
to fire, that is only the beginning of the proximate-cause analysis. A tabbed trigger
will prevent an accidental firing only under specific circumstances: an object must
graze the side of the trigger with sufficient force to fire the gun, but without
making contact with the center-face of the trigger, where the tab sits. We therefore
must ask whether a jury could rationally conclude not only that a foreign object
made contact with the trigger of Sergeant Colwell’s P320 and caused it to fire, but
also that the foreign object, more likely than not, did so without making contact
with the center-face of the trigger, in which case a trigger tab would have made no
difference.
In my view, the answer to that question is “no.” As the majority seems to
acknowledge, the record is completely devoid of any evidence suggesting what
kind of object caused Sergeant Colwell’s gun to fire, how that object made its way
into the gun’s trigger guard, or what part of the trigger was touched. The majority
brushes these evidentiary gaps aside on the ground that “the Colwells need not
prove certainty.” Maj. Op. at 25. True enough, but they must still offer
significantly probative evidence tending to show that whatever entered the trigger

-- 45 of 52 --

7
guard was likely to have acted on the trigger in a way that a tabbed trigger would
have prevented.
On this score the record is bare. Neither the Colwells nor their experts have
explained what potential objects could have gotten into the trigger guard; nor have
they shown that all potential objects that could have gotten into the trigger guard
would have acted on the trigger in the same way. It may be that a solid object, like
a pen cap, has a significantly different chance of sufficiently depressing a trigger
through a side-pull or graze than a soft object like a strand of clothing, which may
need to wrap more completely around the face of the trigger to be capable of
actuating it at all. In the absence of any evidence concerning the nature of the
foreign object allegedly responsible for Sergeant Colwell’s accident or how it
contacted the trigger, it is difficult to see how a jury could rationally conclude that
the object – whatever it may have been – was more likely to have depressed the
trigger through a side-pull or graze than through a frontal pull. See Anderson,
477 U.S. at 252 (emphasizing that “there must be evidence on which the jury could
reasonably find for the plaintiff” at the summary-judgment stage).
That brings me to the role of expert testimony. As the majority notes, New
York law generally requires expert testimony regarding proximate cause in cases

-- 46 of 52 --

8
where the relevant issues are “beyond the sphere of the ordinary juryman.” Fane v.
Zimmer, Inc., 927 F.2d 124, 131 (2d Cir. 1991). In the federal-law context, we have
similarly explained that expert testimony “is often required” when “the nexus
between the injury and the alleged cause would not be obvious to the lay juror.”
Wills v. Amerada Hess Corp., 379 F.3d 32, 46 (2d Cir. 2004) (Sotomayor, J.) (internal
quotation marks omitted).
This is such a case. An unknown foreign object that pressed an unknown
part of the trigger of a holstered pistol, thereby causing the gun to discharge in an
unknown way, is not the sort of situation that an “ordinary person” – or even an
experienced gun owner – would typically encounter in his lifetime. Fane, 927 F.2d
at 131. The probability that the unknown object would have been thwarted by a
trigger tab – that is, the odds that the object would have depressed the pistol’s
trigger through a grazing side-pull rather than a frontal pull – is therefore not the
kind of question that can be reliably answered through “processes familiar to the
average person in everyday life.” United States v. Garcia, 413 F.3d 201, 215 (2d Cir.
2005).
Instead, that question turns on complicated issues of physics, engineering,
and ergonomics, all in the context of a particular incident. It is the kind of question

-- 47 of 52 --

9
that even a sophisticated gun manufacturer would presumably have to spend
countless hours testing and analyzing during the process of designing and
implementing safety features on its firearms. And it should not be lost on us that
this is now the fourth P320 case in which even the plaintiffs’ gun experts have been
unable to come up with anything better than rank “speculation” as to whether a
tabbed trigger was likely to have prevented a particular accident. Slatowski,
148 F.4th at 138; see also Davis v. Sig Sauer, Inc., 126 F.4th 1213, 1224–26 (6th Cir.
2025); Herman v. Sig Sauer Inc., No. 23-6136, 2025 WL 1672350, at *5–6 (10th Cir.
June 13, 2025) (non-precedential). With no evidence pointing to the identity of the
foreign object that supposedly found its way into the trigger guard of Sergeant
Colwell’s gun or how it might have contacted the trigger, and zero incident-
specific testing to determine how likely different objects would be to overcome a
tabbed trigger under the circumstances of this accident, the jury could do little
more than guess at the nexus between the injury and the cause.2
I therefore do not agree with the majority that a jury could find proximate
cause based on a “series of reasonable determinations,” even with the aid of
2 The Majority argues that “the Colwells were not required to show the nature of the foreign object that
actuated the trigger.” Maj. Op. at 27 (internal quotation marks omitted). In essence, the majority suggests
that a jury may, without the aid of expert testimony, simply guess at what foreign object, if any, struck the
trigger. But it cannot be that a jury may find that some unidentifiable, hypothetical object caused the
accident without at least some guidance as to what that object might be.

-- 48 of 52 --

10
Vigilante’s and Tertin’s surviving testimony regarding general cause. Maj. Op.
at 23. The first two determinations that the majority proposes – that the P320 “is
more prone than certain other firearms to discharging inadvertently” and that “a
tabbed trigger reduces the likelihood of inadvertent discharges,” id. at 24 – merely
establish “that the gun could fire more easily in theory.” Slatowski, 148 F.4th at 138
(internal quotation marks omitted). But the record still lacks any evidentiary basis
for concluding how much more likely Sergeant Colwell’s P320 was to have fired
accidentally compared to a gun equipped with a tabbed trigger under the
circumstances of this case. Indeed, no one can say what those circumstances are;
as explained above, the record contains no evidence about what kind of object
made contact with the trigger, how it did so, or whether Sergeant Colwell’s gun
went off due to a full trigger depression or a partial depression combined with the
sear jostling loose. Because the probability that a tabbed trigger would have
prevented the accident could vary dramatically depending on these unknown
variables, a jury would have to engage in unsupported speculation to find
proximate cause by a preponderance of the evidence.
I do not share the majority’s confidence that a jury could overcome these
evidentiary gaps through “its own inspection of [Sergeant] Colwell’s holster, a

-- 49 of 52 --

11
P320, and a gun equipped with a tabbed trigger.” Maj. Op. at 24. Such an
inspection is not likely to generate any useful information beyond the mere fact
that a tabbed trigger theoretically could have prevented the accident. Nor am I
confident that “the jury could use its common sense to determine whether any
foreign object shown to be in the vicinity of the incident would have been likely to
actuate the P320’s trigger but not a tabbed one.” Id. On this record, there has been
no “show[ing]” that any particular kind of foreign object actuated the trigger or
how it might have done so. Id. And how various objects are likely to interact with
a tabbed trigger versus a non-tabbed trigger in the context of Sergeant Colwell’s
particular motions is, again, the kind of question that can be answered only
through testing, not through mere intuition.
The sole New York case that the majority relies on in deciding that expert
testimony on proximate cause is unnecessary is Voss, 59 N.Y.2d 102. There, the
New York Court of Appeals held that a jury could conclude that a longer blade
guard would have prevented a falling power saw from cutting the plaintiff’s hand.
See id. at 110. But Voss differs in critical respects from this case. First, the plaintiff
there could testify as to how the accident happened, whereas Sergeant Colwell
cannot. Second, and most importantly, the causal relationship between the alleged

-- 50 of 52 --

12
defect in Voss (a blade guard that left too much of the saw’s blade exposed) and
the plaintiff’s injury (a thumb severed by the falling saw) was obvious based on
the circumstances of the accident. Here, on the other hand, the probability that a
tabbed trigger would have prevented a foreign object from firing Sergeant
Colwell’s gun could vary dramatically based on the nature of that object, how it
got into the trigger guard, and whether the gun was fired through a complete
trigger depression or a partial depression coupled with a slipping of the sear. But
on this record, a jury could do no more than guess at these significant confounding
factors.
To be clear, I do not mean to suggest that expert testimony on proximate
cause is always required in P320 design-defect cases like this one. Courts, after all,
must be careful “not [to] underestimate jurors.” Slatowski, 148 F.4th at 134. In
another case, evidence short of expert testimony on proximate cause could very
well be enough to support a claim that the absence of a tabbed trigger on a P320
pistol caused the plaintiff’s injuries. But, in my view, a rational finding of
proximate cause must be based on more than the mere facts that a tabbed trigger
can prevent a P320 from accidentally going off, that a P320 can fire more easily than

-- 51 of 52 --

13
some other guns, and that something caused the plaintiff’s P320 to go off by
accident.
On those bare facts, a plaintiff has done nothing more than create
“metaphysical doubt” as to whether a tabbed trigger would have made a
difference in his particular case. Matsushita, 475 U.S. at 586. And if those facts are
enough to allow a jury to find proximate cause by a preponderance of the
evidence, then any P320 owner who simply testifies, “my gun went off by accident,
but I have no idea which part of the trigger was touched or by what,” would also
be entitled to roll the dice before a jury. In my view, it takes more to survive
summary judgment.
* * *
If there is one thing that everyone can agree on, it is that Sergeant Colwell
suffered a terrible injury. The Colwells’ desire to blame the accident on Sig Sauer’s
design choices is understandable, as is the majority’s eagerness to give them a
chance at trial. But without significantly probative evidence of proximate cause, a
jury will be required to rely on impermissible speculation. I therefore respectfully
dissent from the majority’s decision to vacate the district court’s grant of summary
judgment in favor of Sig Sauer.

-- 52 of 52 --