Roger Parks v. Kia Motors Am., Inc.

NOT RECOMMENDED FOR PUBLICATION
File Name: 22a0081n.06

Nos. 20-5690/5693

UNITED STATES COURT OF APPEALS
FOR THE SIXTH CIRCUIT FILED
Feb 24, 2022
DEBORAH S. HUNT, Clerk
AARON HILL and LYNETTA HILL, individually )
and as next friends and next of kin of John Hill, )
Deceased, and James Hill, Deceased (20-5690); ) APPEAL FROM THE UNITED
ROGER DALE PARKS, LEE JUNE CASTOR, ) STATES DISTRICT COURT FOR
NINA JOY RICE, AND JIMMIE RUTH ) THE EASTERN DISTRICT OF
NORTHCUTT (20-5693), ) TENNESSEE
)
Plaintiffs – Appellants, )
)
v. ) OPINION
)
KIA MOTORS AMERICA, INC., et al., )
)
Defendants – Appellees.
)

Before: CLAY, GIBBONS, and BUSH, Circuit Judges.

CLAY, Circuit Judge. Plaintiffs Aaron Hill, Lynetta Hill, Roger Dale Parks, Lee June

Castor, Nina Joy Rice, and Jimmie Ruth Northcutt appeal from the district court’s order granting

summary judgment to Defendant Kia Motors America, Inc., et al., and holding as moot the motions

to exclude two expert witnesses in this case alleging negligent design and manufacture of the

subject 2008 Kia Optima under the Tennessee Products Liability Act (“TPLA”), Tenn. Code Ann.

§ 29-28-101, et seq. For the reasons set forth below, we REVERSE the district court’s holding

as moot the motions to exclude expert witnesses Kress and Loudon, REVERSE the district court’s

order granting summary judgment to Defendants, and REMAND the case for further proceedings

consistent with this opinion.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 2

I. BACKGROUND

A. The December 31, 2015 Accident

This case is about a sudden and unintended vehicular acceleration event on December 31,

2015, that resulted in a triple fatality. Mid-to-late morning that day, eighty-three-year-old Mary

Jean Parks was driving her 2008 Kia Optima on Dinah Shore Boulevard in Winchester,

Tennessee.1 Parks and her seventy-five-year-old sister, Plaintiff Jimmie Northcutt, were en route

from an appointment to the local Kroger to purchase milk, a familiar drive just four miles from

their shared home. Parks was known to be a careful and cautious driver. The roadway conditions

that day were unremarkable, and visibility was unobstructed. The posted speed limit was 30 miles

per hour. Upon entering the intersection of Dinah Shore Boulevard and Bypass Road, Parks’

vehicle suddenly accelerated to 90 miles per hour with over 4,000 revolutions per minute (“rpm”)

for a half-mile.

Eyewitnesses recalled seeing “[Parks’ vehicle] flying” and “not slowing down;” the

Optima was “going . . . too fast to be able to stop in time with the red light.” (Kim Taylor Dep.,

R. 317-22 at PageID ## 11726, 11727). The car was traveling at such a high rate of speed that “it

shook [an eyewitness’ car].” (Tommy Philpot Dep., R. 317-21, PageID # 11671). Parks’ car “just

kept getting faster and faster.” (John Dance Dep., R. 317-17, PageID # 11535). Eyewitnesses

attested that the Kia was “zigging in and out of these cars,” (Kim Philpot Dep., R. 317-20, PageID

# 11638), in an apparent effort to “dodg[e]” other vehicles. (Tommy Philpot Dep., R. 317-21,

PageID # 11669; see also id. at PageID ## 11675–76 (“She was trying to warn us to get out of the

way. She was letting us know something’s wrong.”)). One eyewitness remembered hearing “the

1
Parks’ vehicle was manufactured in February 2008 and purchased on July 30, 2008.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 3

car making . . . [a] weird sound,” which sounded like “the engine revving . . . up and then down.”

(Bobby Metcalf Dep., R. 317-18, PageID ## 11579, 11581). Parks reportedly engaged her hazard

lights or flashed her headlights to warn motorists that something was amiss, though not all

eyewitnesses reported seeing any lights. A defense expert stated that none of the surveillance

camera footage shows that the brake lights were illuminated.

The Kia Optima crashed into a 2003 Ford Windstar. Plaintiffs Aaron Hill and Lynetta Hill

and their two seven-year-old sons, John and James, occupied the Windstar and were stopped at a

red light. The force of the crash propelled the Ford Windstar into a Ford F-150 pickup truck. At

impact, the Kia’s vehicle’s readings were 4,300 rpm on the tachometer and 92 miles per hour on

the speedometer; the electronic throttle control angle was around 80%.

Eyewitnesses ran to help. These good Samaritans recalled seeing Parks pinned under the

dashboard. Parks told eyewitnesses “over and over and over” (Tommy Philpot Dep., R. 317-21,

PageID # 11700) that she was sorry and that “she couldn’t stop the car . . . [t]he car had a mind of

its own.” (Kim Philpot Dep., R. 317-20, PageID # 11644). Similarly, Plaintiff Northcutt testified

that at some point, either before or right after the accident, Parks told her that “something [was]

wrong with this car” and that she was unable to stop accelerating, despite apparently applying the

brakes. (Jimmie Northcutt Dep., R. 317-19, PageID ## 11603, 11612). Emergency personnel

arrived, and Parks told officers that “something happened to [her] car,” and that “[she] could not

control it.” (Compl., R. 1, PageID # 10).

Parks and the Hill twins suffered fatal injuries. James Hill died on December 31, 2015, at

the scene of the accident; Parks died on January 1, 2016; and John Hill died on January 3, 2016.

Plaintiffs Aaron and Lynetta Hill and Plaintiff Northcutt all sustained injuries. Plaintiffs, the next-

of-kin of Parks and the Hill twins, filed suit, and this appeal followed.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 4

B. Background on Unintended Acceleration Cases

A meaningful portion of automobile products-liability caselaw is devoted to unintended

acceleration cases.2 Experts theorize variously on the cause of these unintended accelerations.

Driver error, via accidental application of the accelerator pedal (also referred to as “pedal

misapplication”), or pedal entrapment by a floormat, are recognized causes of unintended

acceleration events. On the other hand, a considerable school of thought maintains that unintended

acceleration can occur without driver error or pedal entrapment. (See Richard M. Goodman, et

al., Toyota Unintended Acceleration, Auto. Design Liability 3 (2016) (“With the advent of

electronic ignition systems and cruise control systems in the late 1970[]s and early 1980[]s

unintended acceleration complaints without clear mechanical failures began to appear.”)). Any

number of malfunctions in a vehicle’s electronics could cause sudden and unintended acceleration.

Throughout this litigation, Plaintiffs have posited many theories about what might have caused the

2015 crash. These theories include: (1) a worn clock spring; (2) a brake-lamp stop switch issue;

(3) a voltage drop; (4) malfunctions in the cruise control; (5) brake issues; (6) unforeseen issues

with the engine control unit (“ECU”); and (7) electromagnetic interference (“EMI”) or cross-talk.

It is first helpful to define some of this vocabulary.

The engine control module (“ECM”), electronic engine controller (“EEC”), electronic

control unit (“ECU”), and Powertrain Control Unit (“PCU”) are various terms used to refer to the

2008 Kia Optima’s central computer. The ECM controls the throttle. Engine power requires air,

and the throttle regulates how much air flows to the engine. When a driver presses the accelerator

2
A Minnesota Supreme Court case from 1961 is illustrative: “‘[A]ll of a sudden’ the Oldsmobile ‘took off and
jerked us back.’ He claims that he put on the brakes, as well as the emergency brake, and turned off the ignition; that
by that time he was going about 90 miles per hour; that he pulled over into the left lane to avoid hitting traffic ahead
of him and went between cars that were facing him; but that his automobile just kept on going.” Grant v. Malkerson
Sales, Inc., 259 Minn. 419, 420–21 (1961).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 5

pedal, a wire signals the ECM to open the throttle, and thus, accelerate. If a driver is “flooring it,”

then the throttle is “wide open.” Releasing the accelerator reverses the process; the car slows, and

the throttle closes when the vehicle comes to a complete stop. Post-crash vehicle inspections here

showed that the throttle plate was in a nearly wide-open orientation, indicating that Parks’ car was

delivering almost full engine power at the time of the crash.

Some defect experts theorize that electronic throttle control systems can be susceptible to

unintended, wide-open throttle acceleration due to electromagnetic interference or EMI.3 The

malfunction central to the EMI theory is a so-called “cross-talk.” A “cross-talk” is a form of EMI

where the wires in close proximity “talk” to each other, causing a malfunction. The theory is that

EMI can cause the cruise control to send a sustained wide-open command to the electronic throttle

control system without driver input.

The cruise control, a system at least conceptually familiar to any driver of a modern

automobile, is a sub-component of the electronic throttle control system, or “ETC.” In the 2008

Kia Optima, there are four cruise control functions: (1) On/Off; (2) Set/Coast; (3) Cancel; and

(4) Resume/Accel. A driver activates any of these functions by pressing the corresponding button

on the steering wheel. The buttons on the steering wheel, in turn, are connected to wires in a “clock

spring.” The clock spring is a circular ribbon cable located inside the steering column, behind the

driver’s airbag storage module; its function is to provide electrical continuity between the steering

wheel and the rest of the car. The clock spring is routed to the control buttons through a multi-

function switch. The multi-function switch is connected to the main computer through a single

3
At the district court, two of Plaintiffs’ four experts, Samuel Sero and Byron Bloch, suggested that EMI
within the Parks’ 2008 Kia Optima caused an unintended acceleration that Parks could not stop by depressing the
brake pedal. The district court excluded Sero’s and Bloch’s EMI testimony as unreliable.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 6

signal wire. The cruise control, via the resume/accel button, is the only feature, other than the

accelerator pedal, that directly communicates to open the throttle. The clock spring also

incorporates the airbag, steering wheel radio buttons, and horn functions. Each cruise-control

function operates in a specified voltage range. The ECU software is designed to turn off the cruise

control if the voltage exceeds the specified range.

C. Procedural History

On December 20, 2016, Plaintiffs Aaron Hill and Lynetta Hill, the next-of-kin of James

and John Hill, and Plaintiffs Roger Dale Parks, Lee June Castor, Nina Joy Rice, and Jimmie

Northcutt, the next-of-kin of Mary Parks, sued Defendants Kia Motors America (“KMA”), Kia

Motors Corporation (“KMC”), Hyundai-Kia Automotive Group (“Hyundai-Kia”), Hyundai

America Technical Center, Inc. (“HATCI”), Hyundai Motor Company (“HMC”), Hyundai Motor

Group (“HMG”), and Hyundai Motor America, Inc. (“HMA”).4 The complaints identified

multiple causes of action under the TPLA, including strict liability in tort, negligence, pre-and

post-sale failure to warn, and breach of warranty.

The district court consolidated the cases for discovery on July 20, 2017 and designated the

Hill Plaintiffs’ suit as the lead case. After two years of discovery, on May 30, 2019, Defendants

Kia Motors America, Inc. and Kia Motors Corporation (“Kia Defendants”) filed a motion for

4
Defendant KMA is a wholly owned subsidiary of Kia Motors Corporation and is the sole distributor of the
Kia-trademarked brand automobiles. Defendant KMC is a publicly traded foreign corporation headquartered in South
Korea. Defendant Hyundai-Kia is also a foreign corporation organized under the laws of South Korea; it merged with
Defendant KMA in 1988 to become the largest majority shareholder of Defendant KMC stock. Defendant HATCI is
a Michigan corporation that designed and engineered the 2008 Kia Optima. Defendants HMC and HMG are
headquartered in South Korea, and each separately owns at least 10% of Defendant KMA’s stock. Defendant HMA
is a California corporation and has a 34% ownership stake in Defendant KMA and is a wholly owned subsidiary of
Defendant HMC.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 7

summary judgment. The Kia Defendants also moved to exclude Plaintiffs’ experts, Samuel Sero,

Tyler Kress, Steven Loudon, and Byron Bloch.

The district court granted the motion for summary judgment, granted in part the motions

to exclude Sero and Bloch, and denied as moot the motions to exclude Loudon and Kress. It

deemed the motions to exclude Sero and Bloch partially meritorious because key portions of the

proffered testimony were unreliable within the meaning of Rule 702 and Daubert v. Merrell Dow

Pharmaceuticals, Inc., 509 U.S. 579 (1993). After exclusion of Sero and Bloch, the district court

held that Plaintiffs had not established a genuine dispute of material fact as to causation, an element

common to all their products liability claims. It then granted summary judgment to all Defendants,

including the Hyundai Defendants who had not joined the Kia Defendants’ motion. This appeal

followed. Plaintiffs appeal the district court’s grant of summary judgment to Defendants and the

holding as moot the motions to exclude Kress and Loudon.5

II. DISCUSSION

A. The District Court’s Evidentiary Rulings

1. Standard of Review

This Court reviews the district court’s decision to exclude the testimony of a party’s expert

witnesses for an abuse of discretion. Pride v. BIC Corp., 218 F.3d 566, 575 (6th Cir. 2000). A

trial court must be satisfied that a proffered expert is qualified to testify on the technical subject

matter at issue such that an expert’s expertise is sufficiently reliable to assist the trier of fact in

disposing of the relevant issues. Sigler v. Am. Honda Motor Co., 532 F.3d 469, 478 (6th Cir.

2008). To this end, while not limiting the admission of relevant evidence, the trial court must

5
They do not appeal the partial grant of the motions to exclude experts Sero and Bloch.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 8

insure that an expert uses in the courtroom the “same level of intellectual rigor that characterizes

the practice of an expert in the relevant field.” Best v. Lowe’s Home Ctrs., 563 F.3d 171, 177 (6th

Cir. 2009) (quoting Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999). A trial court’s

assessment of an expert’s testimony is limited “solely [to] principles and methodology,” so “the

conclusions that they generate” should not affect the court’s conclusion. Daubert, 509 U.S. at 595.

When reviewing a district court’s decision to exclude expert testimony under this standard, this

Court reverses “only if . . . firmly convinced of a mistake that affects substantial rights and amounts

to more than harmless error.” Pressman v. Franklin Nat’l Bank, 384 F.3d 182, 187 (6th Cir.

2004) (citation and internal quotation marks omitted).

2. Motions to Exclude

We now turn to review the district court’s evidentiary rulings holding as moot the motions

to exclude Kress and Loudon. In the absence of the testimony of defect experts Sero6 and Bloch,7

6
Plaintiffs retained electrical engineer Samuel Sero to determine “within a reasonable degree of engineering
certainty, whether the design of the 2008 Kia Optima is defective in that its cruise control system can cause an
unwanted acceleration.” (Sero Report, R. 307-1, PageID # 6943). Sero conducted a post-crash examination of Parks’
Kia Optima and determined that the vehicle was susceptible to malfunction “by unwanted changes in input voltage
values that can be misinterpreted by the computer programming and result in an unwanted acceleration.” (Id.).
Beginning Sero’s report is a review of the history of EMI-induced acceleration events, particularly starting in the early
1980s, which noted “[t]he rapid expansion of the use of electronic and electrical vehicles.” (Id. at PageID # 6946).
Acknowledging the “consummate hazard” in any vehicle “that it will accelerate on its own and without warning” due
to “uncontrolled inputs into the [electronic engine controller],” Sero went on to identify various design options that
can mitigate or prevent a sudden acceleration event. (Id. at PageID ## 6943–44; see id. at PageID ## 6946–47
(exploring design alternatives)). He then turned to the instant case. Sero postulated that the 2008 Kia Optima was
defective because it may cause an unwanted activation of the cruise control in which the throttle motor is activated,
thereby causing an un-commanded acceleration. Put differently, Sero stated that the throttle motor went into near
wide-open throttle; the malfunction of the cruise control would have sustained that condition by ignoring any brake
input.
The district court excluded Sero’s evidence as unreliable within the meaning of Rule 702 and Daubert.
According to the district court, to allow his testimony that the 2008 Kia Optima was defective insofar as EMI could
cause the cruise control to send a sustained open command to the ETC would be to admit the “untested, unpublished,
and unaccepted view that EMI can cause a sustained, unintended acceleration.” (Order, R. 341, PageID # 13021). To
the extent Sero posited a design defect claim, that too was excluded because it presumed that EMI could cause
unintended acceleration.
7
Plaintiffs retained automobile safety-design and vehicle crashworthiness expert Byron Bloch to review what
he viewed as various design defects in the subject car, focusing on the clock spring and the possibility of unanticipated
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 9

the district court denied as moot Defendants’ motions to exclude Kress and Loudon, reasoning that

their testimonies did not bear on the dispositive question of whether a specific defect could cause

unintended acceleration.

A brief review of Loudon’s and Kress’ expert reports and the parties’ concomitant briefing

is illuminating.

After conducting testing in an exemplar 2008 Kia Optima, Steven Loudon, an electrical

engineer, concluded that Parks’ vehicle, on the day of the accident, experienced errant signals from

the cruise control switch that caused the Optima to accelerate for a prolonged period and caused

the crash. Loudon theorized that the design of the cruise control system in the 2008 Kia Optima

was defective because it used a single wire through an unreliable clock spring harness.

Undergirding Loudon’s argument is a criticism of Kia’s diagnostic software routines, which he

describes as defective and incapable of detecting and mitigating the effects of the errant signals.

Human-factors engineering science expert Dr. Tyler Kress opined that “th[is] accident is

consistent with an electronic vehicle malfunction” (Kress Report, R. 317-3, Page ID # 10136), and

there was “a feasible alternative design that could have prevented the subject accident.” (Id. at

PageID # 10142). Kress positioned his findings from a human factors’ perspective, proffering that

a driver typically perceives and corrects accidental application of the accelerator pedal within 1.5

erratic voltages, i.e., EMI or “cross-talk.” Bloch posited that the clock spring design was susceptible to EMI because
it contains multiple electrical connections in close proximity; and EMI within the clock spring could send a sustained,
wide-open throttle command to the ETC. He concluded “that the clock-spring and its connections, in the 2008 Kia
Optima at-issue, caused . . . cross[-]talk (e.g., unanticipated erratic voltages) that adversely affected the Cruise Control,
and which prompted it to kick into an open-throttle condition.” (Bloch Report, R. 313-1, PageID # 7795).
The district court found that Bloch’s conclusions were not reliable within the meaning of Rule 702 and
Daubert because the theory (that EMI can occur in the clock spring) was too attenuated from the conclusion (the clock
spring EMI can cause the automobile to accelerate without driver input). Bloch’s four other theories were irrelevant
because they did nothing to fill this gap. As with Sero’s EMI theory, Bloch’s parallel cross-talk theory lacked testing,
peer review, publication, and general acceptance, prompting the court to grant in part the motion to exclude Bloch’s
testimony.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 10

seconds. (Id. at PageID # 10206; id. at PageID # 10136 (“Like vehicle malfunctions, such a series

of human errors (pedal misapplication, flooring it, and failing to correct these actions) is very rare

. . . .”)). From post-crash vehicle inspections and the injuries to Parks’ lower extremities, Kress

concluded Parks’ right foot could have been on or off the brake pedal before impact. After finding

that circumstantial evidence evinced proper handling, Kress concluded that the subject

acceleration event was not attributable to driver error but to some malfunction in the vehicle’s

computer-based, electronic operation.

The question on appeal is whether the district court’s holding the testimony of Kress and

Loudon as moot amounted to an abuse of discretion. Under Defendants’ telling, these two

remaining experts had a narrow evidentiary role at the trial court in which they only alluded to an

unidentified malfunction in the Optima that might have caused the unintended acceleration—but

they left it to Sero and Bloch to identify the malfunction. So, Defendants argue, Plaintiffs may not

now refashion their surviving experts, Kress and Loudon, as defect experts on appeal.

Contrary to the district court’s holding and the dissent’s belief, the testimonies of Kress

and Loudon were not dependent on the excluded testimony of Bloch and Sero. It would also be a

bridge too far, as the district court reasoned and the dissent asserts, to find that Kress and Loudon

do not offer a defect theory. They do and did so consistently at the lower court. Unlike their

excluded brethren, Kress and Loudon premise their defect theories primarily upon circumstantial

evidence.

Plaintiffs made this clear below as to Loudon: “Mr. Loudon’s testing, on the other hand,

provides strong circumstantial proof of the cause of Mrs. Parks’ event.” (Pls.’ Resp. & Mem. in

Opp’n to Exclude Loudon, R. 322, PageID # 12637). They also made it clear that Kress planned

to offer circumstantial proof of a defect; citing Kress’ report, Plaintiffs stated: “There is
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 11

circumstantial evidence to show that more-likely-than-not the subject accident was caused by a

sudden unexpected acceleration” and “that Mary Parks clearly implemented collision avoidance

driver input to attempt to control the out-of-control” vehicle, thereby demonstrating that “the

collision was not the result of a pedal error or pedal misapplication by Mary Parks.” (Pls.’ Resp.

& Mem. in Opp’n to Exclude Kress, R. 208, PageID # 11755; see also id. (reviewing Kress expert

report on “feasible technologies . . . that could have been used to reasonably address hazards

associated with potential vehicle malfunctions that can cause [sudden unintended acceleration]

incidents”)). Because Kress and Loudon offered circumstantial evidence of the source and cause

of the accident, we find that the district court abused its discretion in denying as moot the motions

to exclude Kress and Loudon.8

We now turn to the remaining basis of Plaintiffs’ appeal, i.e., the district court’s grant of

Defendants’ motion for summary judgment.

B. Summary Judgment

1. Standard of Review

This Court reviews de novo a district court’s grant of summary judgment. Sigler, 532 F.3d

at 482. “In deciding an appeal of a grant of summary judgment, we view the evidence and draw

all reasonable inferences in favor of [plaintiffs], the non-moving party.” Singfield v. Akron Metro.

Hous. Auth., 389 F.3d 555, 560 (6th Cir. 2004) (citing Matsushita Elec. Indus. v. Zenith Radio

Corp., 475 U.S. 574, 587 (1986)). The moving party bears the burden of showing the absence of

any genuine issues of material fact. Plant v. Morton Int’l, 212 F.3d 929, 934 (6th Cir. 2000).

8
Plaintiffs have not waived the argument that their remaining experts offer proof of a specific defect. (Contra
Dissent at 41, 52–57). Rather, Plaintiffs said time and again at the district court that Kress and Loudon offered theories
of defect based on circumstantial evidence. The issues are thus preserved for appellate review.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 12

Once the movant has satisfied its burden, the non-moving party must “come forward with evidence

showing that there is a genuine issue for trial.” Id. (citing another source). The non-movant,

however, must “do more than simply show that there is some metaphysical doubt as to the material

facts.” Matsushita, 475 U.S. at 586. This Court must “assess the proof to determine whether there

is a genuine need for trial,” and “[t]he proper inquiry is whether the evidence is such that a

reasonable jury could return a verdict for the plaintiff.” Weigel v. Baptist Hosp. of E. Tenn., 302

F.3d 367, 375 (6th Cir. 2002) (citing another source).

2. Tennessee Products Liability Act

The Tennessee Products Liability Act of 1978 (“TPLA”) provides the relevant statutory

framework. When sitting in diversity, the Sixth Circuit has set forth the requirements to establish

a plaintiff’s prima facie case under the TPLA as follows: (1) the product was defective and/or

unreasonably dangerous; (2) the defect existed at the time the product left the manufacturer’s

control; and (3) the plaintiff’s injury was proximately caused by the defective product.9 Sigler,

532 F.3d at 483. “The general rule in Tennessee is that the issue of whether a product is defective

or unreasonably dangerous is one for the jury.” Curtis v. Universal Match Corp., Inc., 778 F.

Supp. 1421, 1427 (E.D. Tenn. 1991) (quoting Jackson v. Gen Motors Corp., 60 S.W.3d 800, 806

(Tenn. 2001)).

As to the merits of their appeal, Plaintiffs state that the 2008 Kia Optima is subject to a

defect known as a “runaway throttle condition” that meets the tests of strict liability and negligence

because they have established a genuine issue of material fact under the relevant prongs of the

TPLA. This Court agrees and reverses the grant of summary judgment.

9
Only the first and third prongs are significant issues here, so the second prong is not explicitly discussed.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 13

3. The District Court’s Grant of Summary Judgment was Improper

Based largely on what it perceived to be the deficiencies in Sero’s and Bloch’s testimony,

the district court granted summary judgment in favor of Kia. Specifically, the district court found

that Plaintiffs failed to identify a specific defect that could cause unintended acceleration in a 2008

Kia Optima and thus could not trace the injury to the defect, even assuming one existed. It deemed

causation to be the dispositive issue and only minimally discussed the first prong of the TPLA.

When drawing all reasonable inferences in Plaintiffs’ favor, a grant of summary judgment to Kia

is unsupported.

a. A Specific Defect Must be Proven Under all Theories of Liability

As a preliminary matter, Plaintiffs’ baseline argument is that the district court applied an

improperly heightened legal standard to their claim. They maintain that a party need not identify

a specific error in a product’s construction or design when proceeding under a theory of liability

other than negligence. In cases of strict liability, for example, Plaintiffs argue that they need to

prove only causation. The question is: must a specific defect be proven for all theories of liability

(the district court’s holding), or only when arguing negligence (Plaintiffs’ theory)?

No matter what theory (or theories) of liability under which a party proceeds—e.g., strict

liability, negligence, breach of warranties, or failure to warn—the same three-pronged prima facie

case applies in all TPLA cases. Tenn. Code Ann. § 29-28-102(6). In rejecting Plaintiffs’ plea for

a different standard, the district court looked to the Tennessee Supreme Court’s seminal case of

Browder v. Pettigrew, 541 S.W.2d 402 (Tenn. 1976). The Browder plaintiffs brought a products-

liability action arising out of an automobile accident. The court compared what is required when

proceeding under various theories of liability and stated:
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 14

We agree with counsel that in a products liability action in which recovery is sought
under the theory of negligence, the plaintiff must establish the existence of a defect
in the product just as he does in an action where recovery is sought under the strict
liability theory or for breach of warranty, either express or implied . . . The only
significant difference is that under the negligence theory[,] the plaintiff has the
additional burden of proving that the defective condition of the product was the
result of negligence in the manufacturing process or that the manufacturer or seller
knew or should have known of the defective condition.

Id. at 404 (citing another source) (emphases added). The upshot of Browder is that a plaintiff

alleging negligence is burdened with proving breach of a duty of care—a burden of which an

aggrieved party is relieved when proceeding under strict liability.

The Browder court also discussed Greco v. Bucciconi Eng’g Co., 283 F. Supp. 978 (W.D.

Pa. 1967), upon which Plaintiffs heavily rely. In a negligence-turned-strict liability action, the

Greco court asked whether “the mere occurrence of a malfunction by machinery evidence[s] a

‘defective condition.’” Id. Of Greco, the Browder Tennessee Supreme Court said: “[It] suggests

that while proof of a malfunction alone should be sufficient under the strict liability and warranty

theories in a products liability case, a higher standard of [s]pecificity of proof of defect is required

. . . to recover under the negligence theory.” Browder, 541 S.W.2d at 404 (emphasis added)

(citations omitted).

From this, Plaintiffs contend that all they need to prove for theories other than negligence

is “malfunction alone” and that they are unburdened from any obligation to establish a specific

defect.10 This is unsupported. Regardless of the fact that Greco is an out-of-circuit district court

case from 1967, the Tennessee Supreme Court in Browder does not adopt or base its conclusion

on Greco. All that Greco signified for the Tennessee Supreme Court’s review in Browder was

10
As applied, the malfunction doctrine theory Plaintiffs set forth is that the Parks’ 2008 Kia Optima
malfunctioned during regular operation and thus a legal inference of a defect or unreasonably dangerous condition
arises under the TPLA.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 15

that the laws of strict liability and warranty (which assign liability based on the product’s lack of

fitness) were resemblant theories, in contradistinction to the law of negligence (which assigns

liability upon proof of a breach in the applicable duty of care and thus necessitates the obligation

to trace the injury to some specific defect).

Furthermore, there is in-circuit, recent caselaw reasoning that a plaintiff must prove a

specific defect under the TPLA, even for those cases not sounding in negligence. See, e.g., Tilden

v. Gen. Elec. Co., No. 3:11-CV-628, 2012 WL 1023617, at *2 (E.D. Tenn. Mar. 26, 2012) (“Thus,

regardless of plaintiff’s theory of recovery—which includes strict liability, negligence, and

implied and express breach of warranty—plaintiff must allege facts in her complaint for the Court

to infer that the MRI machine was ‘defective’ or ‘unreasonably dangerous’ at the time it left the

control of the manufacturer.”) (citing King v. Danek Med., Inc., 37 S.W.3d 429, 435 (Tenn. Ct.

App. 2000)). A plaintiff must prove a specific defect under the TPLA, regardless of the theory of

liability. Defect (or causation) might be shown with direct or circumstantial evidence. What

matters is that the defect is proven.

b. A Genuine Issue of Material Fact Exists Regarding a Defect or
Unreasonably Dangerous Condition in the Parks’ 2008 Kia Optima

This subsection, itself subdivided, concerns the first prong of the TPLA prima facie case,

i.e., proof that the product was defective and/or unreasonably dangerous. A plaintiff may show

that a product was defective or unreasonably dangerous through direct evidence, circumstantial

evidence, or a combination. Sigler, 532 F.3d at 483. The trial court in this case should not

have granted the motion for summary judgment because there are genuine issues as to whether the

Parks’ 2008 Kia Optima was in a “defective condition” or “unreasonably dangerous.” We address

each sub-prong in turn below.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 16

i. Evidence of a “Defective Condition”

The TPLA defines “defective condition” as “a condition of a product that renders it unsafe

for normal or anticipatable handling and consumption[.]” Tenn. Code Ann. § 29-28-102(2). The

district court, focused as it was on the question of causation and traceability, danced around

discussing the arguments surrounding a defect. The question on appeal is whether Plaintiffs have

established, either by direct or circumstantial evidence, proof that could have led a rational trier of

fact to conclude that Parks’ Optima was defective. For four reasons, this Court answers that

question in the affirmative.

First, the district court summarily declared that Loudon did not provide a defect theory.

Our interpretation of the record is different. The results of Loudon’s testing in an exemplar 2008

Kia Optima could lead a rational trier of fact to conclude the design of the vehicle was defective.

In that testing, Loudon simulated two acceleration scenarios to determine the cause of the

December 31, 2015 accident, querying whether the acceleration was caused by either wide-open

throttle (which would presume pedal misapplication) or whether the cruise control subsystem

received an incorrect request for the “resume/accel” function on the steering wheel (which would

presume some defect in the car). He concluded that the signature on Parks’ instrument panel—

about 4,300 revolutions-per-minute (“rpm”) on the tachometer and 92 miles per hour on the

speedometer—was most consistent with cruise control application and less consistent with wide-

open throttle application.11 A reasonable jury might well choose to believe that the circumstantial

11
At wide-open throttle, the exemplar vehicle’s tachometer indicated rpm between 5,000 and 6,000 and thus
did not match the 4,300-rpm reading on Parks’ tachometer. In the second scenario, Loudon simulated where a driver
inadvertently presses the “resume-accel” button on the steering wheel, thus allowing the vehicle, through cruise
control, to accelerate. In the cruise-control scenario, the exemplar vehicle’s tachometer was 4,119 rpm, so much closer
to the signature on Parks’ vehicle.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 17

evidence demonstrates that some defect in the cruise control was more likely than not the cause of

the unintended acceleration.

Second, the district court employed a cursory review of Kress’ report, only to conclude that

he did not offer any defect theory. That is not so. On the contrary, he considered relevant human

factors, i.e., that most drivers who accidentally apply the acceleration pedal correct that error

within 1.5 seconds, and concluded: “The unintended acceleration behavior of the subject 2008 Kia

Optima preceding . . . the collision was not the result of a pedal error or pedal misapplication by

Mary Parks.”12 (Kress Report, R. 317-3, PageID # 10149). By ruling out driver error as the cause,

the remaining primary inference is that some defect prompted an un-commanded acceleration.

Taking Plaintiffs’ argument as true that if Parks accidentally accelerated, then she would have

quickly corrected the error and applied the brakes, a jury could reasonably find that some defect

in the cruise control caused the acceleration of Parks’ Optima.

Third, at the district court, Plaintiffs drew two conclusions from the post-crash vehicle

inspections, which showed a half-inch gap between the accelerator pedal and its bracket. The first

conclusion is that the accelerator was not pressed to the floor or entrapped in a “floored” manner

and that it was completely released if not depressed at all; with regard to the second conclusion, it

is also asserted that this gap was consistent with the electrical measurements showing the

accelerator pedal in the “idle” position at the time of impact. The district court did not discuss this

theory as related to the first prong of the TPLA.

12
What is more, Kress’ report was not solely oriented to retorting the defense theory of pedal misapplication,
as the district court stated. Instead, Kress also argued that feasible design alternatives existed that would have
addressed the known hazards that can cause unintended acceleration events. This is the subject of the Court’s later
discussion.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 18

On appeal, Plaintiffs claim that pedal misapplication can be ruled out if the accelerator

pedal was in the idle position, and the possibility of some unspecified cruise control malfunction

is rendered more probable. If Parks’ right foot was not on the accelerator pedal at the time of the

crash, i.e., the inference Plaintiffs urge, then the defense theory of pedal misapplication is

weakened considerably, and Plaintiffs’ rebuttal theory of proper use is strengthened. Thus, the

pedal position and the gap could support a finding that something other than driver error caused

the accident, thus raising, circumstantially, the possibility of some unspecified cruise control

malfunction as the source of the acceleration.

Fourth and finally, Plaintiffs contend that reasonable minds could conclude, based on

witness and eyewitness testimony, that a sudden acceleration event would not have happened in

the absence of a defect.13 This res-ipsa-loquitur-like theory is similar to the Browder-Greco

discussion analyzed previously. Some courts call this the “malfunction doctrine” or “malfunction

theory.” The malfunction doctrine allows plaintiffs in strict liability cases to infer defectiveness

from the negation of other causes. See, e.g., Balducci v. Hyundai Motor Am., Inc., 406 F. App’x

517, 518 (2d Cir. 2011) (“[Plaintiff] offered no evidence demonstrating an ‘absence of other

identifiable causes’ [whereas] defendant’s evidence shows that the air bag’s non-deployment was

13
The dissent accuses this opinion of self-contradiction. It states: “In repeatedly and inexplicably claiming
that plaintiffs may prove a specific defect with a non-specific defect, the lead opinion shirks the clear requirement
under Tennessee law that a specific defect be shown.” (Dissent at 57). This mischaracterizes the opinion.
It seems that the dissent incorrectly deems synonymous a “specific defect” and “direct evidence.” However,
that a party may make her case out by circumstantial evidence under the TPLA is an unremarkable and irrefutable
concept; so, the dissent’s confusion over how a party might plead a case without direct evidence is misplaced. “Where
a plaintiff is dependent upon circumstantial evidence [to prove a defect in a product], it is sufficient if he makes out
the more probable hypothesis, and the evidence need not []rise to that degree of certainty which would exclude every
other reasonable conclusion.” Sigler, 532 F.3d at 486 (Tenn. Ct. App. 1982) (quoting Motley v. Fluid Power of
Memphis, Inc., 640 S.W.2d 222, 225).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 19

caused by the low speed of [her] car and the type of crash, not a malfunction[.]”) (quoting another

source)).14

Reference to this theory does not relieve Plaintiffs of demonstrating a defect under the

TPLA. Plaintiffs do more than solely argue that the facts of the malfunction alone furnish

circumstantial evidence of a defect. They point to specific circumstantial evidence and testimony

to support this conclusion. Such testimony includes statements from Parks herself, Northcutt, and

the eyewitnesses. Parks was known to be a safe and careful driver; Parks’ sister, Northcutt,

recalled Parks saying, “Jimmie, there’s something wrong with this car,” and the car would not

stop. (See Northcutt Dep., R. 317-19, PageID ## 11603, 11612). When emergency personnel

arrived to extract the fatally injured Parks, Parks said, “[she] could not stop.” (Hasty Dep., R. 321-

6, PageID # 12484). Eyewitnesses observed the Kia’s rapid and seemingly uncontrolled

acceleration, as well as Parks’ efforts to avoid crashing into others. If Parks was unable to stop or

control the vehicle, Plaintiffs’ argument would negate the possibility of causes for which

Defendants might not be liable. Viewing the record in the light most favorable to the non-movant,

as we must, Plaintiffs have presented proof via circumstantial evidence supporting an inference of

a defect, thus meeting the required elements for a triable TPLA cause of action.

In summary, then, this Court holds that Plaintiffs have put forth sufficient evidence creating

a material factual dispute on the question of a defect under the first half of the TPLA’s first prong.

There is a genuine issue of material fact as to whether a defect existed in the Optima.

14
The dissent contends that Plaintiffs never argued such a theory. The record indicated otherwise. That
“reasonable minds could conclude, based on the testimony of the witnesses . . . that a sudden acceleration occurred,
and that it would not have happened in the absence of some defect” is precisely what Plaintiffs contend. (Pl.’s Br. at
18; see also Pl.’s Mot. in Opposition, R. 325, PageID ## 12729–30) (“[T]here is . . . ample . . . circumstantial evidence
that Mrs. Parks’[] event was caused by a malfunction of the cruise control,” including evidence from Kress concluding
it “highly unlikely” for an “unintended acceleration event with a duration over 30 seconds and over a half-mile distance
to be . . . attributed to driver error or pedal misapplication.”)).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 20

ii. Evidence of an “Unreasonably Dangerous” Condition

Pivoting to the latter half of the defect and/or unreasonably dangerous condition prong of

the TPLA, Plaintiffs argue that a factual dispute exists as to whether the 2008 Kia Optima and its

parts were “unreasonably dangerous,” under Tenn. Code Ann. § 29-28-102(8). “Unreasonably

dangerous” means:

[A] product is dangerous to an extent beyond that which would be contemplated by
the ordinary consumer who purchases it, with the ordinary knowledge common to
the community as to its characteristics, or that the product because of its dangerous
condition would not be put on the market by a reasonably prudent manufacturer or
seller, assuming that the manufacturer or seller knew of its dangerous condition.

Id. (emphases added).

The statute’s plain language supplies two tests to evaluate if a product is unreasonably

dangerous: the consumer-expectation and the prudent-manufacturer tests. The consumer-

expectation test applies only to products in which the everyday experience of ordinary consumers

can fairly be employed. Ray v. BIC Corp., 925 S.W.2d 527, 531 (Tenn. 1996). Claims involving

more complex products are better evaluated under the prudent-manufacturer test in which expert

testimony is required.15 Under the prudent-manufacturer test, the buyer’s expectations are

irrelevant, and instead, courts ask whether a prudent manufacturer would have placed the product

in the stream of commerce. Brown v. Crown Equip. Corp., 181 S.W.3d 268, 282 (Tenn. 2005)

(quoting Ray, 925 S.W.2d at 532).

The two tests are neither mutually inclusive nor exclusive. Courts often (incorrectly)

approach the inquiry as if the tests were diametrically opposed. One reason for this confusion is

15
As a legal matter, expert testimony is only required under the prudent-manufacturer test: “In contrast to
the consumer expectation test, [under] the prudent manufacturer. . . . expert testimony about the prudence of the
decision to market [the product] would be essential.” Ray, 925 S.W.2d at 531; Sigler, 532 F.3d at 485 (explaining
that the prudent manufacturer test “obligate[s]” an injured party “to provide expert testimony”).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 21

that even where the consumer-expectation test might “technically[] apply” in cases involving

highly complex products, “it may be difficult for plaintiffs” to prove that product is unreasonably

dangerous. Jackson, 60 S.W.3d at 806. The dividing line between the two tests is not always

apparent, but the following caselaw marshals us in the right direction: Courts interpreting

Tennessee law have used the consumer-expectation test in cases involving seatbelts,16 tires,17

airbags,18 all-terrain vehicles,19 the third-row folding seats in sport utility vehicles and station

wagons,20 a hedge trimmer,21 and even where a restaurant failed to disclose its vegan pizza was

topped with pecan chips.22 Elsewhere, courts have found that the complexity of other products

precludes the use of the consumer-manufacturer test, thereby prompting the application of the

16
Jackson, 60 S.W.3d at 804 (“We are unwilling to accept the defendant’s argument that ordinary consumers
cannot form expectations about the safety and performance of seat belts.”).
17
Tatham v. Bridgestone Americas Holding, Inc., 473 S.W.3d 734, 751 (Tenn. 2015) (“We hold that a tire,
though vastly different in function, is comparable in its analysis to that of a seat belt.”).
18
Sigler, 532 F.3d at 486 (“Sigler offered evidence that an airbag is such a familiar product and that
consumers—and, indeed, manufacturers like Honda—have expectations about the product’s performance and
safety.”).
19
Whirley v. Kawasaki Motors Corp., USA, No. 1:04CV1145 T/AN, 2007 WL 9706819, at *9 (W.D. Tenn.
Feb. 21, 2007) (finding that the popularity and prolonged existence of ATVs, and the familiar design—i.e., four
wheels, suspension, disc brakes, and a wide wheel base—the average consumer possessed a degree of knowledge and
familiarity with the product’s performance).
20
Kines v. Ford Motor Co., No. 119CV01054JDBJAY, 2021 WL 3910323, at *10 (W.D. Tenn. Aug. 31,
2021) (explaining that such foldable seats “have been marketed in the United States for many years and the ordinary
consumer is familiar with the function and characteristics of such features.”).
21
Seaton v. Black & Decker (U.S.), Inc., No. 2:20-CV-124, 2021 WL 1395560, at *7 (E.D. Tenn. Apr. 13,
2021) (“There is testimony from Defendant’s witnesses that a consumer would know the battery is attached to the
hedge trimmer ‘[b]y looking.’”).
22
Jones v. WFM-Wo, Inc., 265 F. Supp. 3d 775, 779 (M.D. Tenn. 2017) (“An ordinary consumer purchasing
a slice of vegan pizza may well expect that it would not be topped with pepperoni, sausage or other meats, but the
Court cannot say, as a matter of law, that the same consumer would understand that the pizza slice could contain pecan
chips.”).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 22

prudent-manufacturer test; these cases have involved industrial forklifts,23 car radiators,24

automotive repair tools,25 medical bronchoscopes,26 steel rod passenger restraints on an

amusement park ride,27 a heater,28 and boom-truck cranes.29 It is important to note that in all these

cases, the courts found that the consumer-expectation or prudent-manufacturer test applied—not

necessarily that the plaintiffs carried their burden thereunder.

The district court did not employ any analysis under the “unreasonably dangerous” test,

preferring instead to focus on the causation prong of the TPLA. Plaintiffs now argue that a factual

dispute exists as to whether Defendants’ 2008 Kia Optima was “unreasonably dangerous” and that

material issues of fact exist precluding summary judgment to Defendants under either test. The

parties’ arguments under the consumer-expectation test and prudent-manufacturer test are assessed

below, in turn.

23
Brown v. Raymond Corp., 432 F.3d 640, 647 (6th Cir. 2005) (noting that an allegation that a forklift was
defectively designed is precisely the type of situation in which the ordinary consumer would not have an expectation
regarding the safety of a product).
24
Simpson v. O’Reilly Auto. Stores, Inc., No. 2:13-CV-2684-SHL-CGC, 2014 WL 11514969, at *6 (W.D.
Tenn. Dec. 30, 2014) (finding that radiators, while common, are “highly complex and difficult for plaintiffs to establish
as dangerous to an extent beyond that which would be contemplated by an ordinary consumer”).
25
Coffey v. Dowley Mfg., Inc., 89 F. App’x 927, 929 (6th Cir. 2003) (“The Super Hub Shark is a complex
tool. It is designed to be used for a variety of purposes . . . It is not at all obvious from looking at the tool or even at
diagrams showing it in use how it operates, much less how safe it is.”).
26
Young v. Olympus Am., Inc., No. 07-2547-STA, 2012 WL 252645, at *5 (W.D. Tenn. Jan. 26, 2012)
(“Plaintiffs have adduced no evidence that an ordinary consumer would have an expectation about the performance
of the bronchoscope at issue.”).
27
Alexander v. Zamperla, No. E200901049COAR3CV, 2010 WL 3385141, at *8 (Tenn. Ct. App. Aug. 27,
2010) (contrasting plaintiffs’ failure to prove that the safety system deviated from industry standards with defendants’
showing that it complied with industry standards and governing statutes).
28
Brewer v. Mr. Heater, Inc., No. 13-1330, 2014 WL 1364825, at *1 (W.D. Tenn. Apr. 7, 2014) (reasoning
that the prudent-manufacturer test was applicable given the product’s complexity).
29
Johnson v. Manitowoc Boom Trucks, Inc., 406 F. Supp. 2d 852, 858 (M.D. Tenn. 2005) (finding that “the
appropriate design of a boom truck crane and the safety features of such a crane are not within the ‘common knowledge
of laymen.’”).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 23

Under the “buyer[-]oriented” consumer-expectation test, a product is unreasonably

dangerous if it is “dangerous to an extent beyond that which would be contemplated by the ordinary

consumer who purchases it, with the ordinary knowledge common to the community as to its

characteristics.” Ray, 925 S.W.2d 527 at 530 (internal quotation marks omitted). “[A] product is

not unreasonably dangerous if the ordinary consumer would appreciate the condition of the product

and the risk of injury.” Id. The test is definitionally inapplicable to products about which an

ordinary consumer has no expectation. An ordinary consumer forms an assumption about products

that are commonplace, i.e., those products about which the “everyday experience of the product’s

users permits a conclusion[.]” Id. at 531 (quoting another source). “The manufacturer or seller’s

conduct, knowledge, or intention is irrelevant. What is determinative is what an ordinary purchaser

would have expected.” Id.

The Tennessee Supreme Court has made clear that “the consumer expectation test does not

depend necessarily on a product’s complexity in technology or use.” Jackson, 60 S.W.3d at 806.

Instead, the focus is on whether “prolonged use, knowledge, or familiarity of the product’s

performance by consumers is sufficient to allow consumers to form reasonable expectations of the

product’s safety.” Id. “Even a technically complex failure may involve a subject about which an

ordinary consumer may have an expectation, as discussed in Browder.” Coffey, 187 F. Supp. 2d

at 972.

Here, the question is whether the product at issue is one about which an ordinary consumer

could have any expectation. If yes, then the consequent question is substantive: whether Plaintiffs

have created an issue of fact that the product was unreasonably dangerous under regular use.

We find that ordinary consumers could form expectations about the safety and performance

of the cruise control and braking system in the subject 2008 Kia Optima, meaning that the
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 24

consumer-expectation test can be applied to the facts presented. Based on the ubiquity of cruise

control systems and acceleration and braking pedals in vehicles, their mainstay in everyday life,

coupled with the fact that brakes and accelerator pedals are some of the most elemental, required

components of cars, the average consumer possesses a degree of knowledge of and familiarity with

the product’s performance which not only exceeds that of industrial forklifts and boom truck

cranes, see Brown, 432 F. 3d at 647 (explaining an ordinary consumer does not have expectations

concerning a forklift); Johnson, 406 F. Supp. 2d at 857 (same, but for boom truck cranes), but is

sufficiently developed to form reasonable expectations about vehicular braking and acceleration

safety. This case is placed firmly in the long line of caselaw finding the consumer-expectation test

applicable to various pieces and parts of passenger vehicles. Sigler, 532 F.3d at 486; Tatham, 473

S.W.3d at 751; Jackson, 60 S.W.3d at 804.

One more question in the consumer-expectation test remains. To go forward on a claim

that a product is unreasonably dangerous requires the plaintiff to provide sufficient evidence to

create a question of fact that the product “[was] dangerous to an extent beyond that which would

be contemplated by the ordinary consumer who purchases it[.]” Tenn. Code. Ann. § 29-28-102(8).

Plaintiffs argue that the car suddenly accelerated un-commanded and failed to respond to

Parks’ application of the brakes; eyewitnesses recalled Parks’ efforts to warn and avoid hitting

other motorists, including “flashing” the lights, dodging and warning passersby to get out of the

way, and “zigging in and out of these cars,” (Philpot Dep., R. 317-20 at PageID # 11639); and

Parks’ repeated statements before and after the crash that she was unable to stop the car and that

there was something wrong with the vehicle that commanded its acceleration.

In combination, this evidence is enough to get to a jury under the consumer-expectation

test. Of the three categories of “ordinary consumers” Plaintiffs identify, i.e., the witnesses to the
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 25

accident, Parks herself, and Parks’ sister, Plaintiff Northcutt, Parks’ comments are especially

probative. Recall Parks stated that the car “had a mind of its own” and that she attempted to apply

the brakes, to no avail. Parks certainly had an expectation that her car would not have accelerated

quickly without her intent, and that is enough to establish a consumer’s expectations. Parks’

statements also point to proper handling and weaken the defense retort of pedal misapplication.

Accordingly, the district court should have found that the consumer-expectation test applies here;

application thereof supplies a genuine issue of material fact of whether the Optima was

“unreasonably dangerous” under the TPLA’s consumer-expectation test.

Just as under the consumer-expectation test, the district court did not provide any analysis

on the prudent-manufacturer test. The prudent-manufacturer test will now be discussed.

Liability under the seller-oriented prudent-manufacturer test attaches when a product

would not be put on the market by a reasonably prudent manufacturer or seller because of its

dangerous condition. The manufacturer is presumed omniscient for purposes of this test, meaning

that knowledge of a product’s potentially dangerous condition is imputed. Ray, 925 S.W.2d at

531. Expert testimony about the prudence of the decision to market is essential. Id.

Arguing by omission, Plaintiffs declare that the Parks’ 2008 Kia Optima was unreasonably

dangerous under the prudent-manufacturer test because of what the manufacturer failed to do

(conduct a system-level Failure Modes and Effects Analysis) and what the vehicle was missing

(brake throttle override). Plaintiffs attribute this argument to two individuals: one of their

remaining experts, Loudon, and Defendant Kia’s corporate representative, Pan Sang Kim, an

engineer who designed and developed ECUs. Understanding three engineering concepts—FMEA,

ETC, and brake pedal throttle override system—is required to assess Plaintiffs’ argument.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 26

A Failure Modes and Effects Analysis, or “FMEA,” is an engineering process that tests

and refines designs to improve reliability. In the automobile context, an FMEA detects and

anticipates automotive issues. FMEAs can be limited to a specific component or broadened to

assess a product’s design more comprehensively.

An electronic throttle control, or “ETC,” is a car’s onboard computer. Imagine a traditional

vehicle where a mechanical link—usually, a cable in the carburetor—connects the accelerator

pedal and an engine. The ETC cuts out the cord and replaces it with a complex computer system.

When a driver of a 2008 Kia Optima presses the accelerator, that command makes its way to the

ETC by a single wire. The ETC tells the motor to accelerate and how quickly to do so (or, in

engineering terms, the ETC tells the throttle to open and by how much). When a driver of a 2008

Optima presses any of the cruise-control buttons on the steering wheel (i.e., set, resume, accelerate,

or cancel), that command flows through that same single wire.

A brake throttle override is an engine management software that closes the throttle upon

unintended acceleration. The term is otherwise self-explanatory, except for one important detail:

override systems bring a car’s engine back to idle only when the brake and accelerator pedal are

depressed simultaneously, thereby allowing a driver to stop her vehicle even if the accelerator

pedal is stuck. It is unclear whether Parks’ vehicle had a brake throttle override, a question of fact

that alone might counsel against summary judgment.30

30
Plaintiffs say that the car did not but argue it could and should have been reprogrammed into the Parks’
2008 Kia Optima’s ECU at little cost to Kia. Defense expert Walker also appears to doubt Parks’ car had the
technology; he writes: “In model year 2008, approximately 70% of all passenger cars and light trucks sold in the U.S.,
including the 2008 Kia Optima, were not equipped with a brake-throttle override system.” (Walker Report, R. 317-
15, PageID # 11322). Conversely, Defendants state in their appellate briefing that at least some 2008 Kia Optimas,
including the exemplar vehicle Loudon tested, “have a brake override system.” (Defs.’ Br., ECF No. 36 at 33).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 27

With this technical backdrop established, Plaintiffs rely on Loudon and Kim for the

contours of their prudent-manufacturer argument. From Loudon, Plaintiffs advance the following:

a reasonably prudent manufacturer would not market a vehicle without conducting a system-level

FMEA; if Defendant-manufacturer had done so here, it would have concluded that a single signal

wire for the cruise control had an unacceptable level of risk. From Kim, Plaintiffs infer that a

system-level FMEA was likely never conducted; perhaps even worse still, as Plaintiffs see it, is

that Kim testified it was feasible to reprogram and retrofit the electronic control unit (“ECU”) in

the Parks’ 2008 Kia Optima with a brake override feature.31 It is claimed that Kim’s testimony

demonstrates that the Kia Defendants had actual notice of safety technology that would have

prevented the 2015 accident.

There is an issue of fact as to whether the design of the car’s control and safety systems

was unreasonably dangerous if, as Plaintiffs allege, the manufacturer acted unreasonably in

marketing the vehicle with its then-existing design. See Brown, 432 F.3d at 647. Assuming

Plaintiffs are correct that Parks’ Optima lacked a brake override feature, Defendants’ failure to

retrofit the vehicle with it as an aftermarket feature also counsels against summary judgment under

the prudent-manufacturer test.

c. Proximate Cause

The primary basis on which the district court granted Kia’s motion for summary judgment

was that it found no admissible evidence showing how a malfunction in the Optima’s cruise control

31
The transcript of Kim’s deposition is poorly scanned with parts of it unreadable. The bit that is readable
indicates that Kim was never provided with or reviewed FMEAs. When asked if KMC “conduct[ed] any sort of
feasibility analysis of retroactively reprogramming the ECU in the 2008 Kia Optima 2.4-liter engine in order to put
the Smart Pedal feature into those vehicles,” Kim responded: “I can’t say for the entire company whether there was
an analysis on the company level[;] I myself did not do an analysis.” (Kim Dep., R. 234-3, PageID # 5458). Kim
agreed that he was “[not] familiar with any discussions . . . regarding the retroactive implementation of the Kia Smart
Pedal brake throttle override technology.” (Id.).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 28

could have caused unintended acceleration. The district court erred in finding that Plaintiffs have

not established a dispute of material fact as to causation.

Animating Plaintiffs’ pleadings is a concern that a heightened standard of proof would

make it all-but-impossible to establish proximate cause where an electronic defect leaves no trace

of a malfunction’s occurrence. It is generally accepted that automobile electronics, unlike some

mechanical defects, can fail without a trace, at least as a technical or theoretical matter. See, e.g.,

Johnson v. Ford Motor Co., No. CV 3:13-6529, 2018 WL 1512377, at *2 (S.D.W.Va. Mar. 26,

2018) (“[T]here are transient defects which cannot be replicated and leave no trace.”). The

possibility of those injured by untraceable electronic vehicular defects being shut out of the

courthouse because they are unable to prove causation through direct evidence is obviously

undesirable; however, even assuming the existence of a traceless electronic defect, Plaintiffs are

not relieved from proving causation at all. See Whaley v. Rheem Mfg. Co., 900 S.W.2d 296, 300

(Tenn. Ct. App. 1995) (“It almost goes without saying that the identified product defect must be

the proximate cause of the plaintiff’s injury.”). What varies is the type of proof courts find

persuasive.

To constitute proximate cause, the cause “must be such that had it not happened[,] the

injury would not have been inflicted.” Shouse v. Otis, 448 S.W.2d 673, 676 (1969). “What is

meant by proximate cause is not necessarily that which is next or last in time or place, but that

which is a procuring, efficient[,] and predominate cause. Closeness in causal relation, rather, is

the meaning.” Nash v. Love, 440 S.W.2d 593, 598 (1968) (citation omitted). Circumstantial

evidence, like proof of a defect or unreasonably dangerous condition, may be used to rely upon to

prove proximate causation. Gable v. Tennessee Liquefied Gas Co., 325 S.W.2d 657, 665 (1957).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 29

In this case, the question is, even if the malfunction (assuming one existed) left no traceable

footprint, have Plaintiffs shown that the accident would not have happened if the malfunction had

not been triggered. Because the disputed issues of fact are apparent on the trial court record, the

district court erred in granting summary judgment in favor of Kia. The path to this conclusion

requires discussion of Plaintiffs’ six-fold proximate causation argument.

At least four of these arguments (the post-crash pedal position and half-inch gap;

eyewitness testimony; Loudon’s exemplar testing; and Kress’ theory of driver correction of pedal

misapplication) were the subject of this Court’s earlier analysis. A jury could find each of the six

arguments demonstrates that a vehicular defect caused or contributed to the accident.

First, the district court found that the post-crash pedal position and half-inch gap between

the accelerator pedal and its bracket was not circumstantial evidence that an unspecified cruise-

control malfunction caused the crash. Below and again on appeal, Plaintiffs assert that the pedal

position and the gap indicate that Parks might not have been pressing the accelerator pedal at the

time of the crash and that the damage to the pedals was consistent with her right foot being on or

off the brake pedal. If so, this weakens the defense suggestion of pedal misapplication, bolstering

the possibility of a vehicular malfunction. Summary judgment was not proper because a

reasonable jury could infer from the pedal position and the gap that Parks’ foot was not on the

accelerator pedal.

Second, the district court is correct that the eyewitness testimony does not establish how a

cruise control malfunction could cause an unintended acceleration; however, the eyewitness

testimony might amount to circumstantial evidence that the vehicle was defective or unreasonably

dangerous for ordinary use, leading to an inference of proximate causation. One eyewitness

testified as follows: “That’s the only words [Parks] said to me . . . [‘]sir[,] I could not stop.[’]”
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 30

(Bill Hasty Dep., R. 321-6, PageID # 12484). Recall Plaintiff Northcutt testified that her sister

stated: “Jimmie, there’s something wrong with this car[,]” and that “the car . . . would not stop.”

(Northcutt Dep., R. 317-19, PageID ## 11603, 11612). A jury could credit the various eyewitness

testimony to conclude that an unspecified malfunction was the source of the sudden acceleration

rather than driver error.

Third, the district court determined that Loudon’s cruise-control signature testing failed as

circumstantial evidence that a non-specific cruise control malfunction caused the crash. The

necessary presupposition, the district court said, was missing, i.e., “that an unintended cruise-

control-induced acceleration is possible.” (Order, R. 341, PageID # 13040). Our reading of the

record is different. That an unintended cruise-control-induced acceleration is possible is what

Loudon contends. When deposed, Loudon stated: “I understood that primarily my role in [this

case] was to determine if a failure of the cruise control system could lead to the circumstances of

this accident.” (Loudon Dep., R. 311-2, PageID # 7194). His report also does more than just rely

upon Sero’s and Bloch’s excluded testimonies. He proffers independent evidence that his testing

of the exemplar vehicle coupled with Parks’ vehicle’s digital footprint indicated “that this accident

was much more likely caused by the cruise control system engaging the resume/accel function and

accelerating continuously until the vehicle struck another vehicle.” (Loudon Report., R. 317-5,

PageID # 10263). A jury could find from Loudon’s report that a faulty component of the cruise

control, particularly via the resume/accel button, caused the sudden acceleration here.

Fourth, the district court deemed wanting Plaintiffs’ argument, derived from Kress’ human-

factors report, that pedal misapplication is unlikely to persist for the amount of time that the 2008

Kia Optima accelerated. Kress cited literature indicating that a driver recognizes and corrects a

pedal misapplication within 1.5 seconds. The district court is correct that this is not direct evidence
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 31

of how the cruise control could cause unintended acceleration. But Kress’ report demonstrates

circumstantially that most drivers would, almost reflexively, apply the brakes within 1.5 seconds

of an accidental acceleration event. Parks’ car accelerated for over thirty seconds at speeds over

90 miles per hour. This circumstantial evidence is sufficient to create a jury question and avoid

summary dismissal because a jury could reasonably infer that something other than prolonged

pedal misapplication was the source of the acceleration.

Fifth, the district court held that Plaintiffs failed to connect the prior occurrences of cruise-

control Diagnostic Trouble Codes (“DTCs”) in Parks’ vehicle to any prior un-commanded, sudden

acceleration. To understand the district court’s holding, some background on DTCs is helpful.32

DTCs are error codes. The Kia Optima can display any five codes, some of which appear on the

malfunction indicator lamp (“MIL”) to warn drivers of an issue or malfunction, and some of which

do not. DTCs are either “current” or “historical.” A “current” DTC reveals any current issue with

a vehicle, and “historical” ones indicate any past issues. When the engine control module (“ECM”)

detects a consistent fault in the cruise control switch unit for an established duration, DTC P0564

is triggered. The P0564 error code does not generate a MIL to warn drivers of the issue.

In this case, there were no current DTCs registered during Parks’ last seven driving

records. But the Kia’s computer system (called the Powertrain Control Unit, or PCU) recorded

six historical DTC P0564s in Parks’ vehicle. The “04” refers to a stuck “Resume/Accel” switch.

Thus, those six instances of DTC P0564 mean that the cruise control RES/ACCEL button—located

on the right side of the steering wheel on the Kia Optima—was stuck or depressed for more than

32
The dissent says “the data [from the National Highway Traffic Safety Administration post-crash
investigation report] revealed no ‘trouble codes’ from the car’s onboard computer.” (Dissent at 43). But the dissent
fails to consider that the NHTSA downloaded data only from the vehicle’s airbag control unit. (NHTSA Report, R.
204-5, PageID ## 4343–48). In fact, a subsequent data-pull revealed multiple diagnostic trouble codes downloaded
from the PCU.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 32

sixty seconds at some point predating the December 2015 accident. Plaintiffs emphasize that

P0564 is not triggered unless the stuck condition persists for at least sixty-one seconds; as a result,

it is asserted that such an electrical malfunction might have occurred during the 2015 accident that

was not registered if the button was stuck for less than sixty-one seconds. Taking Plaintiffs’

allegation as true that a vehicle could accelerate to dangerous speeds before a DTC is detectable,

this evidence suggests either a chronically stuck resume/accel switch (which itself might command

an acceleration) or otherwise that the historical DTCs show that some electrical malfunction must

have occurred in the vehicle.

Sixth and finally, the district court found that the depletion of vacuum assist in Parks’

vehicle did not evidence that the cruise control can cause unintended acceleration, let alone the

one at issue. We disagree. A depleted vacuum would have made Parks unable to slow the vehicle

with the braking pedal due to the wide-open throttle condition. Assuming, as Plaintiffs argue, that

Parks attempted to slow the vehicle with the brake pedal during a wide-open throttle condition but

was unable to do so due to the depletion of vacuum assist, then, once again, the defense argument

of pedal misapplication is undermined.

These six theories are circumstantial evidence that an unspecified malfunction caused the

crash. We conclude that Plaintiffs have offered sufficient evidence from which a reasonable jury

could find that a defective or unreasonably dangerous condition of the Parks’ 2008 Kia Optima

proximately caused the accident.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 33

b. Negligence

Plaintiffs assert that the district court erred in granting summary judgment on the theory of

negligence since there are genuine issues of material fact that create a reasonable inference of Kia’s

negligence. For similar reasons explored above, we agree.33

Even though the burden of proof does not shift based on different theories of negligence

under the TPLA, Plaintiffs segregate their negligence claim from those of strict liability, done so

presumably to couch strict liability as a recovery theory different from negligence. Regardless of

which theory of recovery is sought, however, Plaintiffs still must prove that the vehicle suffered

from a defect or was unreasonably dangerous, which proximately caused the injuries. Claims of

negligence under the TPLA are well steeped in Tennessee law. The general rule is that liability

attaches for negligence in the manufacture or sale of a product where the product may be

reasonably expected to cause injury by its negligent design or construction. Dunn v. Ralston

Purina Co., 38 Tenn. App. 229, 233–34 (1954).

Mirroring their prudent-manufacturer argument, Plaintiffs identify four deficiencies

appearing to evince Defendants breached a duty to exercise reasonable care: (1) the lack of the

brake override; (2) failure to conduct a system-wide FMEA; (3) design defects in the DTC system;

and (4) failure to warn. Each argument is, to varying degrees, duplicative of those discussed in

the prudent-manufacturer section, and our findings rise and fall with those discussed therein. The

33
Plaintiffs invoke res ipsa loquitur to supply a reasonable inference of negligence. Tennessee, like all states,
permits negligence liability under the res ipsa loquitur doctrine. In permitting claims to go to a jury because “the
thing speaks for itself,” the TPLA allows a jury to presume negligence where “the accident is such as in the ordinary
course of things does not happen if those who have the management use proper care, it affords reasonable evidence,
in the absence of explanation by the defendants, that the accident arose from want of [proper] care.” Sullivan v.
Crabtree, 258 S.W.2d 782, 784 (1953) (citing another source). Plaintiffs failed to raise res ipsa loquitur in their
response to the motion for summary judgment, so they cannot raise it for the first time on appeal, and this argument
is not preserved for appellate review.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 34

question is whether a genuine issue of material fact exists as to whether Defendants’ negligence

was the proximate cause of Plaintiffs’ injuries. See Leatherwood v. Wadley, 121 S.W.3d 682, 694

(Tenn. Ct. App. 2003). Should it credit Plaintiffs’ evidence, a reasonable jury could find that

Defendants’ negligence was the proximate cause of the injuries.

First, Plaintiffs contend that despite the feasibility and availability of a brake override

system, Defendants negligently failed to install this failsafe system in its 2008 Kia Optima either

in the original design or as an aftermarket feature. The district court, finding Plaintiffs’ case failed

on causation, omitted any meaningful discussion of the brake-override argument as to negligence.

On appeal, Plaintiffs suggest a jury could find that the absence of the brake defect system made

the vehicle unsafe for regular use. We agree. Relying upon Loudon, Plaintiffs introduced the

following evidence that is sufficient to get to a jury: “If the Parks vehicle had” been equipped with

the brake override, “this accident would never have happened.” (Loudon Report, R. 317-5, PageID

# 10270). Even if an ordinary consumer does not harbor a precise expectation concerning the

initial or retrofitted installation of a brake override system, there certainly would be a consumer

expectation that acceleration would stop upon depression of the brakes. Drawing all justifiable

inferences in favor of Plaintiffs, this Court finds that a genuine issue dispute of material fact exists

as to whether Defendants were negligent in failing to install a brake override system.

Second, Plaintiffs, relying upon Loudon, argue that Defendant KMC was negligent in

failing to conduct a system-wide FMEA. If it had, Plaintiffs posit the manufacturer would have

found the use of a single wire in the cruise control system carried an unacceptable level of risk.

This gets to the very core of the case, namely, whether some unidentified defect in the cruise

control caused the December 2015 accident. If Plaintiffs’ argument is to be credited—i.e., that

Defendants should have, but failed to, conduct a system-wide FMEA, which would have detected
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 35

the cruise-control’s susceptibility to command an un-commanded acceleration—then there is a

dispute of fact appropriate for jury consideration.

Third, Plaintiffs submit a design defect argument related to the DTCs. This argument,

drawn primarily from Loudon, requires some patience insofar as it implicates several technical

details. The essential point is that Plaintiffs argue that Parks’ 2008 Kia Optima lacked an adequate

DTC-detection system that was robust enough to anticipate or respond to unintended acceleration.

Plaintiffs say that the design of the 2008 Kia Optima’s cruise control system and its components

(including the clock spring coil, the cruise control switch, the single signal wire for cruise control

operations, and the interconnects between the switch and the ECM input) were negligently

designed. Loudon’s report concluded that an electrical malfunction could have been introduced

into Kia’s electronic throttle control system without recording a DTC, meaning that the system

does not realize or register that a problem occurred. The consequence, if the premise is to be

accepted, would be that an undetected signal could cause the engine’s throttle to open wide without

driver input. As a result, it is asserted that Parks’ vehicle should have had a better-designed failsafe

mode that ignored errant signals, such as an inadvertently opened throttle.

Recall from this Court’s earlier discussion that DTCs related to cruise control can manifest

any five codes. Six historical instances of the P0564 were recorded in Parks’ 2008 Kia Optima,

indicating a “stuck” cruise control “resume/accel” switch. All that error code means is that the

ECM has detected a fault in the cruise control switch, or, in more technical terms, the ECM

detected a signal from the cruise control switch that exceeded the threshold value.

Distilling Plaintiffs’ arguments from the trial-court level and those on appeal, we deduce a

few sub-categories related to the DTC-negligent-design-defect argument.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 36

The first one is temporal, and based on Loudon’s report, that is, the DTCs take too long to

be detected, thereby allowing an unintended acceleration event to occur before a trouble code can

be spotted by a driver or technician. If Plaintiffs are correct that a vehicle can accelerate to

dangerous rates of speed before a DTC is noticed, then a reasonable jury might be able to find that

the historical instances of P0564 in the Optima demonstrated the car suffered from some

acceleration-related malfunctions.

The next sub-argument is a more traditional design-defect argument. Via Loudon,

Plaintiffs argue that the use of a single wire in the cruise-control design makes it difficult for the

system to distinguish between a legitimate press of a cruise control button and one activated in

error. This argument is meritorious and will survive summary dismissal. If the system is unable

to distinguish between a stuck “resume/accel” button, an electrical switch signaling problem, or a

driver accidentally pressing the cruise control switch, then finding that the system could command

an acceleration without driver input is a supported inference.

Finally, we turn to the last category in Plaintiffs’ negligence argument. Plaintiffs contend

that Kia failed to warn of the dangerous flaw in the Parks’ 2008 Kia that allowed a runaway throttle

condition to occur on December 31, 2015. The district court did not assess the failure to warn

claim. In Tennessee, a product is not unreasonably dangerous because of a failure to adequately

warn of a danger or hazard that is apparent to the ordinary user. Tenn. Code Ann. § 29-28-105(d).

For a failure-to-warn claim under the TPLA, “[a]n adequate warning is one calculated to bring

home to a reasonably prudent user of the product the nature and the extent of the danger involved

in using the product . . . . The adequacy of the warning is a question for the jury unless reasonable

minds could agree on the outcome.” Evridge v. Am. Honda Motor Co., 685 S.W.2d 632, 636–37

(Tenn. 1985) (internal quotation marks and citations omitted). The Tennessee Supreme Court held
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 37

last year “that the language of the TPLA and accompanying case law places a duty to warn on a

manufacturer or seller to warn about the condition of the product only if it was defective or

unreasonably dangerous at the time the manufacturer transfers control of the product.” Coffman

v. Armstrong Int’l, Inc., 615 S.W.3d 888, 896 (Tenn. 2021). Because we find Plaintiffs

demonstrated that the product might have been defective or unreasonably dangerous, the duty to

warn is at least implicated. Plaintiffs have also introduced evidence indicating that Defendants

knew of feasible design options that would have prevented the accident here. A jury reasonably

could find that Defendants were negligent in failing to warn of known defects. Accordingly, we

reverse the district court’s grant of summary judgment based on negligence.

III. CONCLUSION

For reasons discussed above, we REVERSE the district court’s holding as moot the

motions to exclude Kress and Loudon, REVERSE the district court’s order granting summary

judgment to Defendants, and REMAND the case for further proceedings consistent with this

opinion.34

34
In addition to being intemperate in its tone and use of language, the dissent inappropriately attempts to
advise the district court on remand to ignore this Court’s remand instructions.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 38

GIBBONS, Circuit Judge, concurring in the judgment. After the district court excluded

the testimony of plaintiffs’ experts Samuel Sero and Byron Bloch, it summarily determined that

defendants’ motions to exclude the expert testimony of Steve Loudon and Tyler Kress were moot.1

It did so based on its understanding that plaintiffs did not contend that either Loudon or Kress

offered testimony relating to whether a specific defect in Mary Parks’s 2008 Kia Optima caused

the fatal crash. That conclusion led to the district court’s failure to consider whether Loudon’s

and/or Kress’s testimony could be used to establish defect as well as causation. This failure was

an abuse of discretion. See Tahfs v. Proctor, 316 F.3d 584, 593 (6th Cir. 2003) (“Abuse of

discretion is defined as a definite and firm conviction that the trial court committed a clear error

of judgment.” (citation omitted)). On remand, the district court must first analyze the defendants’

motions to exclude Loudon and Kress on the merits. Then, the district court should determine

whether summary judgment is appropriate. I therefore agree with the majority opinion to the extent

that it reverses the district court’s holding as to the mootness of the Daubert motions regarding

Loudon and Kress and vacates the district court’s grant of summary judgment.

In his dissent, Judge Bush asserts the district court was correct to consider the motions to

exclude moot because plaintiffs failed to argue to the district court that Loudon’s testimony

established a specific defect. He describes this as a “waiver.” Dissent, at 12, 18. My reading of

the record differs. While plaintiffs did state that Loudon “was not tasked with replicating a defect,”

1
Appellees state in their briefing: “Taking Plaintiffs at their word, and after independently reviewing the
testimony of Kress and Loudon, the district court concluded that neither expert provided evidence of a specific defect
in the Optima and, therefore, the motions to exclude their testimony were moot on the dispositive questions.” CA6
R. 36, Appellee Br., at 22. From its order, we cannot discern whether or not the district court “independently” reviewed
the testimony. While it may have done so, it may also have based its ruling on its understanding of plaintiffs’ initial
position. A review of Loudon’s report reveals he offers an opinion as to defect, as well as opinions bearing on
causation.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 39

they also, citing a specific portion of Loudon’s report, recited his conclusion that “[t]he design of

the cruise control system is defective in that it uses a single wire through an unreliable clockspring

harness.” DE 322, Resp. in Opp’n to Exclude Loudon, Page ID 12635, 12644. Arguing that a

single wire system “is a design flaw because it enables a single point failure to open the throttle

contrary to the driver’s intent without providing a failsafe mode,” plaintiffs used Loudon’s

testimony to establish a specific defect—the use of a single wire in the cruise control system. Id.

at 12640. Plaintiffs also argued that Loudon’s testing and report established causation, asserting

“there was no scenario that could explain the Parks event other than a cruise control-induced

acceleration.” Id. at 12637. Because plaintiffs did argue that Loudon’s testimony established a

specific defect (the use of a single wire) and causation, the district court abused its discretion in

denying as moot defendants’ motions to exclude Loudon and Kress.2

Reversing the district court’s holding that the motions to exclude Loudon and Kress are

moot results in remand for the district court to analyze those motions on the merits. After

considering the admissibility of Loudon’s and Kress’s testimony, it is the district court’s role to

determine whether to grant summary judgment.

2
It is correct to say that plaintiffs’ original plan was to rely on the testimony of Bloch and Sero to establish
defect. But it is also an incomplete characterization of their position to suggest that they conceded that Loudon’s
testimony could not be characterized as proof of a specific defect.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 40

BUSH, Circuit Judge, dissenting. The circumstances of this case are tragic. And there is

an understandable desire to place blame for a crash that led to three fatalities not on an elderly

driver but on something about the car itself. Under the Erie doctrine, however, we are bound to

apply the law of Tennessee to this dispute. See Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78 (1938);

see also Lindenberg v. Jackson Nat’l Life Ins. Co., 919 F.3d 992, 995–96 (6th Cir. 2019) (Bush,

J., dissenting from the denial of rehearing en banc). For liability to arise, Tennessee requires that

plaintiffs do more than simply point to the crash and argue that it wouldn’t have occurred but for

some unspecified defect. Rather, to survive summary judgment, plaintiffs must identify the specific

defect that supposedly caused the crash. See, e.g., Benson v. Tenn. Valley Elec. Coop., 868 S.W.2d

630, 636 (Tenn. Ct. App. 1993) (“[I]t is vital to trace the injury to some specific error in

construction or design of the machinery[.]” (quoting Browder v. Pettigrew, 541 S.W.2d 402, 404

(Tenn. 1976) (emphasis in original)); King v. Danek Med., Inc., 37 S.W.3d 429, 435 (Tenn. Ct.

App. 2000) (explaining that a products-liability plaintiff must “trace [her] injury to the specific

defect”).

My colleagues have split concerning how we should apply that standard to this case. The

lead opinion acknowledges that plaintiffs must identify a specific defect. Lead Op. at 15. Yet it

suggests reversing the district court on the theory that plaintiffs could show some “unspecified”

and “non-specific” defect on remand, based on arguments that plaintiffs never made below. See

id. at 18, 29–32; see also infra 55–56. The lead opinion’s reliance on some “unspecified,” “non-

specific” defect to satisfy what is concededly a specific-defect standard betrays the fact that

plaintiffs have presented no admissible theory about any specific malfunction that caused the car

to accelerate. Nor, as the lead opinion suggests, can plaintiffs prevail based on res ipsa loquitur.

The lead opinion correctly notes that this theory of relief is forfeited because plaintiffs never
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 41

argued it in their response to defendants’ motion for summary judgment. See id. at 33 n.33. Yet

the lead opinion suggests reliance on Connecticut law to apply what it admits is a res ipsa loquitur

theory—a theory plaintiffs have never even mentioned—to suggest summary judgment was

improper. See id. at 18–19. Respectfully, the lead opinion’s reasoning on both issues is not only

self-contradictory but a plain misapplication of Tennessee law.

Unlike the lead opinion, the opinion concurring in the judgment appears to recognize that

proof of a specific defect requires, indeed, proof of a specific defect. Concurrence at 39. But it

suggests that plaintiffs below proffered a theory of specific defect via expert witness Steven

Loudon. Id. Much like the lead opinion, the concurrence misreads the lower-court record. In their

briefing below, plaintiffs never argued that Loudon or their other unexcluded expert, Tyler Kress,

could explain any specific defect that caused the car’s supposed malfunction. To the contrary,

plaintiffs explicitly waived reliance on Loudon and Kress to provide a specific-defect theory. Yet

for reasons left unexplained, both the lead opinion and the concurrence reject those explicit waivers

and indulge plaintiffs’ about-face on appeal, transmogrifying Loudon and Kress from sideshows

to star witnesses. On that basis, those opinions vacate the district court’s order and remand to it a

palimpsest of a case on which the district court itself never had the chance to pass judgment. None

of that is consistent with Rule 56, the law of our circuit, basic requirements of issue preservation,

respect for the district court, or due process for defendants. So I respectfully dissent.

I.

In December 2015, a high-speed collision occurred in Winchester, Tennessee. Mary Parks,

age eighty-three, was driving a 2008 Kia Optima along the Dinah Shore Boulevard. Her sister,

Jimmie Ruth Northcutt, was in the passenger’s seat. They intended to stop at a nearby Kroger to

buy some milk. But near the Kroger’s entrance, Parks’s vehicle began to accelerate. Parks told
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 42

Northcutt that something was wrong and that the car wouldn’t stop. Yet Parks never specifically

stated that her foot was on the brake,1 and Kia says that Parks mistakenly put her foot on the

accelerator.

As the car gained speed, Parks swerved to avoid traffic and flipped her headlights on and

off, apparently trying to warn surrounding motorists. Then, after reaching ninety-two miles per

hour, Parks and Northcutt crashed into the back of the Hill family’s minivan. The impact killed

the Hill family’s twin children and left Parks mortally wounded. At the scene, Parks told a witness

that she tried to stop the car but that it “had a mind of its own.” She died the next day.

No one disputes that the crash was horrific. And no one disputes that it was unintentional.

The only issue is whether Parks’s Optima possessed some specific defect or dangerous condition

that caused the acceleration, or whether Parks caused it herself by confusing the gas for the brakes.

Arguing the latter point, Kia says the collision stemmed from “pedal misapplication”—a well-

documented phenomenon in which drivers, particularly elderly drivers like Parks, mistake the two

pedals. See NTSB Report at 36, R. 143-4; id. at 40 (“Older drivers are generally more susceptible

to conditions that diminish cognition, perception, and physical reaction.”). Relevant here, pedal-

misapplication events are particularly common during “turning movements.” Id. at 37. So, Kia

says, Parks might have thought she was braking to turn into the Kroger’s entrance when she

1
Citing Northcutt’s deposition, the lead opinion claims that Parks “was unable to stop accelerating, despite
apparently applying the brakes.” Lead Op. at 3. Northcutt’s deposition, however, does not support that claim. On the
page of Northcutt’s deposition cited by the lead opinion, Northcutt says nothing about whether Parks was using or
claiming to have used the brakes. In fact, when Northcutt was asked, “[D]o you remember your sister try [sic] to stop
the car with the brakes?” Northcutt responded, “No, I don’t remember that.” Northcutt Dep. at 22:18–20, R. 317-19.
Additionally, when Northcutt was later asked, “Do you remember your sister saying anything, other than there’s
something wrong with the car?” Northcutt responded, “No.” Id. at 23:23–24:1.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 43

unintentionally stepped on the gas pedal. She then kept depressing the gas to “correct” the

subsequent acceleration—thinking she was braking—and gained speed until the collision.

Yet Parks maintained in her final moments that the Optima had caused the crash by itself.

That claim triggered an investigation by the National Highway Traffic Safety Administration

(“NHTSA”) into the circumstances of the crash. NHTSA conducted an on-site investigation,

examined security-camera footage taken shortly before the crash, and reviewed information from

the Optima’s event data recorder (“EDR”) to see whether it contained any record of an electronic

malfunction. NHTSA also thoroughly inspected Parks’s Optima, assessing its brakes, brake lights,

gas pedal, floor mats, and interior and exterior damage.

The inspection revealed “no evidence of pre-crash braking.” NHTSA Crash Report at 4333,

R. 204-5. The car’s brakes showed no signs of burnishing from heavy application. Id. at 4327.

Likewise, if the Optima’s brake lights had been illuminated during the crash, NHTSA explained,

their filaments likely would have been stretched or fractured by the “sudden deceleration” on

impact. Id. at 4327. Yet the filaments were “intact,” leading to the conclusion that “the brake lights

were not on pre-crash and that the driver was not applying the brakes in an attempt to slow the

Kia.” Id. That conclusion accorded with the security-camera footage depicting the last hundred

feet of travel, in which Parks’s brake lights were unilluminated. Id. at 4322. And the EDR data

revealed no “trouble codes” from the car’s onboard computer that might have suggested a

malfunction caused the crash. Id. at 4328. NHTSA thus concluded that the crash had no apparent

“root cause . . . other than inadvertent driver input”—in other words, that Parks had stepped on the

wrong pedal after all. Id. at 4321.

For the past several years, plaintiffs (now appellants) have tried to rebut those conclusions

and explain why the Optima could have caused the crash on its own. Their case eventually
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 44

coalesced around four experts they proposed to introduce at trial: Byron Bloch, Samuel Sero, Tyler

Kress, and Steven Loudon. Kress and Loudon, as explained more fully below, were not tasked

with propounding any specific mechanical or electronic defect with the Optima. Instead, Kress, a

so-called “human factors” expert, proposed to testify from a “biomechanical” standpoint about

why it was more likely that Parks’s foot was on the brake rather than the gas.2 Loudon, for his

part, became involved by virtue of experiments he ran with an exemplar Optima on a closed track

after plaintiffs had hired him as a witness. Loudon’s experiments hinged on the fact that the

tachometer in Parks’s Optima had frozen at 4,300 revolutions per minute (“RPM”) at the moment

of impact. On the theory that Parks’s crash was somehow related to the Optima’s cruise control,

he performed test runs in which he reached ninety-two miles per hour in the exemplar vehicle first

by pressing the gas and then, in a different test run, by pressing the “accelerate” button on the

steering wheel. According to plaintiffs, his experiments showed that reaching ninety-two miles per

hour with the cruise-control button (rather than the gas pedal) resulted in RPMs that more closely

matched the crash-vehicle’s 4,300-RPM figure.

Those respective contributions might have been useful in an ancillary sense. Viewed in the

most favorable light, each may support an inference that some unspecified malfunction occurred

in Parks’s Optima. But such an inference would be manifestly insufficient to get to trial under

Tennessee tort law. Rather, as both the lead opinion and the concurrence acknowledge, Tennessee

requires that products-liability plaintiffs propound evidence about what specific condition with the

product rendered it defective or unreasonably dangerous and proximately caused the plaintiff’s

2
He eventually admitted that he could not say with certainty where her foot had been. See Kress Dep. at
293:6–294:14, R. 317-11 (admitting that Kress had no evidence “either way” concerning whether Parks’s foot was on
the gas or brakes).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 45

injuries. See Lead Op. at 15 (“A plaintiff must prove a specific defect under the TPLA, regardless

of the theory of liability.”); Concurrence at 39 (claiming that “Loudon’s testimony established a

specific defect”); see also Benson, 868 S.W.2d at 636 (“[I]t is vital to trace the injury to some

specific error in construction or design of the machinery[.]” (quoting Browder, 541 S.W.2d at 404

(emphasis in original)). So plaintiffs needed some experts who could testify about what specific

malfunction supposedly occurred in the Optima that produced the unintended acceleration. See id.

For that task, plaintiffs enlisted two different witnesses: Byron Bloch and Samuel Sero.

Sero earned an undergraduate engineering degree in 1967 and, since 1989, has held himself out as

a “forensic consultant” willing to testify about everything from elevators to aquariums. Bloch, for

his part, earned an undergraduate degree in “industrial design” in 1960, formerly designed “avant-

garde cardiovascular surgery facilities,” and eventually transitioned into automotive-safety

consulting. In their respective reports, they referred to the Optima’s purported defect as either

“EMI” (electromagnetic interference) or “crosstalk”—in essence, a theory that one set of wires in

the steering column can erroneously send signals to another, thus opening the throttle and causing

the car to accelerate. Specifically, they contended that a worn “clock spring”—a set of coiled

wires within the steering wheel—had erroneously routed an electrical signal “to the electronic

engine controller,” commanding the throttle to open (and thus the car to accelerate) without Parks’s

input.

Defendants’ response to these four experts was twofold. First, they moved to exclude

Bloch, Sero, Kress, and Loudon on the ground that their opinions were inadmissible under the

Daubert standard and to exclude Kress and Loudon on the additional ground that each was

irrelevant because neither could speak to specific defect. 509 U.S. 579 (1993). Second, defendants

moved for summary judgment. They reasoned that because plaintiffs had no admissible or relevant
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 46

experts capable of testifying about what precise malfunction had induced the acceleration,

defendants perforce had failed to create a genuine issue of material fact about it.

Naturally, in response to the motions to exclude, plaintiffs contended that all four of their

experts satisfied Daubert. But as to defendants’ relevance objections to Kress and Loudon,

plaintiffs argued that defendants’ criticisms of the pair’s relevance was based on a

misunderstanding of Kress and Loudon’s roles. Indeed, plaintiffs explicitly conceded that neither

Kress nor Loudon could testify to any specific defect. But they argued that this fact did not make

them irrelevant—they were simply appearing in supporting roles to bolster the star “defect”

witnesses, Bloch and Sero. In plaintiffs’ own words, “Dr. Kress does not purport to opine as to a

specific defect in the design of Kia’s electronic throttle control system or the cruise control

system. . . . Dr. Kress made it crystal clear that he was relying upon other experts to describe that

condition.” Opp’n to Mot. to Exclude Tyler Kress at 2, R. 320. Likewise, as to Loudon, plaintiffs

explained that “he was not tasked with replicating a defect, but to compare the characteristics of a

cruise control-induced acceleration in a 2008 Kia Optima with the characteristics of an accelerator

pedal-induced acceleration in a 2008 Kia Optima. . . . [H]is purpose was not to replicate a

malfunction. It is the role of Plaintiffs’ other experts Samuel Sero and Byron Bloch to describe

how the cruise control design can cause unintended throttle opening.” Opp’n to Mot. to Exclude

Steven Loudon at 3, R. 322. So, clear as day and in plaintiffs’ own words, neither Kress nor Loudon

could speak to a specific defect.

As for their response to Kia’s motion for summary judgment, plaintiffs put forth six

supposedly independent grounds as to why the case presented a triable issue.3 First, they argued

3
Each of the five “additional” grounds is entirely dependent on the admissibility of the first ground—that
“EMI” or “crosstalk” caused a cruise-control malfunction that caused the acceleration. As was made clear at oral
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 47

that a genuine issue existed concerning whether worn wires in the steering column could cause

“crosstalk.” Second, they argued that if “crosstalk” indeed had occurred, there existed an issue

about whether it could have commanded the cruise-control to accelerate the car without Parks’s

input. Third, they argued that the Optima’s failure to include a method of overriding a crosstalk-

induced cruise-control malfunction was a design flaw. Fourth, in a variation of the third point, they

argued that the Optima should have included a “brake override”—a mechanism that could have

overridden a faulty electrical signal and allowed Parks to properly employ the brake. Fifth, they

argued that the Optima’s mangled gas pedal was frozen in a neutral position, rather than a

depressed position, supporting an inference that Parks’s foot was off the gas at the moment of

impact. And sixth, they argued that Kia had neglected to conduct a “failure modes and effects

analysis” on the Optima when designing it, which could have uncovered its supposed vulnerability

to electronic malfunctions.

The district court adjudicated the motions to exclude and the motion for summary judgment

in a combined opinion. As to Bloch and Sero—the specific-defect experts—it reasoned that neither

came even close to satisfying the strictures of Daubert. Sero, for instance, has never published or

even submitted for publication a scientific paper on “crosstalk.” He used to run experiments on the

theory, but he gave up fifteen years ago when they failed to support it. Likewise, Bloch has never

tested the “crosstalk” theory, could not answer basic questions about how the Optima’s cruise-

argument, none of the five other theories about why Kia should have made provisions to mitigate a crosstalk-induced
cruise-control malfunction is sufficient to defeat summary judgment (or, for that matter, even relevant) unless it can
also be shown that, in the first place, there was a crosstalk-induced cruise-control malfunction to mitigate. See
Recording of Oral Arg., 4:37–5:06 (The Court: “Well if you needed an override or a throttle control, in addition to
needing that device to mitigate a sudden acceleration, wouldn’t there have had to have been an underlying defect to
create the acceleration in the first place, separate and apart from the issue of this electronic throttle control system?”
Counsel: “Yes.”).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 48

control system works, and could not even define the term “volt” when asked to do so in a

deposition. Moreover, the “crosstalk” theory has failed to gain general scientific acceptance.

Concluding that Bloch and Sero were practicing unsupported junk science, the district court

granted Kia’s motions to exclude them.

The district court turned next to the motions to exclude Kress and Loudon. It properly held

that both motions were moot. For even if the pair’s testimony were admissible, plaintiffs had

already explicitly conceded that neither Kress nor Loudon could testify to any specific defect

within the Optima that had supposedly produced the acceleration.4 Once again, Kress did “not

purport to opine as to a specific defect” and Loudon was “not tasked with replicating a defect” or

“malfunction”; it was “the role of Plaintiffs’ other experts Samuel Sero and Byron Bloch to

describe how the cruise control design can cause unintended throttle opening.” Opp’n to Mot. to

Exclude Tyler Kress at 2, R. 320; Opp’n to Mot. to Exclude Steven Loudon at 3, R. 322. Based

on those explicit representations by plaintiffs, as well as the district court’s conclusion that, in fact,

neither Kress nor Loudon could offer a theory of specific defect independent of Bloch and Sero,

the district court granted summary judgment to Kia. Bloch and Sero could not create a genuine

issue on defect because their testimony was inadmissible, and Kress and Loudon could not create

a genuine issue on defect because their testimony concededly did not go to defect. After the district

court entered judgment for Kia, this appeal followed.

4
As explained below, plaintiffs conceded that Kress and Loudon cannot speak to specific defect for an
obvious reason—that, as a matter of fact, neither Kress nor Loudon has any theory of specific defect. See infra 56–
57, 68 n.14, 70–72. And that, in turn, is why neither plaintiffs nor the other opinions can articulate a coherent theory
about how the Optima supposedly malfunctioned.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 49

II.

One might be forgiven for thinking that an issue in this appeal concerns whether the district

court erred in its exclusion of Bloch and Sero under Daubert. Bloch and Sero were, after all, the

only experts plaintiffs proffered who could speak to specific defect. Yet at the same time, we would

have reviewed their exclusion for an abuse of discretion—a “highly deferential” standard under

which we would have reversed only if the district court’s ruling were “manifestly erroneous.”

Johnson v. Manitowoc Boom Trucks, Inc., 484 F.3d 426, 430 (6th Cir. 2007); Gen. Elec. Co. v.

Joiner, 522 U.S. 136, 142 (1997). In tacit recognition of the fact that they never could have made

that showing, plaintiffs have not appealed the exclusions of Bloch and Sero.5 Instead, they claim

that the district court erred because it did not treat Loudon as a specific-defect expert—even though

plaintiffs explicitly told the district court they were not using Loudon to explain any specific defect

that supposedly caused the cruise-control to malfunction. See Appellant’s Br. at 12; Recording of

Oral Arg., 7:14–8:06. Some time after the notice of appeal was filed, apparently, Loudon

5
Counsel for the Hill plaintiffs opened the reply portion of his oral argument by claiming that Bloch and
Sero were “very qualified, longtime auto-safety experts.” Recording of Oral Arg., 30:23–30:34. But if plaintiffs really
believed that comment, they presumably would have appealed the district court’s decision to exclude Bloch and Sero’s
testimony. They pointedly did not, and yet counsel still thought it appropriate to refer to Bloch and Sero’s excluded
theories several times throughout his argument as legitimate bases to reverse the district court. Of course, the Hills’
counsel is in good company: The lead opinion too recites Bloch and Sero’s theories at length, despite their
unchallenged exclusion and thus irrelevance to this appeal. See, e.g., Lead Op. at 8–9 nn.6–7. Likewise, the lead
opinion claims that “a considerable school of thought maintains that unintended acceleration can occur without driver
error or pedal entrapment. Any number of malfunctions in a vehicle’s electronics could cause sudden and unintended
acceleration.” Id. at 4. The lead opinion fails to cite any record evidence for this statement, nor does it address the
record materials from NHTSA and the National Transportation Safety Board reflecting that the vast majority of
unintended-acceleration cases are actually because of driver error rather than electronic malfunctions. See NTSB
Report at 36–38, R. 143-4 (explaining that while unintended acceleration is rare, pedal misapplication is a “well-
documented” and common cause, particularly among elderly drivers).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 50

underwent this metamorphosis and, based upon his newfound knowledge of specific defect, say

plaintiffs, the district court must be reversed.

The other opinions’ indulgence of this about-face is inexplicable under the ordinary rules

of appellate procedure. Consider first the nature of the order that we are reviewing. Summary

judgment in the district court is governed by Rule 56 of the Federal Rules of Civil Procedure. Rule

56 provides that summary judgment is proper “if the movant shows that there is no genuine dispute

as to any material fact” and that “the movant is entitled to judgment as a matter of law.” Fed. R.

Civ. P. 56(a). In other words, “the moving party bears the initial burden of identifying those parts

of the record which demonstrate the absence of any genuine issue of material fact.” Johnson v.

Ford Motor Co., 13 F.4th 493, 502 (6th Cir. 2021) (quoting White v. Baxter Healthcare Corp., 533

F.3d 381, 389–90 (6th Cir. 2008)). But “[o]nce the moving party”—here, Kia—“has met its

burden, the burden shifts to the non-moving part[ies]”—here, plaintiffs—“to demonstrate ‘specific

facts showing that there is a genuine issue for trial[.]’” Id. (quoting Celotex Corp. v. Catrett, 477

U.S. 317, 324 (1986) (emphases added)). Further, those “designated portions of the record”

identified by the nonmovant “must be presented with enough specificity that the district court can

readily identify the facts upon which the nonmoving party relies.” Guarino v. Brookfield Twp.

Trustees, 980 F.2d 399, 405 (6th Cir. 1992) (citation omitted). So the question under Rule 56 is

not whether the record somewhere contains a triable issue, but whether the nonmovants—here,

plaintiffs—met their burden and demonstrated those issues with specific record citations in their

response to Kia’s motion for summary judgment.

Put differently, Rule 56 codifies a summary-judgment-specific party-presentation

principle. We evaluate the arguments the nonmovants actually made in their response to the motion

for summary judgment, asking whether those arguments suffice to create a triable issue. Johnson,
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 51

13 F.4th at 502; cf. United States v. Sineneng-Smith, 140 S. Ct. 1575, 1579 (2020) (“In our

adversarial system of adjudication, we follow the principle of party presentation. [I]n both civil

and criminal cases, in the first instance and on appeal, we rely on the parties to frame the issues

for decision and assign to courts the role of neutral arbiter of matters the parties present.” (cleaned

up)). If the arguments actually made create no triable issue, then summary judgment is proper and,

under Rule 56, our inquiry proceeds no further. See U.S. Structures, Inc. v. J.P. Structures, Inc.,

130 F.3d 1185, 1191 (6th Cir. 1997) (“It is well settled that the non-moving party must cite specific

portions of the record in opposition to a motion for summary judgment, and that the court is not

required to search the record for some piece of evidence which might stave off summary

judgment.”); see also Guarino, 980 F.2d at 410 (“Neither the trial court nor the appellate

court . . . will sua sponte comb the record from the partisan perspective of an advocate for the non-

moving party.”).

On appeal, we undertake precisely the same inquiry—asking whether the nonmovants

demonstrated a triable issue in their response to the motion for summary judgment while granting

the district court no deference to its conclusions on that question. See Franklin v. Kellogg Co., 619

F.3d 604, 610 (6th Cir. 2010) (explaining that we review summary-judgment orders de novo). And

our duty on appeal to consider those arguments actually made is only compounded by the

fundamental requirement of issue preservation: If litigants want us to consider certain arguments

on appeal, they must have asserted those same arguments below. See, e.g., Hayward v. Cleveland

Clinic Found., 759 F.3d 601, 614 (6th Cir. 2014) (“The Supreme Court has repeatedly held it will

not consider an argument that a petitioner failed to assert in the courts below.”); accord Kreipke v.

Wayne State Univ., 807 F.3d 768, 781 (6th Cir. 2015) (“As a general rule in this Circuit, arguments

raised for the first time on appeal are forfeited.”); Armstrong v. City of Melvindale, 432 F.3d 695,
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 52

700 (6th Cir. 2006) (“[T]he failure to present an issue to the district court forfeits the right to have

this argument addressed on appeal.”); Am. Copper & Brass, Inc. v. Lake City Indus. Prods., Inc.,

757 F.3d 540, 545 (6th Cir. 2014) (explaining that the failure to make an argument in response to

a dispositive motion forfeits the argument on appeal). In short, because we are a court of review

rather than first view, we inspect the arguments the parties actually made below—not a new-and-

improved case “fashioned after a district court’s unfavorable order.” Conlin v. Mortgage Elec.

Registration Sys., Inc., 714 F.3d 355, 360 n.5 (6th Cir. 2013) (quoting Hall v. Warden, Lebanon

Corr. Inst., 662 F.3d 745, 753 (6th Cir. 2011)).

Application of these basic principles would have rendered our decision a straightforward

affirmance. Once again, plaintiffs below not only failed to argue that Loudon was their specific-

defect expert, but they explicitly disclaimed reliance on Loudon to expound whatever specific

defect could have caused the acceleration. Their behavior thus ventured beyond mere forfeiture—

the passive failure to develop and preserve an argument—and into the realm of waiver—the

“intentional relinquishment” of the argument. United States v. Montgomery, 998 F.3d 693, 697

(6th Cir. 2021). Because waiver represents an argument’s affirmative repudiation, it is supposed

to “come[ ] with the strictest consequences”—that we do not review the waived claim “at all.” Id.

(citing United States v. Olano, 507 U.S. 725, 733 (1993)); accord United States v. Baskerville, 820

F. App’x 383, 385 (6th Cir. 2020) (noting that “[w]aived claims are unreviewable on appeal”);

Gentek Bldg. Prods, Inc. v. Sherwin-Williams Co., 491 F.3d 320, 327–28 (6th Cir. 2007).

Accordingly, we should have rejected out of hand plaintiffs’ attempted reinvention of Loudon as

a specific-defect expert. Instead, and without explanation, the other opinions rebuff these first

principles and entertain Loudon’s appellate renaissance as a defect expert.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 53

Considering such explicitly waived arguments distorts the integrity of the judicial process

itself. Below, both the district court and defendants shared the same fundamental impression, based

on plaintiffs’ own representations in their responses to the motions to exclude Kress and Loudon,

that neither witness could opine on specific defect.6 Those representations engendered a double

reliance: First, defendants relied on them in choosing how to frame their arguments and briefing

to the district court. And second, the district court itself relied on them in crafting its summary-

judgment order, which was predicated on plaintiffs’ explicit representation that neither Kress nor

Loudon was a specific-defect expert. Plaintiffs on appeal now take a directly contrary position:

that Loudon is a specific-defect expert and thus the district court erred in declining to evaluate the

admissibility of his testimony and in granting summary judgment. Obviously, however, the district

court never had the chance to evaluate that argument, since plaintiffs repudiated it below. The

other opinions’ indulgence of such flip-flopping gives parties the impression that they may freely

“chang[e] positions according to the exigencies of the moment.” New Hampshire v. Maine,

532 U.S. 742, 750 (2001). But precisely because it undercuts fair notice and encourages

gamesmanship, courts do not allow parties “absent any good explanation . . . to gain an advantage

by litigation on one theory, and then seek an inconsistent advantage by pursuing an incompatible

theory.” Id. at 749 (quoting 18 Charles A. Wright et al., Federal Practice and Procedure § 4477,

782 (1981)). Here, of course, plaintiffs have given no explanation at all for their about-face, save

6
The lead opinion comments that “[u]nder Defendants’ telling, [Kress and Loudon] had a narrow evidentiary
role at the trial court in which they only alluded to an unidentified malfunction in the Optima that might have caused
the unintended acceleration—but they left it to Sero and Bloch to identify the malfunction.” Lead Op. at 10 (emphasis
added). Apparently the lead opinion is under the impression that the view that Kress and Loudon could not opine on
specific defect was somehow an invention of defendants—an impression that betrays the lead opinion’s basic
misunderstanding of the record. Defendants (and the district court) thought Kress and Loudon could not opine on
specific defect because plaintiffs said Kress and Loudon could not opine on specific defect. All that defendants and
the district court did was repeat plaintiffs’ explicit descriptions of Kress and Loudon’s limited roles.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 54

for the fact that having advanced failed arguments below, they should now like to salvage their

claims with novel bases for reversal.

III.

I will now give special attention to the deficiencies of the lead opinion before turning to

the similar errors in the concurrence. However bad are the lead opinion’s misunderstandings of the

procedural standards governing this appeal, those misunderstandings are merely the gateway

through which it reaches its misapplication of Tennessee products-liability law. As everyone

agrees, this suit is governed by the Tennessee Products Liability Act of 1978, or “TPLA.” See

Tenn. Code. Ann. §§ 29-28-101–108. The TPLA establishes two relevant theories of recovery—

“defective condition” and “unreasonably dangerous condition”—both of which plaintiffs here

have invoked. See Smith v. Detroit Marine Eng. Corp., 712 S.W.2d 472, 474–75 (Tenn. Ct. App.

1985). I will briefly describe each theory before examining their misapplication by the lead

opinion. Under either theory, the burden of proof remains on the plaintiff throughout the

proceeding to illustrate a specific defect or dangerous condition. Fulton v. Pfizer Hosp. Prods.

Grp., 872 S.W.2d 908, 911 (Tenn. Ct. App. 1993). The mere occurrence of an accident is not proof

that a product is defective or unreasonably dangerous. Id. Rather, the plaintiff must trace the injury

to a specific defect or condition in the product. Benson, 868 S.W.2d at 636 (quoting Browder, 541

S.W.2d at 404); King, 37 S.W.3d at 435. To survive summary judgment under the defective-

condition theory, then, plaintiffs had to propound testimony illustrating what “specific defect”

caused the Optima to malfunction. See id.

Similarly, under the unreasonably-dangerous-condition theory, plaintiffs had to propound

testimony illustrating what condition the Optima possessed, that it was unreasonably dangerous

for normal use, and that the condition proximately caused plaintiffs’ injuries. See, e.g., Jackson v.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 55

Gen. Motors Corp., 60 S.W.3d 800, 805–06 (Tenn. Ct. App. 2001). Under this theory, the

dangerous condition technically need not be a defect—an imperfection—but may include design

features as well. See, e.g., Ray v. BIC Corp., 925 S.W.2d 527, 532 (Tenn. 1996) (explaining that a

manufacturer may still choose to market a dangerous product if it determines that the product’s

usefulness outweighs its risks). To show that a product is unreasonably dangerous, the TPLA

establishes two tests—a “consumer-expectation” test and a “prudent-manufacturer” test. Id. at 530.

The former asks whether a product performed in some way that an ordinary consumer would not

have expected, while the latter asks whether a prudent manufacturer with knowledge of the

condition still would have chosen to place the item in the stream of commerce. Id.

These distinctions matter little for purposes of the present appeal. Plaintiffs have

consistently asserted that some supposed design flaw (whatever it was) rendered the Optima both

defective and unreasonably dangerous. See Irion v. Sun Lighting, Inc., No. M2002-00766-COA-

R3-CV, 2004 WL 746823, *4 (Tenn. Ct. App. Apr. 7, 2004) (“The alternatives make little

difference in the case before us because [plaintiff] claims that a design defect was the defective

condition and also made the lamp unreasonably dangerous.”). So the relevant inquiry is simply

whether plaintiffs have articulated what specific, identified condition with the Optima caused the

unintended acceleration. See Benson, 868 S.W.2d at 636 (quoting Browder, 541 S.W.2d at 404);

see also King, 37 S.W.3d at 435.

The lead opinion apparently has no quarrel with these principles in the abstract. Indeed, it

explicitly recognizes that “[a] plaintiff must prove a specific defect under the TPLA, regardless of

the theory of liability.” Lead Op. at 15. The issue is not the lead opinion’s accurate recitation of

the standard, but its misapplication of that standard to the facts of this case. Shortly after

recognizing that plaintiffs “must prove a specific defect,” id. (emphasis added), the lead opinion
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 56

suggests reversing the district court on the ground that plaintiffs could establish “some

unspecified” defect in the Optima’s cruise control. Id. at 18 (emphasis added). Yet the lead

opinion’s reference to proving a specific defect with an unspecified defect is not a mere lapsus

linguae. Over and over, it suggests that we should reverse and remand because:

• “[A] jury could reasonably find that some defect in the cruise
control caused the acceleration of Parks’[s] Optima.”
• “A reasonable jury might well choose to believe that the
circumstantial evidence demonstrates that some defect in the
cruise control was more likely than not the cause of the
unintended acceleration.”
• The pedal having been in the idle position at impact raises
the inference that “some unspecified cruise control
malfunction” caused the crash.
• “[T]he pedal position and gap could support a finding that
something other than driver error caused the accident, thus
raising, circumstantially, the possibility of some unspecified
cruise control malfunction.”
• “[C]ircumstantial evidence” suggests “that an unspecified
cruise-control malfunction caused the crash.”
• “A jury could credit the various eyewitness testimony to
conclude that an unspecified malfunction was the source of
the sudden acceleration[.]”
• The district court erred in concluding that “Loudon’s cruise-
control signature testing failed as circumstantial evidence
that a non-specific cruise control malfunction caused the
crash.”
• “[A] jury could reasonably infer that something other than
prolonged pedal misapplication was the source of the
acceleration.”
• “[T]he very core of the case” concerns “whether some
unidentified defect in the cruise control caused the December
2015 accident.”
• Once again, that “circumstantial evidence” suggests “that an
unspecified malfunction caused the crash.”
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 57

Id. at 16–18, 29–32, 34 (emphases added). In repeatedly and inexplicably claiming that plaintiffs

may prove a specific defect with an unspecified, non-specific defect, the lead opinion shirks the

clear requirement under Tennessee law that a specific defect be shown.

Of course, the notion that plaintiffs may prove a specific defect by proving a non-specific

defect is a self-evident contradiction. But the problems with the lead opinion’s view that Loudon

suffices to defeat summary judgment extend even deeper. Set aside for a moment the fact that

plaintiffs below explicitly waived reliance on Loudon to illustrate a specific defect with the

Optima, and assume further that Loudon actually did proffer a theory about how a specific defect

with the Optima caused the acceleration. If Loudon in fact had offered such a theory below, why

wouldn’t the lead opinion simply repeat that theory, rather than taking the inexplicable step of

declaring that plaintiffs can prove a specific defect with a non-specific defect?

The answer is this: the lead opinion cannot articulate Loudon’s theory of specific defect

because Loudon has, in fact, never offered a theory about any specific mechanical or electronic

malfunction with the Optima. He has no theory for the lead opinion to recite.7 All he ever did was

attempt to lend credence to Bloch and Sero’s theory of specific defect—that “EMI” or “crosstalk”

from a worn clock spring erroneously routed electrical signals to the cruise control, causing the

throttle to open—by driving a properly functioning exemplar Optima around a closed track. So all

he has shown is that hitting ninety-two miles per hour by pressing the “accelerate” button produces

RPMs consistent with the figure at which Parks’s tachometer was frozen. He has never

explained—nor even attempted to explain—at an electronic or mechanical level how the

7
And, of course, the fact that plaintiffs articulated no admissible theory of specific defect means that their
failure-to-warn claim must fail as well. Plaintiffs necessarily had to first put forth evidence that the Optima was, in
fact, defective, before they could claim that Kia breached a duty to warn about the putative defect. See Flax v.
DaimlerChrysler Corp., 272 S.W.3d 521, 542 (Tenn. 2008).
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 58

“accelerate” button could have operated without Parks’s input. See, e.g., Loudon Dep. at 153:21–

156:2, 27:4-7, R. 311-2 (explaining that he neither performed nor even attempted to perform

testing about whether or how the cruise-control system could fail). In other words, plaintiffs’

remark below that Loudon was not being used to establish a specific defect was not only an explicit

waiver of his use for that purpose, but a simple restatement of reality: that Loudon has no theory

of specific defect.8 Indeed, the lead opinion’s repeated recognition that all his testimony can

achieve is an inference of some “unspecified,” “non-specific” defect constitutes the lead opinion’s

implicit concession that plaintiffs cannot satisfy the specific-defect standard.

As with its revelation that one can prove a specific defect with proof of a non-specific

defect, the lead opinion’s discussion of res ipsa loquitur similarly misunderstands the procedural

and substantive standards that govern this case. In their appellate brief, plaintiffs for the first time

invoked res ipsa loquitur—a doctrine “that permits, but does not compel, a jury to infer negligence

from the circumstances of an injury,” Seavers v. Methodist Med. Ctr. of Oak Ridge, 9 S.W.3d 86,

91 (Tenn. 1999)—in their attempt to patch the holes created by the exclusion of Bloch and Sero.

8
Tellingly, counsel for the Hill plaintiffs could not articulate at oral argument what, specifically, Loudon
claims malfunctioned with Parks’s Optima. Indeed, he reconfirmed that Loudon was enlisted not to replicate
malfunctions, but to corroborate Bloch and Sero’s EMI theory. The full exchange was as follows:
The Court: Didn’t you state in your opposition to the motion to exclude that Loudon, quote, “was
not tasked with replicating a defect. His purpose was not to replicate a malfunction.”?
Counsel: That’s right.
The Court: So how can you now be arguing before us that that was his task given that you
represented to the district court that it was not his task?
Counsel: So, to clarify Judge Bush, and this is an excellent question, we are not asking Mr. Loudon,
Mr. Loudon as part of his—he didn’t replicate a defect. He didn’t inject fault signals, faulty signals,
or try to put additional electronic magnetic interference to basically scramble the wires. He didn’t
do that. What he did do was take a 2008 Kia Optima exemplar, and he took it on a closed track, and
he looked at the different forms of acceleration through the gas pedal, which is what the defendants
are saying, that it’s a basically a pedal-to-the-metal, wide-open throttle condition, and he found that
the RPMs shot way too high and that it accelerated too fast compared to the facts in our case.
Recording of Oral Arg., 8:07–9:03.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 59

The lead opinion remarks that the “[p]laintiffs failed to raise res ipsa loquitur in their response to

the motion for summary judgment, so they cannot raise it for the first time on appeal, and this

argument is not preserved for appellate review.” Lead Op. at 33 n.33. On this point the lead opinion

is right. As explained above, the failure to include arguments in a response to a dispositive motion

forfeits those arguments on appeal. See supra 50–54. The problem once again is not the lead

opinion’s recitation of legal principles in the abstract, but its misapplication of those principles to

this case.

Consistent enforcement of the lead opinion’s own forfeiture rule—for instance, to

plaintiffs’ late-blooming arguments about Loudon—would make this case an easy affirmance.

Plaintiffs never argued below in their response to Kia’s motion for summary judgment that Loudon

was a specific-defect expert or that he could independently articulate a theory of specific defect

that did not hinge on the excluded testimony of Bloch and Sero. Again, in their responses to the

motions to exclude, plaintiffs were contemporaneously arguing the very opposite—that Loudon’s

task was not to “replicate a defect” or “malfunction” and that “[i]t [wa]s the role of Plaintiffs’ other

experts Samuel Sero and Byron Bloch to describe how the cruise control design can cause

unintended throttle opening.” Opp’n to Mot. to Exclude Steven Loudon at 3, R. 322. The lead

opinion never explains why it selectively applies the forfeiture rules to res ipsa loquitur but not to

Loudon’s appellate reinvention as a defect expert.

Ironically, though, the lead opinion’s claim that it is treating res ipsa loquitur as forfeited

is not even true. Just a few pages before its invocation of forfeiture, the lead opinion described the

following argument as one of its suggested bases for reversal of the district court:

Fourth and finally, Plaintiffs contend that reasonable minds could conclude, based
on witness and eyewitness testimony, that a sudden acceleration event would not
have happened in the absence of a defect. This res-ipsa-loquitur-like theory is
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 60

similar to the Browder-Greco discussion analyzed previously. Some courts call this
the “malfunction doctrine” or “malfunction theory.” The malfunction doctrine
allows plaintiffs in strict liability cases to infer defectiveness from the negation of
other causes. See, e.g., Balducci v. Hyundai Motor Am., Inc., 406 F. App’x 517,
518 (2d Cir. 2011).

Lead Op. at 18–19. So, much like the notion that one can prove a specific defect with proof of a

“non-specific” defect, the lead opinion says that res ipsa loquitur is forfeited yet suggests applying

a concededly “res-ipsa-loquitur-like theory” to reverse the district court. Id. The lead opinion

never explains how a theory of liability that was forfeited below can serve as a basis for reversal

simply because it is repackaged with a different label.

Aside from being another self-evident contradiction, however, the lead opinion’s suggested

resurrection of res ipsa loquitur under the banner of “malfunction theory” engenders two further

problems.

First, much as plaintiffs never argued res ipsa loquitur in their response to Kia’s motion

for summary judgment, plaintiffs there also never even mentioned “malfunction doctrine” or

“malfunction theory,” much less claimed that “malfunction doctrine” or “malfunction theory”

would permit them to “infer defectiveness from the negation of other causes.” Lead Op. at 18–19.

Under the lead opinion’s own standard for forfeiture, then, these theories are themselves “not

preserved for appellate review.” Id. at 33 n.33.

Second and more importantly, the lead opinion’s invocation of “malfunction doctrine”

represents yet another misapplication of Tennessee law, and thus yet another violation of the Erie

doctrine. The lead opinion apparently draws “malfunction doctrine” from the Second Circuit’s

Balducci decision, which itself was applying the law of Connecticut, not the law of Tennessee. See

Balducci, 406 F. App’x at 518. Unlike Connecticut, Tennessee does not permit products-liability

plaintiffs “to infer defectiveness from the negation of other causes.” But see Lead Op. at 18–19.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 61

To the contrary, it keeps the burden on the plaintiff to affirmatively show what specific defect

caused a malfunction and then to trace the plaintiff’s injury back to that specific defect. See Fulton,

872 S.W.2d at 911 (“The burden is upon the plaintiff to show that there is something wrong with

the product.” (cleaned up)); see also Benson, 868 S.W.2d at 636 (quoting Browder, 541 S.W.2d at

404); King, 37 S.W.3d at 435; Fulton, 872 S.W.2d at 911 (explaining that an injury by itself “is

not proof of a defect and thereby raises no presumption of defectiveness”).9 For that very reason,

Tennessee courts have repeatedly held that res ipsa loquitur cannot overcome a plaintiff’s failure

to illustrate what specific defect caused her injury. Fulton, 872 S.W.2d at 912 (“Under Tennessee

law the doctrine of res ipsa loquitur is not a substitute for proof of defect.”) (citing Browder, 541

S.W.2d at 404). Here, of course, plaintiffs possess no viable theory of specific defect, as the lead

opinion’s constant references to a “non-specific” and “unspecified” defect plainly illustrate. And,

unsurprisingly, a doctrine from Connecticut cannot remedy a fatal omission under the laws of

Tennessee.

So what does the lead opinion have to say in response to all that? As for its importation of

Connecticut law into a Tennessee diversity dispute, nothing. As for its baffling invention of the

“non-specific defect” standard, its brief attempt at self-defense—apparently deemed unworthy of

inclusion in the lead opinion’s main text—hides away in footnote 13. There, seeking to accuse this

opinion of having “mischaracterize[d]” the lead opinion, the lead opinion, ironically,

9
The lead opinion’s invocation of Connecticut’s “malfunction doctrine” is particularly misplaced because
malfunction doctrine permits an even more radical inference than classical res ipsa loquitur, which itself cannot
supplant a plaintiff’s duty under the TPLA to prove a specific defect. Under classical res ipsa loquitur, after plaintiffs
had already identified a defect and shown causation, they could raise an inference that the manufacturer had been
negligent in permitting the defect. See Fallon v. Matworks, 918 A.2d 1067, 1075–76 (Conn. Super. Ct. 2007).
Malfunction theory, by contrast, allows plaintiffs to raise an inference that a product is defective simply because a
malfunction occurred during normal use. Id. Such a presumption of defectiveness would directly contradict Tennessee
courts’ longstanding interpretation of the TPLA, which requires that plaintiffs affirmatively show what specific defect
caused the malfunction.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 62

mischaracterizes this opinion. Lead Op. at 18 n.13. It claims that this opinion wrongly assumes

that a TPLA plaintiff may only prove specific defect with direct evidence, thus discounting the

“unremarkable and irrefutable” fact that a TPLA plaintiff may also prove specific defect with

circumstantial evidence. Id. Yet what is truly irrefutable (and also remarkable) is the irrelevance

of this argument. No one disputes that a TPLA plaintiff may prove a specific defect with

circumstantial evidence. The foundational problem with the lead opinion, once again, is that it

permits recovery even if all the plaintiff has proven—whether by evidence circumstantial or

direct—is merely some “non-specific,” “unspecified” defect. Therein lies the true violation of the

Erie doctrine, given that no matter whether the plaintiff employs direct or circumstantial evidence,

Tennessee law requires that she prove the specific defect having proximately caused her injuries.

That the lead opinion expends its micro-rebuttal swinging at shadows, though, is at least consistent

with its general dearth of illumination.

IV.

Perhaps because of its erroneous treatment of the TPLA, the lead opinion’s reasoning on

“non-specific defect” has failed to garner a second vote. Indeed, because the concurrence is only

in the judgment, the lead opinion’s discussion of the TPLA does not constitute a holding of this

Court, and it binds no one—not even the district court on remand. The concurrence instead

apparently agrees with me that proof of a specific defect requires proof of, indeed, a specific defect.

And, it turns out, the lead opinion and concurrence disagree on more than just that. The

lead opinion suggests “reversal” of the district court’s grant of summary judgment on the grounds

that (1) a specific defect can be proven with proof of a non-specific defect and (2) the plaintiffs

have established a “non-specific defect” with Loudon, at least to the extent his testimony is

admissible under Daubert. Lead Op. at 30. The concurrence, by contrast, does not say that
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 63

summary judgment was affirmatively wrong on the merits, so it recommends that we “vacate” the

decision below rather than “reverse” it.10 Concurrence at 38. And, having recognized that the

TPLA requires proof of a specific defect for liability to arise, the concurrence claims that plaintiffs

did establish a specific defect with Loudon’s report that he prepared for the litigation. Id. at 39.

Last, the concurrence apparently agrees with the lead opinion that the district court’s first task on

remand should be to evaluate Loudon’s admissibility under Daubert.11 Id.

Having attempted to discern the effect of today’s seriatim opinions, I turn now to the flaws

in the concurrence. Aside from its proper refusal to endorse the lead opinion’s view that one can

prove a specific defect with a non-specific defect, the concurrence simply replicates in miniature

the foundational errors of the lead opinion. Start with the fact that the concurrence disregards the

plaintiffs’ explicit waiver below of the use of Loudon to illustrate specific defect. Id. at 38–39. If

the concurrence were following basic principles of issue preservation, it would “have rejected out

of hand plaintiffs’ attempted reinvention of Loudon as a specific-defect expert.” Supra at 52.

Instead, citing no caselaw, engaging with none of my relevant arguments, and providing no

10
But perhaps both the other opinions actually agree on this point. Despite its facial suggestion that we should
“REVERSE the district court’s order granting summary judgment to Defendants,” Lead Op. at 1, the lead opinion
never suggests that we must, in turn, award summary judgment to plaintiffs, as would technically be entailed by a
reversal. Compare Jon O. Newman, Decretal Language: Last Words of an Appellate Opinion, 70 Brooklyn L. Rev.
727, 728 (2005) (“If the appellate ruling orders the complete opposite of what the district court has ruled, e.g., the
district court has entered judgment for the plaintiff (for example, on a motion for summary judgment) and the court
of appeals orders entry of judgment for the defendant . . . the decretal language should include the word ‘reversed.’”),
with E-Pass Tech., Inc. v. 3Com Corp., 473 F.3d 1213, 1218 (Fed. Cir. 2007) (“By vacating we signaled that, although
the district court’s prior decision rested upon erroneous grounds, a proper claim construction might support a judgment
(summary or otherwise) in favor of either party, depending on the evidence and argument submitted to the district
court on remand and considered by the district court in the first instance.”). Pedantry aside, whether we employ the
term “reverse” or the term “vacate,” the real point is that at least two (and seemingly all) opinions today agree that
summary judgment at least could be proper for Kia on remand.
11
The concurrence also claims that a remand is necessary to evaluate Kress’s admissibility under Daubert,
Concurrence at 39, but it never suggests how Kress offers any theory of specific defect. So it implicitly recognizes
that Kress’s admissibility is irrelevant to whether summary judgment was proper. Still, the district court should make
Daubert rulings on remand as to both Kress and Loudon, lest it exclude solely Loudon and we get a second appeal
about how, all along, it was Kress who was the real specific-defect expert.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 64

meaningful explanation, it entertains plaintiffs’ clearly unpreserved and explicitly repudiated claim

that Loudon goes to specific defect.

True, unlike the lead opinion, the concurrence at least acknowledges that plaintiffs

conceded below that Loudon “was not tasked with replicating a defect.” Concurrence at 38–39.

Yet note how the concurrence does not finish the rest of that quotation: “[H]is purpose was not to

replicate a malfunction. It is the role of Plaintiffs’ other experts Samuel Sero and Byron Bloch to

describe how the cruise control design can cause unintended throttle opening.” Opp’n to Mot. to

Exclude Steven Loudon at 5, R. 322 (emphasis added). Once again, defendants and the district

court thought that Loudon could not describe how the cruise-control design could cause unintended

throttle opening because plaintiffs told defendants and the district court that it was Bloch and

Sero’s role—not Loudon’s—to describe how the cruise-control design could supposedly cause

unintended throttle opening. See id.

The concurrence also tries to enlist portions of plaintiffs’ response in opposition to

defendants’ motion to exclude Loudon to claim that, all along, plaintiffs “used Loudon’s testimony

to establish a specific defect—the use of a single wire in the cruise control system.” Concurrence

at 39. Of course, plaintiffs’ response in opposition to defendants’ motion to exclude Loudon is the

same document in which plaintiffs stated that Loudon’s “purpose was not to replicate a

malfunction. It is the role of Plaintiffs’ other experts Samuel Sero and Byron Bloch to describe

how the cruise control design can cause unintended throttle opening.” Opp’n to Mot. to Exclude

Steven Loudon at 5, R. 322. So the concurrence’s theory, apparently, is that in the same document,

plaintiffs were secretly making and preserving an argument that they were simultaneously and

explicitly repudiating. The concurrence never explains how the district court was supposed to

decode this secret message from plaintiffs’ briefing.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 65

Try as it might to get around plaintiffs’ waiver, the concurrence itself gives up the ghost in

its final footnote. There, it remarks, “[i]t is correct to say that plaintiffs’ original plan was to rely

on the testimony of Bloch and Sero to establish defect.” Concurrence at 39 n.2. In other words,

plaintiffs’ “original plan”—what they actually argued below—was that Bloch and Sero, not

Loudon, went to specific defect. Id. The concurrence thus concedes that plaintiffs on appeal are

not arguing their “original” theory from the district court, but a new theory “fashioned after [the]

district court’s unfavorable order.” Conlin, 714 F.3d at 360 n.5 (quoting Hall, 662 F.3d at 753).

Yet that is precisely what a litigant cannot do under the law of our circuit and the requirement of

issue preservation. See, e.g., id.; see also Ghandi v. Police Dep’t of City of Detroit, 747 F.3d 338,

343 (6th Cir. 1984) (“Having presented their claims in the district court under one theory, plaintiffs

cannot save their claim against [defendants] by proceeding under a new theory on appeal.”); Mich.

Bell Tel. Co. v. Strand, 305 F.3d 580, 590 (6th Cir. 2002) (“Propounding new arguments on appeal

in attempting to prompt us to reverse the trial court—arguments never considered by the trial

court—is not only somewhat devious, it undermines important judicial values[.]”). Much like the

lead opinion, then, the concurrence also disregards well-established procedural rules governing

this appeal.

Having concluded that plaintiffs may now advance what concededly was not their

“original” theory in the district court, the concurrence next claims that a remand is necessary so

that the district court can evaluate the substance of Loudon’s supposed contributions. Concurrence

at 38 n.1. The concurrence claims that this is necessary because of its erroneous assertion that the

district court refused to “independently” review Loudon’s materials in light of plaintiffs’ waiver

of his use to show specific defect. Id. It quotes appellee’s brief that, “[t]aking Plaintiffs at their

word, and after independently reviewing the testimony of Kress and Loudon, the district court
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 66

concluded that neither expert provided evidence of a specific defect in the Optima.” Id. (quoting

Appellee’s Br. at 22). Yet, says the concurrence, “[f]rom its order, we cannot discern whether or

not the district court ‘independently’ reviewed the testimony.” Id. In other words, supposedly, the

district court’s order granting summary judgment is unclear as to whether the district court formed

its conclusions about Loudon based on how plaintiffs said he would be used (which the

concurrence apparently considers insufficient) or whether the district court proceeded to

“independently” evaluate whether Loudon could create a genuine issue of material fact on specific

defect.

To begin with, the fact that the concurrence considers it necessary for a district court to

“independently” comb the record in search of a genuine issue of material fact further exposes the

extent to which it, like the lead opinion, has cast off the ordinary rules of summary judgment, party

presentation, and issue preservation. The district court was perfectly entitled to rely on plaintiffs’

description of the purpose for which they were using Loudon. It had no obligation to scour the

record to “independently” formulate different arguments on behalf of plaintiffs. See, e.g., Guarino,

980 F.2d at 410 (“Neither the trial court nor the appellate court . . . will sua sponte comb the record

from the partisan perspective of an advocate for the non-moving party.”).

In any event, though, the concurrence’s claim that the district court failed to review

Loudon’s materials as part of its summary-judgment order is wrong. The district court’s order cites

both Loudon’s deposition testimony and his reports at multiple points as part of its conclusion that

Loudon offered no theory of specific defect that was not itself contingent on Bloch and Sero. See,

e.g., Order at 22 n.18, R. 341 (“Loudon’s report provides technical information regarding P0564

DTCs recorded in the 2008 Kia Optima, and explains that P0564 is specific to the cruise control
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 67

system. [Doc. 311-1, at 14].12 As is discussed in Plaintiffs’ opposition to summary judgment, no

DTC related to the cruise control or otherwise was recorded at the time of the Parks accident.”);

id. at 33 (“Loudon’s rebuttal report responds to the report of Defense expert James Walker, Jr.,

whose testimony was not subject to a motion to exclude.”); id. at 33–34 (“Loudon’s report does

not appear to discuss the crumpling or bracket dislocation—instead focusing on the pedal position

itself, which is not in dispute—and explained at his deposition that he ‘was not asked to evaluate’

the question of ‘how and to what extent’ the assembly ‘was out of position post-crash’ and that

‘that’s not what I was . . . looking at in my rebuttal report.’ [Doc. 317-13, at 64].13 In other words,

Loudon concedes that he does not know or opine as to how the pedal could have been trapped in

an idle position[.]”); id. at 34 (“Loudon’s failure to dispute Walker’s modeling suggests that the

post-crash pedal position actually cuts against plaintiffs.”); id. (“For the [RPM] signature to be

consistent with a cruise-control-induced acceleration is to presuppose, however, that an unintended

cruise-control-induced acceleration is possible. As discussed above, Loudon does not purport to

conclude that the presupposition is possible, and he instead defers to the excluded testimony of

Bloch and Sero on that point.”). So, as should be apparent, Loudon’s materials suffuse the district

court’s order. Frankly, it is difficult to understand the concurrence’s claim that, from the “order,

we cannot discern whether or not the district court ‘independently’ reviewed the testimony.”

Concurrence at 38 n.1. A simple review of that order reveals that the district court engaged in the

very analysis the concurrence now says it omitted.

12
Document 311-1 is, indeed, one of Loudon’s reports. See Loudon Report, R. 311-1.
13
Document 317-13 is Loudon’s May 2019 deposition. See Loudon Dep., R. 317-13.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 68

Last, much like the lead opinion leveraged a score of procedural errors to reach its

substantive error that Loudon speaks to a “non-specific” defect, the concurrence leverages its own

erroneous understanding of issue preservation to reach its own substantive error that Loudon

speaks to a specific defect. The concurrence cites a line from plaintiffs’ response in opposition to

defendants’ motion to exclude Loudon—again, the same document in which plaintiffs said Bloch

and Sero rather than Loudon went to specific defect—to assert that “plaintiffs used Loudon’s

testimony to establish a specific defect.” Id. at 39. In particular, says the concurrence, the brief

“recited [Loudon’s] conclusion that ‘[t]he design of the cruise control system is defective in that

it uses a single wire,” itself vulnerable to “a single point failure,” “through an unreliable

clockspring harness.” Id. And so the district court abused its discretion, supposedly, “[b]ecause

plaintiffs did argue that Loudon’s testimony established a specific defect (the use of a single wire).”

Id.

First, note where the concurrence discovered this argument: from plaintiffs’ response in

opposition to defendants’ motion to exclude Loudon—not from plaintiffs’ response in opposition

to defendants’ motion for summary judgment. Even if we assumed that Loudon could speak to

specific defect (as explained below, he can’t), the concurrence apparently concedes that plaintiffs

did not make this argument where it counted: in the relevant brief opposing summary judgment.14

14
Indeed, plaintiffs’ response in opposition to defendants’ motion for summary judgment never claimed that
the “single wire” theory was a standalone theory of specific defect and causation, much less one that Loudon could
propound independently of Bloch or Sero. According to Sero—not Loudon—a single wire was problematic because
it could somehow allow an “on” signal to “bypass” an “off” signal sent to the cruise control, thus causing an
inadvertent throttle opening. Response in Opp’n to Mot. for Summ. J. at 14, R. 325 (citing Sero for the “bypass”
theory). On Sero’s assumption that such a “bypass” was possible, Loudon diagnosed a single wire as the “lack of a
failsafe” and a design flaw. Id. So not only was the “single wire” theory contingent on Sero, but it was merely pitched
as the “lack of a failsafe” rather than the underlying defect that would have been required to generate an erroneous
“on” signal sent across the wire. Id. As counsel for the Hill plaintiffs conceded at oral argument, the lack of a failsafe
would only matter if it could also be shown that there was an underlying defect that caused an errant signal in the first
place for the failsafe to have mitigated. See Recording of Oral Arg., 4:37–5:06 (The Court: “Well if you needed an
override or a throttle control, in addition to needing that device to mitigate a sudden acceleration, wouldn’t there have
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 69

Once again, under basic summary-judgment principles, the district court had no obligation to look

beyond that brief to other parts of the record in a self-directed attempt to drum up a genuine issue

of material fact. See, e.g., Skotak v. Tenneco Resins, Inc., 953 F.2d 909, 915 (5th Cir. 1992)

(“Although these articles are part of the total record, [plaintiffs] never referred to them in their

response in district court to [defendant’s] motion for summary judgment, nor did they argue in

district court in their summary judgment response brief that such evidence created a genuine issue

of fact[.] . . . Because [plaintiffs] failed to refer to these articles in district court in their summary

judgment response, the articles were not properly before that court in deciding whether to grant

the motion; therefore, they will not be considered here.”); see also Wimbush v. Wyeth-Ayerst Lab’y

Co., 619 F.3d 632, 638 n.4 (6th Cir. 2010) (“[Plaintiff] suggests that the fact the evidence was in

the record somewhere is sufficient to create a question of fact and survive summary judgment.

This is simply incorrect. Even if the evidence to which [plaintiff] now refers was sufficient to rebut

[defendant’s] evidence of adequate warning, it was [plaintiff’s] job to point to the evidence with

specificity and particularity in the relevant brief rather than just dropping a pile of paper on the

had to have been an underlying defect to create the acceleration in the first place, separate and apart from the issue of
this electronic throttle control system?” Counsel: “Yes.”). The reason for that concession is obvious: the “lack of a
failsafe” could not have been a cause-in-fact of the collision unless there was an underlying defect producing an errant
signal that could have been arrested by the inclusion of the failsafe. See Tatham v. Bridgestone Am. Holding, Inc., 473
S.W.3d 734, 751 (Tenn. 2015) (explaining that the TPLA requires cause-in-fact and proximate causation). In the same
vein, the response in opposition to summary judgment frankly acknowledged that the “single wire” theory could not
by itself have caused an acceleration. Response in Opp’n to Mot. for Summ. J. at 15, R. 325. Just after setting out the
“single wire” theory, it queried, “But the question remains: What caused the Parks UA [unintended acceleration]
incident? Given the nature of electronic malfunctions, it would be impossible to examine the vehicle and pinpoint the
precise manner in which the unintended throttle opening occurred.” Id. It then suggested, as does today’s lead opinion,
that “circumstantial evidence” showed that “[s]omething” was wrong with the cruise control. Id. So not only did the
response concede that plaintiffs have no idea about whether the “single wire” theory played a causal role in the
collision, but it also conceded that it is not even possible to meet plaintiffs’ burden under the TPLA—to “pinpoint”
what specific defect caused the supposed unintended throttle opening. Benson, 868 S.W.2d at 636 (quoting Browder,
541 S.W.2d at 404); King, 37 S.W.3d at 435. Perhaps because of this slew of problems with how plaintiffs actually
presented the “single wire” theory in their response in opposition to the motion for summary judgment, the
concurrence instead relies on the response in opposition to the motion to exclude Loudon to claim that the district
court somehow erred by failing to credit the “single wire” theory.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 70

district judge’s desk and expecting him to sort it out.” (emphases added)). So, once again, the

concurrence imposes a spurious obligation on the district court to have divined plaintiffs’

subtextual, “real” arguments against summary judgment—the arguments plaintiffs actually made

in their “relevant brief” notwithstanding. See Wimbush, 619 F.3d at 638 n.4.

Yet even if we had to consider this plainly unpreserved argument, the notion that Loudon

can testify to specific defect and causation because of a wire in an “unreliable clockspring” is false.

Loudon himself in his sworn deposition on May 3, 2019, explained that he had done nothing to

evaluate the possibility of “a worn clock spring as a possible scenario to explain Ms. Parks’[s]

crash” because that was the role of Bloch and Sero. Loudon Dep. at 153:21–154:6, R. 311-2.

Indeed, he was asked point-blank:

Q: Have you done anything to evaluate the possibility of a worn clock spring as a
possible scenario to explain Ms. Parks’[s] crash?
A: There were other experts15 that worked on that aspect of the—
Q: So the answer is no?
A. I was not asked to do that and it really fell outside my—my purview.

Id. (emphasis added). Indeed, not only had he failed to investigate an “unreliable clockspring” as

a cause of the supposed malfunction, he had never even seen the Parks vehicle’s clockspring. See

id. at 177:3–6 (“Q: Has anybody on the plaintiffs’ side examined the clock spring in Ms. Parks’[s]

vehicle? A: I—I don’t know. I was not asked to do that specifically.”). Nor had he done anything

to measure faulty “voltages in the wires going from the cruise control resume accel switch.” Id. at

104:7–11. Nor was he testing how the cruise-control system could fail. See id. at 27:4–7 (“Q. [N]ot

15
That is, Bloch and Sero. Note that I have omitted plaintiffs’ counsel’s various objections from the quoted
portions of the deposition transcript, as they are not relevant to the present inquiry.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 71

only didn’t you induce a failure, you didn’t try to induce a failure; correct? A. That’s correct.”);

id. at 109:5–6 (“I wasn’t testing the failure of those systems in this [experiment].”).

Indeed, because it was Bloch and Sero’s role, not Loudon’s, to provide a theory of how a

supposed defect could have caused an unintended acceleration, Loudon had investigated none of

the plaintiffs’ theories about how the cruise-control had malfunctioned:

Q: Have you tried to evaluate a possible issue with the brake lamp stop switch as a
possible scenario to explain Ms. Parks’[s] crash?
A: Again, I was not asked to look at that specifically—
Q: Same question—
A: —in my work.
Q: Same question with regard to a possible voltage drop as a possible scenario to
explain Ms. Parks’[s] crash. Have you evaluated that?
A: I was not asked to, no.
Q: Have you evaluated possible issues with the smooth operation of the cruise
control as a possible scenario to explain Ms. Parks’[s] crash?
A: Again, I was not asked to evaluate that specifically.
Q: Have you evaluated possible brake issues as a possible scenario that could—that
could cause Ms. Parks’[s] crash?
A: I didn’t look specifically at—at that, no.
Q: Have you evaluated possible unforeseen circumstances with the ECU as a
possible scenario that could explain Ms. Parks’[s] crash?
A: I did not specifically look at that.
Q: Did you look at possible EMF cross-talk as a possible scenario that could have
caused Ms. Parks’[s] crash?
A: I think there were other experts16 looking at EMF and EFI and I was not asked
to look at that specifically.

Id. at 154:7–155:22 (emphasis added). Thus, as his deposition makes plain, Loudon himself

exposes the fallacy of the concurrence’s attempt to recast him on appeal as a specific-defect expert.

16
Again, the “other experts” were Bloch and Sero.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 72

Loudon conceded that he had performed no investigation into the clockspring, id. at 177:3–6, no

investigation into whether it was vulnerable to faulty voltage signals, id. at 104:7–11, and was

relying on “other experts” to describe the specific electronic malfunction that supposedly could

have caused the acceleration. Id. at 153:21–154:6; 154:7–155:22. So while the concurrence at least

attempts to shoehorn Loudon into the correct legal standard—specific defect—it too cannot escape

the reality that Loudon “never offered” a standalone theory of defect that was not itself contingent

on the excluded opinions of Bloch and Sero. Supra at 57. The district court recognized precisely

this point in its summary-judgment order. Order at 34, R. 341 (“For the [RPM] signature to be

consistent with a cruise-control-induced acceleration is to presuppose, however, that an unintended

cruise-control-induced acceleration is possible. As discussed above, Loudon does not purport to

conclude that the presupposition is possible, and he instead defers to the excluded testimony of

Bloch and Sero on that point.”). Yet rather than affirm the district court’s thoughtful opinion, the

concurrence casts the deciding vote to authorize a pointless remand so the district court can assess

the admissibility of testimony the declarant himself has already admitted does not opine on a

specific electronic malfunction.

V.

The facts of this case are tragic, and plaintiffs are no doubt sympathetic. Yet even in the

face of such circumstances, it remains “the duty of the judiciary calmly to poise the scales of

justice.” United States v. Bollman, 24 F. Cas. 1189, 1192 (C.C.D.C. 1807) (Cranch, J.). That

exercise would produce a straightforward affirmance here. Under the relevant rules of issue

preservation, we should hold plaintiffs’ novel arguments waived or forfeited. And under Tennessee

law, plaintiffs have failed to illustrate how any specific defect caused the collision.
 Nos. 20-5690/5693 Hill, et al. v. Kia Motors Am., Inc., et al. Page 73

I will close with this thought. At several points the lead opinion speaks as if it were

remanding this case for a trial on the merits. It is not. What it actually remands for is the district

court’s determination of whether Kress and Loudon can satisfy the Daubert standard. Just as it did

with Bloch and Sero, the district court remains perfectly free on remand to exclude Kress and

Loudon and thus to reimpose summary judgment.

I respectfully dissent.