OPINION

CORNELIUS, Chief Justice.

The plaintiffs Susan Renae Miles, individually and as next friend of minors Willie Searcy and Jermaine Searcy, and Kenneth Miles sued Ford Motor Company and Douglas Stanley, Jr., doing business as Doug Stanley Ford, to recover damages for personal injuries suffered by Willie Searcy as a result of a two-car collision that occurred in Dallas County. The plaintiffs asserted liability on the basis of defective product, negligence, and breach of warranties, alleging that a faulty seat belt caused Willie Searcy's injuries. Before trial, the court granted Ford and Stanley's summary judgment motion on the loss of consortium claims of Kenneth Miles, Willie's stepfather, and Willie's brother Jermaine. The jurors found for the plaintiffs on all claims against Ford but not against Stanley, and awarded $30 million compensatory damages. The jurors also found that Ford had acted with gross negligence and malice, and awarded exemplary damages of $10 million.

Both sides appeal. Ford and Stanley contend that the trial court erred in refusing to transfer venue to Dallas County from Rusk County; in allowing the plaintiffs to present certain expert testimony and refusing to allow Ford to present contradicting expert testimony; in admitting the plaintiffs' test evidence and excluding the defense's test; in admitting voluminous exhibits en masse without individually considering authenticity, hearsay, and relevance, and without balancing their probative value against possible prejudice and juror confusion; by instructing the jurors that Ford had a duty to "prevent all unnecessary and avoidable injuries"; in rendering judgment and in failing to grant Ford's motion for new trial because the jurors' liability findings were fatally conflicting; and in denying Ford's motion for new trial because the evidence was factually insufficient to support the jury's liability and punitive damage findings.

The plaintiffs contend that they should have a new trial on their claims against Stanley because the jury's failure to find liability against Stanley was against the great weight and preponderance of the evidence and was manifestly unjust, and that the trial court erred in granting Ford and Stanley's motion for partial summary judgment on plaintiffs' loss of consortium claims.

Because we find factually insufficient evidence to support the findings of gross negligence and malice, we will reverse the judgment and remand the cause for a new trial on the issues of gross negligence and malice. In all other respects, we will affirm the judgment.

Susan Miles is Willie and Jermaine Searcy's mother, and her husband Kenneth Miles is their stepfather. Willie was fourteen and Jermaine was twelve in April of 1993, the time of the collision. The collision occurred when the 1988 Ford Ranger pickup Kenneth Miles was driving struck a Mercury Cougar that had crossed the median into the truck's path. Kenneth Miles and Willie were wearing seat belts and shoulder harnesses, and Jermaine was wearing a lap belt only.

Kenneth Miles and Jermaine were injured, but generally recovered. Willie suffered a spinal cord injury rendering him quadriplegic. He needs round-the-clock care and can breathe only with the help of a ventilator. Willie's injuries form the basis of the suit.

The plaintiffs' theory was that, some time before the collision, Willie, who was riding on the right passenger side, leaned forward to pick up trash on the floorboard, which caused the webbing on his shoulder harness to spool out, creating six to eight inches of slack. A device on the truck called a "tension eliminator" prevented the webbing from rewinding onto the spool, much as the pull-down device on a window shade can be set to prevent the shade from rewinding. Plaintiffs say that when Willie sat up straight, the webbing folded behind his back, creating the appearance of a snug shoulder harness when the harness really was quite loose. When the collision occurred, Willie was thrown forward and his torso moved forward until it hit the webbing when the slack ended. The belt then rode up under Willie's neck. This force caused the ligaments attaching Willie's skull to his spine to tear apart. One expert also theorized that the collision caused Willie's body to "submarine," or slide under his lap belt. When his shoulder harness caught his head, the force separated his skull from his spine. The plaintiffs argue that the tension eliminator created the slack in the shoulder harness and thus caused the injury.

The defense's theory was that Willie was properly wearing his shoulder harness and lap belt, with no undue slack, and that the force of the crash simply was too great for Willie's neck muscles and ligaments, which were immature for his age and size. The injury occurred simply by his head and neck snapping forward, they argue.

I. VENUE

Ford and Stanley, in their position as appellants, contend that the trial court should have transferred the suit from Rusk County to Dallas County because the plaintiffs are Dallas County residents, Doug Stanley Ford is a Dallas County business, and the accident occurred in Dallas County.

We must uphold the trial court's venue determination if, after viewing the entire record, including the trial record, we find any probative evidence that venue was proper in Rusk County. TEX.CIV.PRAC. REM.CODE ANN. § 15.064(b) (Vernon 1986); Ruiz v. Conoco, Inc., 868 S.W.2d 752, 758 (Tex. 1993). At the time this suit was filed, a plaintiff could sue an out-of-state corporation doing business in this state in any county where the company had an agency or representative. TEX.CIV.PRAC. REM.CODE ANN. § 15.037, Act of June 3, 1987, 70th Leg., 1st C.S., ch. 4, § 1, 1987 Tex.Gen.Laws 52-53. When there were two or more defendants in the same action and venue was proper in the county of suit as to any one defendant, venue could be maintained there against the other defendants. TEX.CIV.PRAC. REM.CODE ANN. § 15.061. For the purposes of the venue statute, an agency means a more-or-less permanent business operation, and a representative means someone who has broad powers to act for the corporation. Ruiz v. Conoco, Inc., 868 S.W.2d at 759. An ordinary employee does not qualify as an agency or a representative because an ordinary employee does not have the broad power and discretion to act for the corporation. Id. To come within the term agency or representative, there must be some discretionary power conferred on the entity. Brazos River Transmission Elec. Coop. v. Vilbig, 244 S.W.2d 266, 268 (Tex.Civ.App. — Dallas 1951, no writ). The agency or representative must be an entity authorized to bring about a business relationship between the principal and a third party, an authority not characteristic of a mere servant or employee. Allis-Chalmers Mfg. Co. v. Coplin, 445 S.W.2d 627, 628 (Tex.Civ.App. — Texarkana 1969, no writ).

The plaintiffs based their claim of venue in Rusk County on the fact that Premier Ford Mercury, Inc., a Ford dealership, is located in Rusk County. Ford and Stanley argue that Premier is not an agency or representative of Ford Motor Company as those terms were used in the venue statute. The Sales and Service Agreement detailing the relationship between Ford and Premier prohibits the dealership from "act[ing] or attempt[ing] to act, or represent himself, directly or by implication, as agent of the company or in any manner assum[ing] or creat[ing] any obligation" and disclaims any general liability by Ford for dealership operations. Thus, Ford says, the dealership has practically no discretion to act for it. It argues that the dealer's obligations, such as promotion and advertising, are obligations to do things for the dealership itself, not for Ford. Other Premier duties, Ford contends, such as handing out warranty booklets and owners' manuals, are not discretionary but ministerial. Premier cannot alter a Ford warranty and it has no discretion in deciding whether to deliver a warranty. Ford also points out that Premier's predelivery inspections, as well as all recall and correction work, are nondiscretionary and are done only under Ford's express direction.

The plaintiffs, on the other hand, contend that the dealership is authorized and is responsible to Ford for several discretionary duties specified by the Sales and Service Agreement, including (1) promoting the retail sale of Ford cars and trucks; (2) advertising Ford vehicles; (3) incorporating Ford's warranty into sales documents for Ford vehicles; (4) performing predelivery inspection, conditioning, and repair of Ford vehicles; (5) performing all warranty service on Ford products; and (6) performing recall and correction work on Ford products. The Premier Ford manager has final approval of all warranty work done at Premier Ford and final approval on all vehicles received new from Ford. There was also evidence that Ford has told the dealership it represents Ford in dealing with the public. The dealership agreement also requires Premier to "develop, maintain and direct a trained, quality" vehicle sales organization and service organization.

Premier was incorporated by Ford under a dealer development program designed to increase minority ownership of dealerships. Premier's original directors were Ford employees, and Ford handled the legal paperwork for establishing Premier.

The decisions dealing with agency and representation in this arena are not uniform. See, e.g., Colorado Interstate Gas Co. v. MAPCO, Inc., 570 S.W.2d 164 (Tex.Civ.App. — Amarillo 1978, no writ), where the court held that a pipeline company with a $31 million facility, 140 nonsupervisory employees and fourteen supervisors in the county of suit was not an agent or representative for venue purposes because no employee had the power to bring about business relationship between the company and a third party; but see General Motors Corp. v. Ramsey, 633 S.W.2d 646, 648 (Tex.App. — Waco 1982, writ dism'd), where the court held that a Chevrolet dealership was a General Motors agent because the dealership's contract required it to make predelivery inspections and adjustments of each new vehicle before delivery.

Ford argues that Ruiz v. Conoco, Inc., supra, requires that to be an agent or representative, one must have broad discretionary powers, and Premier's powers are so circumscribed by Ford that the relationship does not meet the Ruiz standard. The evidence here, however, is subject to different interpretations. There is a large amount of evidence on this issue, and we believe that at least some of that evidence can reasonably be construed to show that Premier had broad discretion to act in some respects for Ford. When some evidence supports the trial court's determination of venue, we are not at liberty to overturn that determination. Ruiz v. Conoco, Inc., supra at 758.

II. JURY FINDINGS

All parties complain about the jury findings on liability. Plaintiffs say that because Doug Stanley Ford made no change whatsoever in the pickup's seat belt mechanism, and the jury found that the mechanism was defective when it left Ford's possession, the jury's failure to find that it was defective when it left Stanley's possession is against the great weight of the evidence. Ford says that, since no change was made in the seat belt mechanism after it left Ford's possession, the jury findings that it was defective in Ford's possession but not defective in Stanley's possession are in total conflict. We disagree with both of these contentions.

The court submitted the following questions on liability:

QUESTION NO. 1.

Was there a design defect in the occupant restraint system of the truck at the time it left the possession of the persons listed below that was a producing cause of the injuries to Willie Searcy?

A "design defect" is a condition of the product that renders it unreasonably dangerous as designed, taking into consideration the utility of the product and the risk involved in its use.

Answer "Yes or No" for each of the following:

Answers:

A. Ford Motor Company: Yes

B. Doug Stanley Ford: No

QUESTION NO. 2.

Was there a manufacturing defect in the occupant restraint system of the truck at the time it left the possession of the persons listed below that was a producing cause of the injuries to Willie Searcy?

A "manufacturing defect" means a condition of the product that renders it unreasonably dangerous as manufactured. An "unreasonably dangerous" product is one that is dangerous to the extent beyond that which would be contemplated by the ordinary user of the product, with the ordinary knowledge common to the community as to the product's characteristics.

Answer "Yes" or "No" for each of the following:

Answers:

A. Ford Motor Company: Yes

B. Doug Stanley Ford: No

QUESTION NO. 3.

Was there a marketing defect in the occupant restraint system of the truck at the time it left the possession of the persons listed below that was a producing cause of the injuries to Willie Searcy?

A "marketing defect" with respect to the product means the failure to give adequate warnings of the product's dangers that were known, or by the application of reasonably developed human skill and foresight, should have been known, or the failure to give adequate instructions to avoid such dangers, which failure rendered the product unreasonably dangerous as marketed.

"Adequate" warnings and instructions means warnings and instructions given in a form that could reasonably be expected to catch the attention of a reasonably prudent person in the circumstances of the product's use; and the content of the warnings and instructions must be comprehensible to the average user and must convey a fair indication of the nature and extent of the danger and how to avoid it to the mind of a reasonably prudent person.

In connection with this question, you are further instructed that when a manufacturer fails to give adequate warnings or instructions, the law presumes that the user would have read and heeded adequate warnings or instructions. This presumption may be overcome if other evidence in the case shows that adequate warnings or instructions would not have been heeded or followed if given.

Answer "Yes" or "No" for each of the following companies:

Answers:

A. Ford Motor Company: Yes

B. Doug Stanley Ford: No

QUESTION NO. 4.

Did the negligence, if any, of the persons named below proximately cause the injuries to Willie Searcy?

Answer "Yes" or "No" for each of the following:

A. Ford Motor Company: Yes

B. Doug Stanley Ford: No

QUESTION NO. 5.

Was the occupant restraint system of the truck provided by the persons named below unfit for the ordinary purposes for

which occupant restraint systems are used because of a defect, and, if so, was such unfit condition a proximate cause of the injuries to Willie Searcy?

As used in this question, the word "defect" means a condition of the occupant restraint system that renders it unfit for the ordinary purposes for which occupant restraint systems are used because of the lack of something necessary for adequacy.

Answer "Yes" or "No" for each of the following:

Answers:

A. Ford Motor Company: Yes

B. Doug Stanley Ford: No

QUESTION NO. 6.

Was there a breach of an implied warranty of fitness for a particular purpose by the persons named below that was a proximate cause of the injuries to Willie Searcy?

There was a breach of implied warranty of fitness for a particular purpose if, at the time of contracting:

1. The person named below had reason to know the particular purpose for which its product was required;

2. The person named below had reason to know that the buyer was relying on the skill and judgment of that person to select or furnish a suitable product; and

3. The occupant restraint system of the truck supplied by the person named below was unfit for the particular purpose for which it was purchased.

Answer "Yes" or "No" for each of the following:

Answers:

A. Ford Motor Company: Yes

B. Doug Stanley Ford: No

The plaintiffs argue that the jurors' failure to find liability against Stanley in connection with design defects and marketing defects was against the great weight and preponderance of the evidence. It is undisputed that Doug Stanley Ford did not alter the restraint system after it left Ford Motor Company. The dealership had the opportunity to deactivate the tension eliminator by cutting the button that activated the device when the door was shut, but apparently did not. The plaintiffs reason that, because the jurors found the restraint system was defective when it left Ford's possession and because Doug Stanley did nothing to the system, the restraint system must have been defective when it left the dealership's possession and so the jury's finding was against the great weight and preponderance of the evidence.

When reviewing factual sufficiency challenges when the complaining party had the burden of proof and the jury failed to find a fact, see M.J. Sheridan Son v. Seminole Pipeline, 731 S.W.2d 620, 623 (Tex.App. — Houston [1st Dist.] 1987, no writ), the court of appeals must first examine all of the evidence, Lofton v. Texas Brine Corp., 720 S.W.2d 804, 805 (Tex. 1986), and having considered and weighed all of the evidence, should set aside the verdict only if the failure to find is so against the great weight and preponderance of the evidence that it is clearly wrong and unjust. Cain v. Bain, 709 S.W.2d 175, 176 (Tex. 1986). When considering great-weight points complaining of a jury's failure to find a fact, a reviewing court should remember that a jury was not convinced by a preponderance of the evidence. Herbert v. Herbert, 754 S.W.2d 141, 144 (Tex. 1988). Courts of appeals may not reverse merely because they conclude that the evidence preponderates toward an affirmative answer. They should reverse only after detailing the evidence indicating that the great weight of the evidence supports an affirmative answer. Id.

The jury's answer was not a finding, but a failure to find. The jurors' answer did not amount to a finding that the restraint system was not defective when it left Stanley Ford. It amounted only to a finding that plaintiffs failed to prove their case on that issue by a preponderance of the evidence. The plaintiffs had the burden of proving the dealership refrained from altering the belts. Although the fact that no change was made is not disputed here, we cannot say that the great weight of the evidence at trial required the jury to make an affirmative finding of liability against Stanley.

Ford argues that the jury's findings are in fatal conflict because the jury failed to find that the restraint system had a defect when it left Stanley's possession, while at the same time it found that the restraint system left Ford's possession with a defect.

Ford also argues that, because Ford and Stanley supplied essentially the same restraint system, the jurors' failure to find a breach of warranty as to the dealership fatally conflicts with its breach of warranty finding against Ford. Moreover, Ford contends that the jurors' failure to find a breach of warranty and product defect against the dealership proves, in essence, that Ford provided a defect-free truck and so it cannot be found negligent; that is, Ford was not negligent because it did not breach a warranty or supply a defective product.

Again, however, the jurors' answers were not affirmative findings but failures to find. The jury's failure to find that the plaintiffs proved a defect in the restraint system when it left the dealership does not mean the product was not defective. Similarly, the jurors' failure to find a breach of warranty against the dealership does not conflict with their findings against Ford, and their failure to find negligence against the dealership does not preclude their finding of negligence against Ford. The jurors' findings could reflect the general lack of direct evidence against the dealership and the plaintiffs' apparent failure to meet their burden as to the dealership.

III. SUFFICIENCY OF THE EVIDENCE

Ford challenges the factual sufficiency of the evidence to support the jury's findings of liability against it. When reviewing factual sufficiency challenges when the complaining party's opponent had the burden of proof and the jury answered affirmatively, M.J. Sheridan Son v. Seminole Pipeline, supra, a reviewing court considers and weighs all of the evidence, Lofton v. Texas Brine Corp., supra, and should set aside the verdict only if the evidence is so weak that the finding is clearly wrong and unjust. Cain v. Bain, supra.

In essence, Ford argues that because tension eliminators made seat belts more comfortable to wear, more people were willing to use seat belts and thus more lives were saved and injuries avoided. The plaintiffs argue that Ford ignored evidence suggesting that tension eliminators were hazardous and ignored safer alternatives.

1. Design Defect

The jury's determination of whether the restraint system was unreasonably dangerous as designed required the jurors to balance the product's utility against the risk of its use. If the risk outweighs the benefits, the product can be considered to be defectively designed. Ford relied on documents from the National Highway Traffic Safety Administration showing that adding tension eliminators to restraint systems increased seat belt use and thus saved lives overall. Dr. Rose M. Ray, a statistician, testified about a General Motors study finding that people used restraint systems more after the tension eliminators were introduced. She cited a study that found seat belt use in Chrysler vehicles rose from 8.3% to 19.3% after the company introduced tension eliminators.

The risk of tension eliminators, Ford argues, is small or nonexistent. The risk, Ford says, depends on the convergence of several events, viz: (1) an occupant of a 1988 Ford Ranger will inadvertently introduce slack into the shoulder harness, (2) the slack will be excessive enough to significantly increase the risk of injury in an accident, (3) the accident will occur before the occupant corrects the slack problem, and (4) the excessive slack will cause additional injury over that which would have occurred anyway. The studies Ford relied on concluded that the number or severity of injuries did not increase with the introduction of tension eliminators. NHTSA documents also found no problems associated with tension eliminator use.

The plaintiffs produced expert opinions that tension eliminators are unreasonably dangerous because of the possibility that an occupant could introduce slack into the shoulder strap.

Plaintiffs' expert evidence suggested that Ford had reasonable alternatives to tension eliminators, such as:

— A tension reducer, which eliminates some of the pressure on the occupant but does not allow the occupant to introduce slack into the harness.

— An electric break switch to electrically deactivate the tension eliminator.

— An adjustable D-ring and anchor combination. The D-ring is the loop attaching the shoulder harness to the car pillar immediately to the right and behind the outboard front seat passenger. An expert testified that Ford studies showed that such an arrangement could accommodate about 85% of the people who had problems with shoulder harness neck chafing.

— The same type belts Ford used in Europe and Australia, which do not use tension eliminators.

— A system similar to that installed in some Ford Probes, some of which Ford and Mazda jointly produced. Mazda refused to install tension eliminators in its version.

— A pretensioner in conjunction with a tension eliminator. A pretensioner reels in any slack in an emergency and "loads" the belt so that it will work more quickly and reduce the occupant's movement.

The plaintiffs also rely on the testimony of Roger E. Maugh, Ford's Automotive Safety Office director, who testified that Ford knew excessive slack was dangerous. He also testified that if the federal government required Ford to conduct crash tests with the one to one and a half inches of slack recommended in its owner manual, Ford would have difficulty certifying the tests because it could not predict how test dummies would move around in the cabs in the crashes. When asked if Ford tested its system with slack in the torso belt, he replied, without saying that Ford conducted no tests, that engineers "can judge pretty much" what is going to happen in such a situation. He also said that, while Ford did not test tension eliminators on the public, the company had to get its decision to use tension eliminators "confirmed."

The plaintiffs introduced a letter from Ford, Plaintiffs' Exhibit 103, that suggested the company found no correlation between seat belt comfort and seat belt use, although in the same letter Ford told the NHTSA that it feared outlawing tension-relieving devices would discourage seat belt use.

There was other evidence suggesting that tension eliminators created risks.

— A Ford document showed that the tension eliminator's primary disadvantage was that it reset "rather easily from occupant movement, such as reaching for the radio or glove compartment, and must be manually cancelled and reset to avoid excessive and inadvertent belt slack." The document also said, "Excessive slack, after reaching for something, is the most frequent complaint among drivers with [tension eliminators]. Between thirty and fifty percent of the drivers expressed this concern. When excessive slack does occur, most drivers remove the slack but describe this need to be bothersome." The document further said, "Inadvertent and excess slack are evidence in current designs because ease of setting the [tension eliminator] was the primary design consideration. The system sets so easily that the occupant frequently does so without being aware."

— Another Ford document shows that in one belt-use survey, Ford observed that 33% of three-point belt system users wore their shoulder belt loosely, with more than three inches of slack. The same document showed that a "decision tree" chart, marked "draft," suggesting that for 1990 passenger cars Ford not use an active belt with a tension eliminator on the passenger side, similar to the system in Miles' truck, for safety reasons.

— A Ford interoffice memo said that one survey showed that 55% of drivers observed in the survey wore their shoulder belts too loosely, with three or more inches of slack.

The plaintiffs also showed that, while the NHTSA says in some documents that tension eliminators contribute to occupant comfort and thus lead to greater usage, another NHTSA document mentioned problems with belt misuse and problems with belt systems that allow users to inadvertently introduce slack.

The plaintiffs also attacked the credibility of Ford's expert, Dr. Rose Ray, because she works for Failure Analysis Associates, which they describe as "industry-dominated," and her findings have been criticized. Her suggestion that tension eliminators are safe because no study shows they are dangerous has been criticized by the NHTSA, plaintiffs say. This reflects merely a lack of data, rather than a positive finding of safety, they contend.

The jury is the exclusive judge of the facts proved, the credibility of witnesses, and the weight to be given their testimony, and a reviewing court may not substitute its findings and conclusions for those of the jury. Benoit v. Wilson, 150 Tex. 273, 239 S.W.2d 792, 796-97 (1951). A reviewing court should not set aside a jury verdict merely because the jurors could have drawn different inferences or conclusions from the evidence. Id.

Rational jurors could have determined that the risk of Ford's using tension eliminators outweighed the benefits associated with increased seat belt use, considering that Ford may have had alternatives. Viewing the evidence as a whole, a rational fact finder taking into consideration the restraint system's utility and the risk involved in its use could have determined that the restraint system had a design defect when it left Ford's possession that was the producing cause of Willie Searcy's injury.

2. Manufacturing Defect

Ford contends that the jury's finding that the restraint system had a manufacturing defect is supported by insufficient evidence.

Roger Maugh testified that a tension eliminator, when correctly made and functioning, will not allow the creation of inadvertent slack as the result of routine movement. Witnesses at the crash scene testified that Willie Searcy was slumped forward in his seat with the belt spooled out. This constitutes some evidence that the restraint system had a manufacturing defect that caused the belt to spool out. Thus, the court correctly denied Ford's motion for an instructed verdict on this issue. Douglas v. Neill, 545 S.W.2d 903, 905 (Tex.Civ.App. — Texarkana 1977, writ ref'd n.r.e.); 4 MCDONALD TEXAS CIVIL PRACTICE § 21:53 (rev. 1992).

A manufacturing defect exists when a product does not conform to the manufacturer's design standards and blueprints and the flaw makes the product more dangerous and therefore unfit for its intended or reasonably foreseeable uses. Ford Motor Co. v. Pool, 688 S.W.2d 879, 881 (Tex.App. — Texarkana 1985), aff'd in part, rev'd in part, 715 S.W.2d 629 (Tex. 1986). The court gave the jury a definition of manufacturing defect, i.e., "a condition of the product that renders it unreasonably dangerous as manufactured," that did not clearly distinguish a manufacturing defect from a design defect. Under the definition, the jurors could have believed that, even though there was no direct evidence of a flaw in the system's manufacture, the restraint system, because of a defective design, was defective as manufactured. Ford does not complain of the definition. The Texas Pattern Jury Charges does not define manufacturing defect.

Considering Maugh's testimony, the eyewitnesses' testimony, and the court's definition, we find sufficient evidence to support the jury's finding.

3. Marketing Defect

Ford urges that the plaintiffs failed to present sufficient evidence of a marketing defect because they conceded that a written warning about excessive slack would have been ineffective, and that the only effective warning would have been a warning buzzer. Ford argues that the lack of a warning buzzer would be a design defect rather than a marketing defect.

The court defined "adequate warning" as "warnings and instructions given in a form that could reasonably be expected to catch the attention of a reasonably prudent person in the circumstances of the product's use." This included, but was not limited to, written warnings, and the jurors could have taken it to include warning buzzers or lights. Ford does not complain of the definition.

A warning about excessive slack was included in the truck's owner's manual. Kenneth and Susan Miles testified they did not read that portion of the manual. Witnesses also testified that neither Ford nor the dealership placed warnings on the seat belt itself.

We find sufficient evidence to justify the jury in finding that Ford failed to give adequate warnings of known product dangers that made the restraint system unreasonably dangerous as marketed.

4. Breach of Warranty of Merchantability and Negligence

Ford also challenges the jury's findings of breach of warranty and negligence. It contends there is insufficient evidence to support the findings.

Some evidence suggested that seat belt users could inadvertently introduce slack into torso belts with tension eliminators by routine movements such as leaning forward to adjust the radio or to open the glove box. The jurors could have found that this was a defect that made the restraint system unfit for the ordinary purpose for which it was used, that is, to restrain passengers in sudden stops, and that the unfit condition was the proximate cause of Willie Searcy's injuries. TEX.BUS. COM.CODE ANN. § 2.314 (Vernon 1994). Rational jurors also could have found that Ford's decision to use tension eliminators was negligent in light of Ford's knowledge and the alternatives available to it.

5. Breach of Warranty of Fitness

The jury found that Ford breached its implied warranty of fitness for a particular purpose. Ford contends there is no evidence to sustain the finding because plaintiffs failed to prove that the truck was sold to Kenneth Miles for a particular purpose apart from its ordinary use. Where the seller has reason to know any particular purpose for which the buyer requires the goods and that the buyer is relying on the seller's skill and judgment to select or furnish such goods, there is an implied warranty that the goods are fit for such purpose. TEX.BUS. COM.CODE ANN. § 2.315 (Vernon 1994).

The plaintiffs argue that Kenneth Miles bought the truck for a particular purpose — that of carrying passengers — and that he intended for the restraint system to provide safety for the driver and passengers in the event of a collision and believed that the truck did so.

A particular purpose envisages a specific use by the buyer that is peculiar to the nature of his business. The ordinary purposes for which goods are used are those included in the concept of merchantability and go to uses that are customarily made of the goods. Texas law is unclear on the question whether the particular purpose contemplated by Section 2.315 can be a specific ordinary purpose or whether the particular purpose must be a nonordinary purpose. In some jurisdictions a specific ordinary purpose will not invoke the warranty. See, e.g., Ford Motor Co. v. Fairley, 398 So.2d 216 (Miss. 1981). In other jurisdictions the warranty can protect a buyer whose particular purpose is the ordinary purpose. See, e.g., Tennessee Carolina Transp., Inc. v. Strick Corp., 283 N.C. 423, 196 S.E.2d 711 (1973).

Most of the Texas cases we have found on this issue have approved recovery for breach of warranty of fitness for a particular purpose. In all of those cases, however, there was evidence that the seller was told about or knew about peculiar or extraordinary uses to which the buyer intended to put the product that were at least somewhat out of the ordinary. See, e.g., Innovative Office Systems v. Johnson, 906 S.W.2d 940 (Tex.App. — Tyler), rev'd and remanded for entry of settlement agr., 911 S.W.2d 387 (Tex. 1995); Lester v. Logan, 893 S.W.2d 570 (Tex.App. — Corpus Christi 1994), writ denied per curiam, 907 S.W.2d 452 (Tex. 1995); Jeep Eagle Sales v. Mack Massey Motors, 814 S.W.2d 167 (Tex.App. — El Paso 1991, writ denied); Lanphier Constr. Co. v. Fowco Constr. Co., 523 S.W.2d 29 (Tex.Civ.App. — Corpus Christi 1975, writ ref'd n.r.e.); Parks v. Glidden Co., 433 S.W.2d 445 (Tex.Civ.App. — Texarkana 1968, writ ref'd n.r.e.). In Crosbyton Seed Co. v. Mechura Farms, 875 S.W.2d 353 (Tex.App. — Corpus Christi 1994, no writ), however, the court held that even though a seed merchant had extensive information about the buyer's situation, in terms of yield and purpose, the buyer's purpose was to produce a commercial cash crop, which was the ordinary purpose of the seed.

We believe the better rule is that a particular purpose must be a particular nonordinary purpose. Even if Kenneth Miles had expressly told either Ford or the dealership that he intended to carry passengers in his truck, that is the ordinary purpose of a passenger truck, rather than a particular purpose. Because there was no warranty for a particular purpose, the trial court should have granted Ford's motion for directed verdict on this question. But because the trial court could have properly based its award of compensatory damages on the plaintiffs' other causes of action, this error is harmless. TEX.R.APP.P. 81(b)(1); McAllen State Bank v. Linbeck Constr. Corp., 695 S.W.2d 10, 21 (Tex.App. — Corpus Christi 1985, writ ref'd n.r.e.).

6. Punitive Damages

Ford challenges the award of punitive damages, arguing that there is factually insufficient evidence to support the jury's finding that Ford was grossly negligent and acted with malice.

The trial court correctly instructed the jury on gross negligence that:

"Gross negligence" means more than momentary thoughtlessness, inadvertence, or error of judgment. It means such an entire want of care as to establish that the act or omission in question was the result of actual conscious indifference to the rights, welfare, or safety of the persons affected by it.

To find Ford Motor Company's negligence to be gross negligence, you must find (1) that, viewed objectively from the standpoint of Ford Motor Company, the act or omission of Ford Motor Company involved an extreme degree of risk, considering the probability and magnitude of potential harm to others, and (2) that Ford Motor Company had actual awareness of the risk involved, but nevertheless proceeded in conscious indifference to the rights, safety or welfare of others.

and on malice that:

"Malice" means an act that is committed by the defendant with a flagrant disregard for the rights of others and with actual awareness on the part of the defendant that the act will, in reasonable probability, result in human death, great bodily harm, or property damage.

The plaintiffs contend that Ford has not preserved error as to the jury's malice finding but only as to the gross negligence finding, and therefore the punitive damage award may be upheld solely on the basis of the malice finding. We disagree with this narrow view of Ford's appeal. Ford complains in one of its points of error that the evidence is factually insufficient to support the jury's punitive damage finding. The punitive damage finding is based on findings of gross negligence and malice on Ford's part. Under its point of error, Ford argues specifically that there is insufficient evidence of gross negligence. Although Ford does not specifically mention malice, it argues there is insufficient evidence of the elements of malice, in that it argues that it relied on national safety standards and studies showing that its restraint system was not unreasonably dangerous, and therefore it could not have acted in flagrant disregard of the rights and safety of its customers and in spite of actual awareness of probable injury. This point of error and the argument under it is sufficient to direct this court's attention to the complaint as to malice as well as gross negligence. TEX.R.APP.P. 74. Ford in its motion for new trial specifically contended that the gross negligence and malice findings were supported by factually insufficient evidence and were against the great weight and preponderance of the evidence. Transportation Ins. Co. v. Moriel, 879 S.W.2d 10 (Tex. 1994).

It should be noted at this point that Ford does not attack the punitive damage award as being excessive; nor does it contend that no evidence supports the underlying gross negligence and malice findings. It only contends that there is factually insufficient evidence to support the findings authorizing punitive damages, viz: the gross negligence and malice findings. That being the case, we are required by Pool v. Ford Motor Co., 715 S.W.2d 629 (Tex. 1986), when we reverse a jury finding on insufficiency grounds, to detail the evidence in our opinion and explain why the jury's finding is factually insufficient or so against the great weight of the evidence as to be manifestly unjust. As to the factual sufficiency to support a punitive damage award, we must conduct such an analysis and give such an explanation when either reversing or affirming such an award. Transportation Ins. Co. v. Moriel, supra.

In detailing the evidence supporting the findings of gross negligence and malice, we will use the very words of the plaintiffs as stated in their briefs: Steven Syson, plaintiffs' accident reconstruction expert, and Dr. Robert Curitz, their human factors expert, testified that, in their opinions, Ford was grossly negligent and acted with malice in designing and using the tension eliminator device. Ford designed and manufactured the tension eliminator for a number of reasons, one of which was a desire to achieve a perceived competitive advantage over other domestic manufacturers. Prior to, during, and following the development of the tension eliminator, Ford had detailed information that the tension eliminator would allow inadvertent slack into the occupant restraint system, seriously degrading the effectiveness of seat belt protection in the event of a serious collision. Having been repeatedly warned about this danger and with knowledge that its customers and dealers did not know how the tension eliminator worked, how dangerous it was, or how to properly use it, Ford failed to provide meaningful warnings or cautionary instructions. In the meantime, Ford refused to admit that it had conducted tests of such a dangerous product. Instead, Ford waited to have its "eyeball judgments" confirmed by the public's experience with the product.

Ford knew there was no reliable national clearinghouse for reporting of fatal or serious automobile collisions, and Ford also knew that most investigating officers would not know that a person had been killed or seriously injured by such a highly technical component of the occupant restraint system. Former Ford employee Maugh estimated that in a "worst case scenario" the tension eliminator would increase fatalities by 10% and serious injuries by 20%.

Syson testified that he was familiar, during the relevant time period, with the corporate policies of Ford Motor Company as they related to potentially defective products. Syson testified that when Ford identified what it believed was a defective product it would first run a "cost benefit" analysis to see what the cost would be to fix or repair the defect. Next, Ford would assign arbitrary values to each death or serious injury and would predict the number of occurrences which would involve either death or serious injury. Finally, Ford would determine the cost to litigate such deaths and injuries. Syson testified that if the cost to repair the defect exceeded the other costs, Ford would not correct the defect.

Ford's evidence against gross negligence and malice includes the following: Much of the plaintiffs' "evidence" that the tension eliminator was dangerous was in the form of some 7,000 pages of documents that were admitted in evidence not for the proof of the matter stated, but only to show that Ford had possession of them. Many of these documents were completely irrelevant to the tension eliminator system. The tension eliminator system used by Ford is not unreasonably dangerous because, in order to constitute a risk of injury, four separate and unlikely events must coincide: (1) the user will inadvertently reset the tension eliminator to introduce excessive slack, (2) the slack will be so excessive that it causes a risk of injury; (3) an accident will occur before the user has an opportunity to see, feel, and correct the excessive slack, and (4) the collision will be so severe that it will cause injury above that which would occur even without slack.

On the risk-utility balance, Ford produced two studies showing that tension eliminators did increase seat belt usage but did not result in an increase in the frequency or severity of injuries to persons wearing seat belts. Ford relied on National Highway Traffic Safety Administration studies and reports that repeatedly concluded that tension eliminators increase seat belt usage, but do not create an increased safety risk. The NHTSA actually concluded that "[T]he cases as a whole do not demonstrate that occupants of window shade equipped cars are injured more often or more seriously than occupants on non-window shade equipped cars." The alleged profit motive of Ford in its risk-benefit policy is no evidence of gross negligence or malice here, because Ford proved that introducing the tension eliminator in its Ranger trucks actually cost it more money than if it had not provided them.

The evidence we have summarized is all the evidence on the question of gross negligence and malice that the respective sides even claim supports their position. Viewing the evidence in its entirety, we conclude that the evidence is factually insufficient to support the jury findings of gross negligence and malice, and that those findings are so against the great weight and preponderance of the evidence as to be manifestly unjust.

The evidence is barely sufficient to find ordinary negligence; it falls far short of supporting a finding of gross negligence, and certainly not malice. The principal reason why this is true is that Ford relied on studies by NHTSA that consistently showed the risk-utility balance of tension eliminators weighed in favor of overall safety, and that the kind of tension eliminators Ford used were not unreasonably dangerous.

When a seller relies in good faith on the current state of the art in safety concerns, and on conclusions by the governmental agencies charged with administering safety regulations in the area of its product that the product is not unreasonably dangerous, it cannot be said to have acted with an entire want of care showing conscious indifference to the safety of the product users, or to have acted with conscious indifference to an extreme degree of risk. Nor can it be said that the seller acted "with a flagrant disregard for the rights of others and with actual awareness . . . that the act will, in reasonable probability, result in human death [or] great bodily harm." In fact, the undisputed evidence here shows that Ford went to great lengths to assay the relative risks and benefits of its restraint system and made a decision based on the evidence it had. What sets gross negligence and malice apart from ordinary negligence is the culpable mental state of the defendant. There is no evidence of such a mental state here. While Ford's decision to use the tension eliminators may have turned out to be a mistake (ordinary negligence), it certainly cannot be said to have been a decision in conscious indifference of the safety of its customers or in spite of a known extreme risk of harm. Although plaintiffs had two expert witnesses testify that Ford's actions were grossly negligent and with malice, they based their opinions entirely on acts and policies that, at the most, constituted ordinary negligence, not gross negligence. The experts simply characterized ordinary negligent acts as gross negligence and malice. Their opinions were thus conclusions without basis in fact.

The jury findings of gross negligence and malice are supported by factually insufficient evidence and are so against the great weight and preponderance of the evidence as to be manifestly unjust. Thus, the award of punitive damages was improper, and we will reverse and remand the gross negligence and malice issues for a new trial.

III. PLAINTIFFS' LOSS OF CONSORTIUM CLAIMS

Ford and Stanley moved for partial summary judgment on Kenneth Miles's and Jermaine Searcy's loss of consortium claims. The court granted the motion on grounds that, as a matter of law, Miles and Searcy did not have the required relationships to recover for loss of consortium.

The relevant facts are not contested. The plaintiffs seek to extend the right to recover for loss of consortium to the relationships between stepparent and stepchild and between siblings.

Texas first recognized a cause of action by a child for loss of consortium of a parent in the case of Reagan v. Vaughn, 804 S.W.2d 463 (Tex. 1990). In that case, however, our Supreme Court limited the cause of action to the parent-child relationship. In analogous situations, both federal and state courts in Texas have specifically declined to extend the wrongful death statute to stepparents. Byrnes v. Ford Motor Co., 642 F. Supp. 309 (E.D.Tex. 1986); Robinson v. Chiarello, 806 S.W.2d 304, 310 (Tex.App. — Fort Worth 1991, writ denied). Ford and Stanley argue that Texas also has no common law action for tortious injury to a family relationship in favor of the siblings of the deceased. Williams v. Texaco Ref. Mktg., 868 S.W.2d 873, 875 (Tex.App. — Houston [1st Dist.] 1993, no writ).

The plaintiffs argue that wrongful death actions are creatures of statute and so may be subject to limits that are not appropriate when considering purely common law actions, such as tortious injury to family relationships. They argue that in these days of mixed families, a step-parent may be closer emotionally to a child than a natural parent would be. They also argue that sibling relationships are close family relationships. This is a persuasive argument, but in view of our Supreme Court's holding in Reagan v. Vaughn, supra, we decline to expand the cause of action beyond the parent-child relationship.

IV. DEFENSE EVIDENTIARY COMPLAINTS

Ford and Stanley contend that the trial court erred in allowing the plaintiffs to present certain expert testimony while refusing to allow Ford and Stanley to present rebutting testimony.

Plaintiffs' expert Steve Syson testified that during the crash Willie "submarined" under his lap belt. He said this explained why Willie's knees, but not his head, hit the dash. As Willie's body slid under his lap belt, his chin got caught in the shoulder harness and this contributed to his neck injury, Syson said. Ford and Stanley argue that this submarining theory was a previously undisclosed theory of injury that the plaintiffs failed to disclose pursuant to TEX.R.CIV.P. 166b(6)(b), which requires parties to reveal the substance of their expert testimony no later than thirty days before trial. They further argue that TEX.R.CIV.P. 215(5) required the trial court to exclude the testimony. They say the error was harmful to them because their trial theory was that Willie wore his torso belt correctly, which was confirmed by evidence that Willie's knees, but not his head, hit the dash. They say plaintiffs' new theory of submarining allowed the plaintiffs to undercut their defense without giving them a chance to counter the new theory.

Ford and Stanley failed to object to the introduction of this purportedly new theory until Syson had completely introduced it and after Ford and Stanley had raised other unrelated objections.

A party opposing the admission of evidence must specifically object at the time it is offered to lay a predicate for appellate review of the trial court's action. TEX.R.APP.P. 52(a); Zamora v. Romero, 581 S.W.2d 742, 747 (Tex.Civ.App. — Corpus Christi 1979, writ ref'd n.r.e.). An objection to the introduction of the evidence if not timely made is waived. An objection made after the evidence has been admitted is too late. Wenco of El Paso/Las Cruces v. Nazario, 783 S.W.2d 663, 665 (Tex.App. — El Paso 1989, no writ); J.A. Robinson Sons, Inc. v. Wigart, 420 S.W.2d 474, 486 (Tex.Civ.App. — Amarillo 1967), rev'd on other grounds, 431 S.W.2d 327 (Tex. 1968); Hix v. Wirt, 220 S.W.2d 530 (Tex.Civ.App. — Waco 1949, writ ref'd n.r.e.).

The jury had already heard the testimony when it was objected to. Even allowing for the complexity of the case and a momentary lapse before Ford and Stanley realized Syson was testifying about a purportedly new theory, the objection came too late.

Ford also argues that the court improperly excluded its rebuttal witnesses in connection with the submarining testimony.

The court in its fourth pretrial order required the parties to list their trial witnesses by September 19, 1994, but excluded "rebuttal or impeachment witnesses, the necessity for whose testimony cannot be reasonably anticipated." On September 12, 1994, the Tyler Court of Appeals stayed further discovery. While the stay was in effect, Ford says, the September 19 deadline for listing trial witnesses passed. On October 5, 1994, the Tyler court lifted the discovery stay, and on November 23, the trial court extended its discovery deadline.

Ford argues that questions about possible marks on Willie's neck first became material on September 6, 1994, when Syson testified in a pretrial hearing that a belt mark on Willie's neck might be significant in describing how Willie's neck was injured in the crash. Ford argues that it could not list its rebuttal witnesses, Dunn and McGrath, until they had been deposed so Ford could learn the substance of their testimony. Dunn and McGrath were deposed in December. Both testified they saw no evidence of neck trauma on Willie immediately after the wreck. Ford then listed them January 13, 1995, on an amendment to its trial witness list. On plaintiffs' motion, the court excluded their testimony because they had not been listed on the September 19 witness list.

The plaintiffs argue that Ford knew the substance of McGrath's and Dunn's knowledge and knew the relevance of Willie's possible neck trauma before the September 19 cutoff. McGrath's notes were part of four volumes of original hospital records provided to the defense on April 25, 1994, the plaintiffs say. The plaintiffs identified Dunn in their original answers to interrogatories on May 19, 1994. The emergency medical services report, which mentioned blunt trauma to neck, was part of the hospital records available to the defense. Plaintiffs' expert Louis D'Aulerio mentioned in his report a "pronounced abrasion on the neck which ran under the jaw and across both sides of the neck." D'Aulerio in his deposition also testified that the evidence showed the torso belt shifted up and rode underneath the chin on the high side of the neck and that Willie's body jackknifed forward until the torso belt slack was taken up and he hit the shoulder harness in and around the neck region. Ford also deposed paramedic Randy Rodriguez on October 19, 1994, and questioned him about marks on Willie's neck. The plaintiffs also say that McGrath and Dunn testified at their depositions that they had no independent recollection of the events apart from their records, which were before the jurors as trial exhibits.

The question of whether the trial court abused its discretion by not allowing McGrath and Dunn to testify about the purported absence of marks on Willie's neck is, in part, a question of whether Ford and Stanley knew of evidence about possible marks on Willie's neck and the possible significance of the marks. This rests in part on the question of whether the "submarine" theory is a new theory introduced at trial to replace or supplement the "jackknife" theory. The trial court was not forced to specifically determine whether the submarine theory was new because the defense failed to timely object to Syson's testimony, which purportedly introduced the theory. If the theory is new, then the testimony could not be reasonably anticipated and the court should have admitted the testimony as rebuttal testimony in accord with its pretrial order. If the theory was not new, Ford and Stanley could have reasonably anticipated the testimony and the court's decision to exclude the testimony was within its discretion.

We have difficulty determining whether the theory is new because we have only the trial record to guide us. The plaintiffs argue that Syson's trial testimony was consistent with his report and deposition. He testified that his report said, "The weakness of the right front passenger seat also contributes to the injury potential at that position by failing to assist the lap belt in restraining [Willie's] lower torso." Plaintiffs' expert D'Aulerio in his report also said the seat belt allowed Willie to move forward and make contact with internal vehicle surfaces.

Ford argues that Syson in his deposition testified there was no evidence of submarining. Syson on cross-examination said that during his deposition defense attorneys asked him whether there were any submarine injuries, that is, injuries to the stomach and lower back, to which he replied negatively. He also testified in his deposition that the seat and the seat belt did not cause submarining. The defense during the deposition apparently did not specifically ask Syson whether submarining had contributed to Willie's neck injury.

It is unclear whether the plaintiffs' version of the crash has Willie submarining under his lap belt and catching his chin on the torso belt, or having him jackknife forward at the hips, hitting the slack torso belt and catching his throat on the torso belt, or a combination of the two motions.

The theory is not, perhaps, wholly new. Whether there was neck trauma and a visible neck injury, whether the result of jackknifing or submarining, always was part of the facts and always has been significant. The question of whether Willie had marks on his neck was addressed in part by the evidence. The trial court must control the flow of trial evidence and there must be an end of discovery. We conclude that the trial court's decision to exclude the testimony falls within that zone of reasonable disagreement. Cf., e.g., Montgomery v. State, 810 S.W.2d 372 (Tex.Crim.App. 1990) (opinion on rehearing). Even if the trial court erred, we cannot say the witnesses' exclusion was reversible error. TEX.R.APP.P. 81(b)(1). Their deposition testimony was based not on independent memory but solely on their records, which were before the jury as trial exhibits.

V. ADMISSION OF TRIAL EXHIBITS

Ford also complains in its fourth point of error because the trial court admitted en masse some 340 trial exhibits, comprising some 7,000 pages. It contends that the trial court ignored considerations of authenticity, hearsay, relevance, and whether the exhibits' probative value outweighed their prejudicial value, and admitted the exhibits as a group.

Both Ford and the plaintiffs on appeal concede that the court granted a limiting instruction advising the jurors that they were to consider the exhibits only for the limited purpose of showing that the exhibits were in Ford's possession. It is difficult to tell from the statement of facts to which exhibits the limiting instruction applied, but we presume it applied to all the listed documents.

The trial court in its fourth pretrial order told the parties to prepare written objections to exhibits. Ford filed more than ninety pages of objections, each page listing multiple exhibits and often with multiple complaints about each exhibit. The court filed written orders in response to the plaintiffs' and defendants' objections to exhibits, in which the court overruled objections, sustained objections, and carried forward complaints.

Ford and Stanley apparently wanted the court to rule on each day's exhibits in the morning of each trial day, whereas the court opted for making most of the evidentiary decisions before trial. In a pretrial hearing, the court gave Ford the opportunity to object to any proposed trial exhibit, and Ford said it was relying on its written objections. The plaintiffs offered the exhibits before trial in broad categories — scientific literature, public records, and Ford documents. When the exhibits were introduced, the court gave its limiting instruction.

The record does not support Ford's allegations that the court admitted the exhibits en masse. Ford submitted objections to the exhibits, the court gave Ford the opportunity to further argue its objections before trial, and the court issued written orders as to the trial exhibits. The record shows the trial court did not rule on the exhibits en masse, but individually before trial.

In its point of error, Ford argues that the following exhibits were irrelevant and prejudicial.

(1) Plaintiffs' Exhibit 557, a 1967 paper summarizing restraint system tests on adult baboons.

(2) Plaintiffs' Exhibit 88, an article from an unidentified newspaper or magazine reporting that Kangol Magnet, Ltd. claimed to have developed a new restraint system that avoided the danger of excess slack.

(3) Plaintiffs' Exhibit 361, a photograph of a crash test of a 1992 Ranger.

(4) Plaintiffs' Exhibit 328, a list of "things gone wrong" with the interior trim of a Lincoln Continental.

(5) Plaintiffs' Exhibit 195, Transport Canada's concerns relating to passive restraint systems.

(6) Plaintiffs' Exhibit 204A, a page from a professor's 1974 report finding that shoulder harnesses do not cause injuries to female breasts.

(7) Plaintiffs' Exhibit 425, a list of "things gone wrong" with 1985 and 1987 F-Series pickups and Broncos.

(8) Plaintiffs' Exhibit 97, a 1980 letter from NHTSA Administrator Joan Claybrook criticizing the industry for not doing enough for auto safety. The letter mentions restraint systems but does not specifically mention tension eliminators.

(9) Other documents discussing various kinds of restraint systems. See, for example, Plaintiffs' Exhibits 352-365 and 368-375.

(10) Other documents discussing the development, design, and testing of air bags and passive belt systems from the 1960s to the 1990s. See, for example, Plaintiffs' Exhibits 174, 201, 203, 204, 206, 208, 270, 299, 302, 307, 310, 312, 314, 322, and 344.

Courts hold all manufacturers to an expert's knowledge and skill. Borel v. Fibreboard Paper Prods. Corp., 493 F.2d 1076, 1088 (5th Cir. 1973), cert. denied, 419 U.S. 869, 95 S.Ct. 127, 42 L.Ed.2d 107 (1974); Fibreboard Corp. v. Pool, 813 S.W.2d 658, 668 (Tex.App. — Texarkana 1991, writ denied), cert. denied, 509 U.S. 923, 113 S.Ct. 3037, 125 L.Ed.2d 724 (1993). In Pool, this court held that a limiting instruction telling the jurors that documents in the defendant's possession could not be considered for the purpose of determining the actual knowledge or state of mind of a defendant was too broad. Fibreboard Corp. v. Pool, 813 S.W.2d at 670.

The jurors could not consider the documents for the truth of any matter asserted, but could consider the documents as proof that Ford had the documents and presumably was aware of any relevant issues, complaints, or criticism contained in the documents.

Ford argues this requires the jurors to consider the truth of the matters asserted because, unless the matters asserted are true, Ford need not act on any assertion leveled in any document. Ford's argument goes too far. Even if matters asserted in some documents are false, for example, in an inaccurate news story, Ford still would have been aware of the criticism and issues the document raises and could be deemed to have the knowledge of the issue, criticism, or complaint, even an unjustified criticism or complaint. This presupposes, however, that the documents, though false, had some relevance to the issues in this case.

The documents numbered (1), (4), (6), (7), and (8) above and specifically challenged here by Ford should not have been admitted. They did not address or apply, even tangentially, to the issues in this case. They had no relevance at all and could have been prejudicial. We conclude, however, that, especially in view of the limiting instruction, they were not so prejudicial that their admittance into evidence probably caused the rendition of an improper verdict.

Reviewing courts generally leave to the trial court's discretion questions of relevancy, Wells v. State, 880 S.W.2d 185, 188 (Tex.App. — Texarkana 1994, no pet.), and the weighing of the probativeness of evidence versus its prejudicial nature, Fibreboard Corp. v. Pool, 813 S.W.2d at 671. The court admitted the documents as general proof that Ford had knowledge of the issues and complaints in its field. It was relevant for the purpose submitted. The limiting instruction also may have cured any problems with prejudice because, for the documents to inflame the jurors' feelings, the jurors would have had to consider the truth of the matters raised, which the trial court instructed the jurors not to do. We assume the jurors followed the court's limiting instruction. Turner, Collie Braden v. Brookhollow, Inc., 642 S.W.2d 160, 167 (Tex. 1982). Even if the trial court abused its discretion, it is not clear the defendants demonstrated harm. TEX.R.APP.P. 81(b)(1).

VI. TESTS

Ford also complains because the trial court admitted into evidence tests offered by the plaintiffs and excluded the tests Ford offered.

1. The D'Aulerio Videotape

The plaintiffs introduced a videotape that showed their expert D'Aulerio seated in a Ford Ranger pickup truck, buckling the seat belt and shoulder harness, and demonstrating how a tension eliminator works. He leans forward and then leans back. Slack accumulates in the shoulder harness when he leans forward and folds behind his back as he leans back. Ford contends that this demonstration should not have been admitted because it was not a proper reenactment of what happened to Willie Searcy.

A trial court should admit evidence of an out-of-court experiment only when there is substantial similarity between conditions existing at the time of the occurrence giving rise to the litigation and the conditions created by the experiment. Kainer v. Walker, 377 S.W.2d 613 (Tex. 1964), overruled on other grounds, Burk Royalty Co. v. Walls, 616 S.W.2d 911 (Tex. 1981); Fort Worth Denver Ry. Co. v. Williams, 375 S.W.2d 279, 281-82 (Tex. 1964). A film intended for even a limited purpose, such as illustrating a principle, must show sufficient similarity to the actual accident to be admissible. Lopez v. Foremost Paving, Inc., 796 S.W.2d 473, 480 (Tex.App. — San Antonio 1990, writ dism'd). The conditions do not have to be identical. Garza v. Cole, 753 S.W.2d 245, 247 (Tex.App. — Houston [14th Dist.] 1987, writ ref'd n.r.e.). When there is dissimilarity in the conditions, the admission of the experiment is within the trial court's discretion if the differences are minor or subject to explanation. City of Dallas v. Cox, 793 S.W.2d 701, 734 (Tex.App. — Dallas 1990, no writ). The offering party's affirmative acknowledgment to the jury of dissimilarities between a taped reconstruction and the actual occurrence can serve to alleviate unfairness. Lopez v. Foremost Paving, Inc., 796 S.W.2d at 480-81.

D'Aulerio testified that he was not Willie's size, that the Ford Ranger was selected at a used car lot, and that he did not intend to demonstrate how the accident occurred but only how tension eliminators work and how the user's movement could inadvertently introduce slack. Although he did use the word "enactment" before the jurors, he corrected himself and said he was not trying to reenact what happened in the accident.

Ford argues that D'Aulerio's movements are markedly different from the movements Willie made and that the circumstances of the demonstration were materially different from the conditions before the accident. It particularly complains about D'Aulerio's size as compared to Willie's size.

Ford introduced a similar videotape in which Ford executive Maugh demonstrated a restraint system with a tension eliminator. The tape purported to show that a person leaning forward would not introduce slack into the shoulder harness.

Plaintiffs' counsel and the witness told the jurors about the dissimilarities between the tape and the occurrence. The trucks and seat belts are similar to Kenneth Miles's truck and the seat belts in his truck. Although the conditions are not identical, the dissimilarities were minor considering the demonstration's purpose, and the plaintiffs explained that the tape was not an exact depiction of what happened to Willie. The decision to admit the demonstration was within the trial court's discretion, and we cannot say the trial court abused its discretion in this instance.

2. Ford Sled Tests

Ford, however, offered evidence of some sled tests it conducted, and the court, after a pretrial hearing, excluded the evidence.

The plaintiffs defend the trial court's exclusion of Ford's tests by saying that the tests were not sufficiently similar to what happened in the collision. They argue that (1) Ford used a dummy with a different seat height, which weighed more and which had its weight distributed differently than did Willie; (2) the shoulder belt was taped to the dummy's chest to keep it from moving, whereas Willie's shoulder belt was not taped to his chest; (3) the crash pulse used by Ford in the test was not substantially similar to the actual event; (4) the sled test represented only head-on forces, whereas the actual crash involved both head-on and lateral forces; and (5) the dummy's seat in the sled test was in the mid-position, whereas Willie's seat in the collision was in full rear position.

Ford counters by asserting that the dummy was as close as currently available to Willie's weight and height; the tape was irrelevant to the test results because the test-crash forces tore the tape off the chest; the crash test pulse used resulted in a more conservative test; evidence showed that the head-on forces, not the lateral forces, were the only forces that injured Willie; and evidence showed that Willie's seat was in the mid-position at the time of the crash.

While Ford's tests had some dissimilarities, so did plaintiffs' videotape. The dissimilarities in Ford's tests were fully admitted and could have been explained to the jury. Thus, having admitted plaintiffs' admittedly dissimilar tests, the trial court erred in excluding Ford's tests, when their dissimilarities were fully explained. The court, having adopted a stance of admitting demonstration evidence, even though slightly dissimilar, should have evenly applied that policy and admitted Ford's sled tests.

We cannot say, however, that the error in excluding evidence of these tests was of such gravity as to have probably caused the rendition of an improper judgment and to constitute reversible error. There was an abundance of evidence about how the restraint system and tension eliminator worked and how they would affect a person of Willie's size and age in a car crash. We do not believe that the exclusion of these tests was so critical as to cause the rendition of an improper judgment.

3. Plaintiffs' Crash Test Tapes

Ford also contends that the trial court erred by admitting videotapes of Ford crash tests that show crash test dummies being decapitated.

The plaintiffs did not claim that these tests were substantially similar to the accident, but rather submitted the tests as impeachment evidence to show Ford knew that slack in belts was dangerous and to refute statements by Ford's witnesses that the company had conducted no dummy tests with slack in the torso belts.

The trial court, if requested, must balance the probative value of impeachment evidence against the possibility of undue prejudice. See Perkins v. State, 887 S.W.2d 222, 226 (Tex.App. — Texarkana 1994, pet. ref'd). The appellate court, however, must accord the trial court wide discretion in weighing the relevant factors and in deciding whether to admit impeachment evidence. See Polk v. State, 865 S.W.2d 627, 631 (Tex.App. — Fort Worth 1993, pet. ref'd); see also Howell v. Burch, 616 S.W.2d 685, 688 (Tex.Civ.App. — Texarkana 1981, writ ref'd n.r.e.); Ledisco Fin. Serv., Inc. v. Viracola, 533 S.W.2d 951, 958 (Tex.Civ.App. — Texarkana 1976, no writ); Roberts v. Dallas Ry. Terminal Co., 276 S.W.2d 575, 577-78 (Tex.Civ.App. — El Paso 1953, writ ref'd n.r.e.); Missouri, K. T. Ry. Co. v. Bailey, 53 Tex. Civ. App. 295 [ 53 Tex. Civ. App. 295], 115 S.W. 601, 607-08 (1908, writ ref'd).

If the tests were admissible as impeachment evidence, similarity to the actual crash events would be immaterial, because the tests' value would not be for the substantive information presented but for their impeachment value. Verbal evidence about the tests was admissible for impeachment purposes and to show that Ford knew of the dangers of slack torso belts. The Court of Criminal Appeals has long held that if a verbal description of an event or scene would be admissible, a photographic depiction of the same is admissible. Emery v. State, 881 S.W.2d 702, 710 (Tex.Crim.App. 1994), cert. denied, ___ U.S. ___, 115 S.Ct. 1257, 131 L.Ed.2d 137 (1995). The Supreme Court, however, in cases predating the Rule of Civil Evidence, has been reluctant to allow photographs "that are merely calculated to arouse the sympathy, prejudice or passion of the jury" where the photographs do not serve to illustrate disputed issues or aid the jury in understanding the case. Heddin v. Delhi Gas Pipeline Co., 522 S.W.2d 886, 889-90 (Tex. 1975), and cases there cited; but see State v. City of Greenville, 726 S.W.2d 162, 168 (Tex.App. — Dallas 1986, writ ref'd n.r.e.); cf. Castro v. Sebesta, 808 S.W.2d 189, 193 (Tex.App. — Houston [1st Dist.] 1991, no writ); 1 STEVEN GOODE, OLIN G. WELLBORN III M. MICHAEL SHARLOT, GUIDE TO THE TEXAS RULES OF EVIDENCE: CIVIL AND CRIMINAL § 403.2 (Texas Practice 1993).

The videotapes in question were not gory, although the test dummies' heads did come off, suggesting severe human injuries. Considering that the tapes were only crash tests with dummies rather than actual crashes, and the tapes were admissible for the limited purpose of showing that Ford had conducted slack belt crash tests, we do not find that allowing the tapes in evidence was reversible error.

4. Data From Unknown Tests

Ford also complains about Plaintiffs' Exhibit 227, one page entitled "Belt Slack Analysis — 3 Point System, Passenger Side, Preliminary," which Ford produced during discovery. A Ford attorney, out of an abundance of caution in responding to discovery requests, wrote a letter to plaintiffs' counsel enclosing the document and saying that the document was the only known evidence suggesting that slack belt tests may have been conducted "by or for Ford," and that Ford did not know further specifics of any tests. The plaintiffs offered the document pretrial as a trial exhibit, and Ford objected on hearsay grounds. The plaintiffs offered the Ford attorney's letter to authenticate the document, but the trial court sustained the objection. Before trial, the judge admitted the document but excluded the letter. At trial when Ford in its cross-examination of the witness Syson attempted to demonstrate that the document could not be authenticated as coming from Ford or as proving that Ford had conducted slack-belt experiments, the plaintiffs again offered the letter as proof of the document's authenticity. The court admitted the letter.

Although the letter might have been admissible as impeachment evidence to show that Ford's counsel denied the authenticity of the document in court while conceding its authenticity in the letter, the letter does not authenticate the document. Authentication as a condition precedent to admissibility is established by evidence that the matter in question is what its proponent claims. TEX.R.CIV.EVID. 901(a). This evidence can consist of testimony by a witness that the matter is what it is claims to be. TEX.R.CIV.EVID. 901(b)(1). A document can be considered authentic if a sponsoring witness vouches for its authenticity. In re G.F.O., 874 S.W.2d 729, 731 (Tex.App. — Houston [1st Dist.] 1994, no writ). The attorney who wrote the letter was not a witness nor was the letter sworn to, so the letter does not authenticate the document. Thus, plaintiffs did not lay a proper predicate for the introduction of the document, and it was error to admit it, even for the limited purpose of impeachment.

There was abundant other evidence, however, that Ford had knowledge of various tests of restraint systems, and we do not believe that the erroneous admission of this single-page document among the volumes of evidence in this trial was so critical that it probably caused the rendition of an improper judgment. TEX.R.APP.P. 81(b)(1).

VII. COURT'S CHARGE

Ford also contends that the trial court erred in its jury instructions. The court told the jurors:

The manufacturer of a motor vehicle has a duty to reasonably design, manufacture and market the vehicle so as not to subject occupants of the vehicle to unnecessary and avoidable injuries in the event of a collision. A motor vehicle manufacturer is not an insurer of the safety of its

product and is only required to design, manufacture and market a product that is reasonably safe for its intended use.

(Emphasis added.)

Ford argues that the instruction materially misstates Texas law and was surplusage that wrongly deflected the jurors' attention from the correct instructions. Specifically, Ford complains because the instruction told the jury that Ford had a duty to prevent "all unnecessary and avoidable injuries." It argues that, by telling the jurors that a reasonably designed product would prevent unnecessary and avoidable injuries, the instruction invited the jurors to find for the plaintiffs if they believe a different design could have prevented Willie's injuries even if the actual design prevented thousands of injuries in other types of accidents.

In the case of Acord v. General Motors Corp., 669 S.W.2d 111 (Tex. 1984), our Supreme Court reapproved the jury charge approved in Turner v. General Motors Corp., 584 S.W.2d 844 (Tex. 1979), and set out in the Pattern Jury Charges as the only approved jury charge in design defect, manufacturing defect, and marketing defect cases. The instruction is applicable to cases where a defect causes a crash, as well as to crashworthiness cases, i.e., where it is alleged that injuries were rendered more severe because of the defect. Moreover, the Supreme Court in Acord v. General Motors Corp., supra, expressly disapproved of adding any other considerations or balancing factors to the approved instruction. It said:

We explicitly approve the Pattern Jury Charges issue and instruction for design defect cases, and disapprove the addition of any other instructions in such cases, however correctly they may state the law under § 402A of the Restatement (Second) of Torts.

Acord v. General Motors Corp., supra at 116. It was therefore error to give the additional instruction in this case.

We must consider the pleadings, the trial evidence, and the charge as a whole in determining whether a jury charge error constitutes reversible error. Island Recreational Dev. Corp. v. Republic of Texas Savings Ass'n, 710 S.W.2d 551, 555 (Tex. 1986). We may find the error reversible only if, when viewed in the light of all these circumstances, the error amounted to such a denial of the rights of the complaining party that it was reasonably calculated to and probably did cause the rendition of an improper judgment. TEX.R.APP.P. 81(b)(1).

Although it was error to give the additional instruction, we do not believe it was reversible error. The added instruction informed the jurors that Ford had the duty to design, manufacture, and market its vehicles so that they would not subject occupants to unnecessary and avoidable injuries. We believe that, standing alone, the instruction so worded would be reversible error, because it expands the manufacturer's duty beyond that given in the instruction approved in Acord v. General Motors Corp., supra. We believe, however, that the additional sentence immediately following the objectionable one, viz: "A motor vehicle manufacturer is not an insurer of the safety of its product and is only required to design, manufacture and market a product that is reasonably safe for its intended use " (emphasis added), sufficiently modified the former sentence to bring the overall instruction back to the requirements of Acord v. General Motors Corp., supra — that defectively designed means unreasonably dangerous, taking into consideration the utility and risk. Thus, although the additional charge should not have been given, we believe that, in its entirety, its gist is consistent with the approved instruction in cases of this kind and that it did not mislead the jury or "nudge" it toward using an improper standard, and therefore did not constitute reversible error.

VIII. DISPOSITION

For the reasons stated, we affirm the judgment in all respects except the findings of gross negligence and malice as a basis for the award of punitive damages. Although we are prohibited from ordering a separate new trial on unliquidated damages when liability is contested, TEX.R.APP.P. 81(b)(1), when the improper damages relate to clearly different and separable theories of liability from those supporting proper damages, we can order a new trial on the improper damage issues. The burden of proof for punitive damages cannot be satisfied by evidence of ordinary negligence. TEX.CIV.PRAC. REM.CODE ANN. § 41.003(b). Because ordinary negligence is based on ordinary care regardless of intent, and gross negligence and malice are based on the mental state and intent of the actor, recovery under those different theories depends on separate and distinct elements of proof, and they are clearly separable. See Alm v. Aluminum Co. of America, 753 S.W.2d 478 (Tex.App. — Houston [14th Dist.] 1988), rev'd on other grounds, 785 S.W.2d 137 (Tex. 1990); Otis Elevator Co. v. Joseph, 749 S.W.2d 920 (Tex.App. — Houston [1st Dist.] 1988, no writ). The issues of gross negligence, malice, and punitive damages are severed from the remainder of the judgment. The judgment in those respects is reversed, and those issues are remanded for a new trial. In all other respects, the judgment is affirmed.

ON REHEARING

Ford argues, among other things, in its motion for rehearing that we cannot properly order a retrial on the issues of gross negligence and malice as a basis for punitive damages. It argues that such a partial remand would (1) violate the Texas rule against remand on damages only when liability is contested, TEX.R.APP.P. 81(b)(1); and (2) would violate its due process rights because such a partial remand and retrial cannot be held without unfairness. We disagree with these contentions for the following reasons.

We first point out that we are not reversing the punitive damage award itself for insufficient evidence or for excessiveness. Rather, we are reversing the underlying liability findings that authorize an award of punitive damages. Consequently, the partial remand will not violate the rule against a partial remand on damages apart from liability; it will rather be a remand on the issues of gross negligence and malice as bases for punitive damages, apart from the issue of ordinary negligence as a basis for compensatory damages. That kind of partial remand has been approved by Texas courts because gross negligence and malice are based on different standards of conduct from those of ordinary negligence, and are clearly distinct. See, e.g., McElroy v. Fitts, 876 S.W.2d 190 (Tex.App. — El Paso 1994, writ dism'd by agr.); Alm v. Aluminum Co. of America, 753 S.W.2d 478, 485-86 (Tex.App. — Houston [14th Dist.] 1988), rev'd on other grounds, 785 S.W.2d 137 (Tex. 1990); Olin Corp. v. Dyson, 678 S.W.2d 650, 659 (Tex.App. — Houston [14th Dist.] 1984), rev'd on other grounds, 692 S.W.2d 456 (Tex. 1985).

Ford relies on Gasoline Products Co. v. Champlin Refining Co., 283 U.S. 494, 500, 51 S.Ct. 513, 515, 75 L.Ed. 1188, 1191 (1931). In that case, the Supreme Court held that an appellate court cannot remand a case for a partial new trial unless the issue to be retried is so distinct and separable from the others that the trial of it alone may be had without injustice. The court found that the question of damages on a counterclaim was so interwoven with that of liability that the jurors could not consider the damage claim without confusion and uncertainty and that this would amount to a denial of a fair trial. The Court said, however, "Here we hold that where the requirement of a jury trial has been satisfied by a verdict according to law upon one issue of fact, that requirement does not compel a new trial of that issue even though another and separable issue must be tried again." Gasoline Products Co. v. Champlin Refining Co., 283 U.S. at 499, 51 S.Ct. at 515.

All of the cases Ford cites involve the impropriety of remanding damages alone for partial retrial, not remanding gross negligence or malice issues, or they involve other facts that distinguish them from the situation here.

In fact, Ford cites McDonald v. Johnson Johnson, 722 F.2d 1370 (8th Cir. 1983), cert. denied, 469 U.S. 870, 105 S.Ct. 219, 83 L.Ed.2d 149 (1984), which supports a partial remand in our situation. There, the trial court gave the jurors improper instructions involving malicious or oppressive conduct, which would support punitive damages. The Eighth Circuit affirmed the trial court's judgment on the fraud verdict and remanded for a new trial on what the court called "punitive damages," but what was really the issue of liability supporting punitive damages. Courts sometimes confuse the issue by referring to punitive damage issues when they actually are discussing gross negligence, malice, or other egregious conduct justifying punitive damages.

Ford relies on three cases involving the remand of underlying liability issues: Burke v. Deere Co., 6 F.3d 497, 513 (8th Cir. 1993), Western Fireproofing Co. v. W.R. Grace Co., 896 F.2d 286 (8th Cir. 1990), and Atlantic Coast Line R.R. Co. v. Bennett, 251 F.2d 934, 938-39 (4th Cir. 1958).

In Burke v. Deere Co., supra, the appeals court, applying Iowa law, found "no evidence of any conduct so egregious as to support an award of punitive damages." The majority said that remand of the punitive damage issue would violate Gasoline Products Co. v. Champlin Refining Co., supra. Among the court's considerations was Iowa's practice of admitting evidence of the defendant's net worth in connection with exemplary damages when jurors also are considering compensatory damages. Texas, however, uses a bifurcated system in which exemplary damages, but not liability, are determined in a separate trial. Moreover, a dissenting judge in that case noted that Champlin specifically would permit retrial of the issue of whether Deere had acted with "willful and wanton disregard of the rights or safety of another." Burke v. Deere Co., 6 F.3d at 514 n. 2 (Heaney, J., dissenting).

In Western Fireproofing Co. v. W.R. Grace Co., supra, the appeals court remanded for trial on all issues after the trial court gave the jurors an improper malice instruction. The court noted, however, that the evidence was legally sufficient to submit the punitive damage issue to the jurors.

In Atlantic Coast Line R.R. Co. v. Bennett, supra, the court found that the evidence of willful misconduct was inextricably intertwined with that relating to primary negligence, and remanded both the willful misconduct and the negligence claims for retrial. The court relied on Smyth Sales v. Petroleum Heat Power Co., supra, in which the court found the amount of punitive damages excessive. There the court said Champlin barred a trial on the amount of damages alone. The amount of damages, not the question of liability supporting the damages, was at issue.

Ford also cites Redman Homes, Inc. v. Ivy, 920 S.W.2d 664 (1996). That case involved a partial remand on actual damages alone, apart from any liability issue. As the Supreme Court noted, that clearly violated TEX.R.APP.P. 81(b)(1). That is not what we have in this case.

Ford also has cited us to Hyman Farm Service, Inc. v. Earth Oil Gas Co., 920 S.W.2d 452 (Tex.App. — Amarillo 1996, n.w.h.), in which the court holds that, even in separate or bifurcated trials of different issues, the same ten jurors must agree and sign the verdict on the same issues. TEX.R.CIV.P. 292. The holding in that case is not in point on the issue we consider. In that case there was no severance — only separate trials of issues in the same case. When there is a severance, the suit is divided into separate and independent causes. Hall v. City of Austin, 450 S.W.2d 836 (Tex. 1970). The Hyman case applies to jury answers in one cause, even though it is tried piecemeal. Our severance and remand of the gross negligence and malice issues as a basis for punitive damages will make those issues a separate cause, and Rule 292 will not require that the same ten jurors who heard the original case hear and concur in the verdict on retrial.

We cannot agree that the partial remand will cause unfairness to Ford. As noted, the elements of ordinary negligence are entirely different from the elements of gross negligence and malice. Although the concepts are similar, the evidence of ordinary negligence cannot satisfy the burden of proof for either gross negligence or malice, TEX.CIV.PRAC. REM.CODE ANN. § 41.003(b), and the issues are not inextricably intertwined.

We find the other points in Ford's motion for rehearing to be without merit. For the reasons stated, the motion for rehearing is overruled.