SPAETH, President Judge:

This is a products liability case arising out of an automobile accident. The product in question is a radial tire, and the allegations are that the tire was defective because it was not embossed with a warning not to mix it with non-radial tires, and that the accident occurred because the tire was mixed with non-radial tires.

Appellees, Joann Dambacher and her parents, were the plaintiffs, and appellant, Sears, Roebuck and Company, the supplier of the tire, was one of the defendants. A jury returned a verdict in favor of appellees. The trial court denied appellant's motion for judgment n.o.v. or in the alternative for a new trial, and granted appellees' motion for a new trial limited to damages. There are two principal issues. The first issue is whether the trial court erred in ruling that certain of appellees' witnesses were qualified to express an opinion that in fact the accident did occur because the tires were mixed. We hold that the court did err, and we therefore order a new trial generally. The second issue concerns what instructions the trial court should give the jury at the new trial. We hold that the court should instruct the jury in accordance with Azzarello v. Black Brothers Co., Inc., 480 Pa. 547, 391 A.2d 1020 (1978). This means that the court should not instruct the jury to determine whether appellant's radial tire was "unreasonably dangerous" because it was not embossed with a warning not to mix it with non-radial tires. Nor should the court otherwise instruct the jury in negligence terms. Instead, the court should instruct the jury in terms of appellant's liability as a guarantor, responsible if the user of its product — the radial tire — was injured as a result of a defect in the product. The court should explain that if the jury finds that when the tire left appellant's control it lacked the warnings necessary to make it safe for its intended use, then the tire was defective, and appellant is liable for the harm caused by the defect.

Although the legal issues presented will require extended discussion, the underlying facts may be briefly stated. On November 6, 1977, Nicholas Mallis, a sixteen year-old high school student, discovered that the right front tire of his grandfather's 1971 Plymouth Fury was flat. He replaced the flat tire with a Sears radial tire. The other tires on the Plymouth were non-radial tires. The next day Nicholas gave seven fellow students a ride home from school in the Plymouth. One of the students was Joann Dambacher. As the youngsters left school, it was drizzling and the highway was wet, with leaves on the surface. Nicholas, driving at about 20 to 25 miles per hour, failed to negotiate an S-curve. He testified:

As I went through the right-hand turn I braked. I went through the right hand turn, started going into the left hand turn, and the back of the car slid out. The back slid to the right . . . It happened quick. As best I remember, the car slid a little sideways and towards the left lane, and it went off the embankment, into a tree.

R.R. 586a.

Joann suffered a fractured dislocation of the cervical spine, and was rendered a quadriplegic who will be confined to a wheelchair for life.

Joann's parents, William J. Dambacher and Joann Dambacher, on Joann's behalf and in their own right, sued Nicholas and his grandfather in negligence and Sears, in strict liability. Sears joined the Commonwealth of Pennsylvania, Department of Transportation, as an additional defendant on the theory that the S-curve was unsafe. Before trial, the Department of Transportation settled with the Dambachers for $87,500. The trial was before a jury, from May 6 to May 20, 1980. After a non-suit was entered in favor of Nicholas's grandfather, the jury returned a verdict in favor of Joann in the amount of $800,000, and in favor of her parents in the amount of $10,000, apportioning responsibility under the Pennsylvania Comparative Negligence Act 50% to Nicholas, 45% to Sears, and 5% to the Department of Transportation. Sears filed a motion for judgment n.o.v. or in the alternative for a new trial, and the Dambachers filed a motion for a new trial as to damages. The trial court, sitting en banc, denied Sears's motions, granted the Dambachers', and entered judgment against Sears as to liability, with delay damages. Sears then filed this appeal, and only its liability is at issue, for the Department of Transportation has settled, as mentioned, Nicholas has not appealed, and the propriety of the nonsuit in favor of Nicholas's grandfather is not questioned. For convenience, in the discussion that follows we shall usually refer to Sears as appellant and to the Dambachers as appellees, without distinguishing between Joann and her parents.

The admissibility of appellees' opinion evidence on causation

Appellees' theory of recovery against appellant was that appellant was strictly liable because the radial tire supplied to Nicholas's grandfather was not embossed with a warning not to mix it with non-radial tires. In support of this theory, appellees introduced documentary evidence and the testimony of two witnesses that Nicholas's mixing of the tires caused the accident. Appellant in turn introduced evidence that if the mixing was dangerous, it was so only at high speeds, not at 20 to 25 miles per hour, the speed at which Nicholas was driving when the accident occurred.

It is settled that in a products liability case the plaintiff must prove that a defective product was the proximate cause of his injuries. Sherk v. Daisy Heddon, 498 Pa. 594, 450 A.2d 615 (1982); Agostino v. Rockwell Manufacturing Co., 236 Pa. Super. 434, 345 A.2d 735 (1975). Appellant argues that "[t]he only testimony that a mixed fitment could have caused the accident was given by two of plaintiffs' [appellees'] witnesses, . . . whom the court below permitted to testify as `experts.' These witnesses were not qualified to give opinions on causation in this case, and therefore the verdict cannot be upheld on the basis of their opinions." Brief for Appellant at 36. Because of the trial court's error in ruling that appellees' two witnesses were qualified to express an opinion on causation, appellant argues, it is entitled not simply to a new trial but to judgment n.o.v.

-a-

It will be convenient to start with appellant's argument that it is entitled to judgment n.o.v., for in considering that argument we may assume, without deciding, that appellees' witnesses were unqualified to express an opinion on causation.

When we review an order denying a motion for judgment n.o.v., we must regard the evidence in the light most favorable to the verdict winner. Evidence supporting the verdict is to be considered, with the rest rejected, Glass v. Freeman, 430 Pa. 21, 240 A.2d 825 (1968), and "only when the facts are such that two reasonable persons could not fail to agree that the verdict was improper," Cummings v. Nazareth Borough, 427 Pa. 14, 25-26, 233 A.2d 874, 881 (1967), should we enter judgment n.o.v. See McKnight v. City of Philadelphia, 299 Pa. Super. 327, 445 A.2d 778 (1982).

Applying these principles here, we reject the testimony of appellant's witnesses and consider, in the light most favorable to appellees, the testimony of appellees' witnesses on causation. Both of these witnesses expressed the opinion that the mixing of the radial tire with three non-radial tires was a cause of the accident. Walter VanNess Pruyn testified as follows:

Q. What is your opinion?

A. It is my opinion that this accident was precipitated by a loss of control of the vehicle by the driver.

THE COURT: What do you mean by precipitated?

THE WITNESS: Initiated, caused to occur because of a combination of circumstances triggered by the mixing of a radial tire on the right front in combination with a cross ply belted tire on the left front of the vehicle which in combination with the highway surface, highway grade, curvature, inexperience of the driver, a lack of advanced knowledge of the potentialities for loss of control, the vehicle suddenly oversteered on the left turn and at the speed at which the vehicle was traveling the driver did not have time in which to correct, to react, to recognize possible solutions and to take action in the time distance available to him from the time the oversteer condition occurred and impact with the tree.

R.R. 630a-31a.

William P. Kelly testified that with a radial on the right front and a non-radial on the left front, the chance of an accident increases. R.R. 338a. He called "ridiculous" the statement of appellant's expert witness, Sidney Bloor, that a mixed fitment of radial and non-radial tires will not cause erratic steering unless the automobile is travelling at 70 miles per hour or more, even on wet ground. R.R. 996a. Other testimony was presented regarding the weather and the condition of the road, of the tires, and of the Plymouth at the time of the accident.

This record demonstrates that appellant's argument for judgment n.o.v. has no merit. For by accepting the testimony of appellees' witnesses, and rejecting that of appellant's, the jury could find that the mixed fitment caused the accident. The trial court therefore properly denied appellant's motion for judgment n.o.v.

Appellant cites several cases in which judgment n.o.v. has been granted when expert testimony was found, on appeal, to have been incompetent. Rennekamp v. Blair, 375 Pa. 620, 101 A.2d 669 (1954); Sinkovich v. Bell Telephone Company of Pennsylvania, 286 Pa. 427, 133 A. 629 (1926); Moyer v. Ford Motor Company, 205 Pa. Super. 384, 209 A.2d 43 (1965). These cases are inapposite, for the court granted judgment n.o.v. not because it found the witness unqualified to express an opinion but because the opinion was so equivocal as to be legally insufficient. See also Niggel v. Sears, Roebuck and Co., 219 Pa. Super. 353, 281 A.2d 718 (1971); McCrosson v. Philadelphia Rapid Transit Co., 283 Pa. 492, 129 A. 568 (1925). When the witness is unqualified, as appellant claims appellees' witnesses were, the proper remedy is a new trial, not judgment n.o.v. As our Supreme Court has said:

Manifestly it would be unfair, where a party has relied upon a favorable ruling on evidence presented by him, to enter final judgment against him without affording him the opportunity to furnish competent proof of which he might have availed himself had the evidence submitted by him been rejected. The only remedy under such circumstances is to grant a new trial.

Hershberger v. Hershberger, 345 Pa. 439, 29 A.2d 95 (1942).

See also, Stewart v. Chernicky, 439 Pa. 43, 266 A.2d 259 (1970); Weaverling v. Smith, 181 Pa. Super. 153, 124 A.2d 509 (1956).

-b-

In considering whether to order a new trial because of the trial court's ruling that appellees' witnesses were qualified to express an opinion on the cause of the accident, we must bear in mind that the ruling was within the sound discretion of the trial court. Accordingly, we will not order a new trial unless the ruling was such clear error as to constitute an abuse of discretion. Handfinger v. Philadelphia Gas Works, 439 Pa. 130, 266 A.2d 769 (1970); George I. Reitz Sons, Inc., 319 Pa. Super. 76, 465 A.2d 1060 (1983); Grubb v. Albert Einstein Medical Center, 255 Pa. Super. 381, 382, 387 A.2d 480 (1978); Flavin v. Aldrich, 213 Pa. Super. 420, 250 A.2d 185 (1968).

When a witness is offered as an expert, the first question the trial court should ask is whether the subject on which the witness will express an opinion is "so distinctively related to some science, profession, business or occupation as to be beyond the ken of the average layman." McCormick on Evidence 33 (3d ed. 1984) (footnote omitted). And see Commonwealth v. Leslie, 424 Pa. 331, 227 A.2d 900 (1967); Commonwealth ex rel. M.B. v. L.D.B., 295 Pa. Super. 1, 11, 440 A.2d 1192, 1197 (1982). If the subject is of this sort, the next question the court should ask is whether the witness has "sufficient skill, knowledge, or experience in that field or calling as to make it appear that his opinion or inference will probably aid the trier in his search for truth." McCormick on Evidence, supra at 33 (footnote omitted). See also In Re Involuntary Termination of Parental Rights, 449 Pa. 543, 297 A.2d 117 (1972) (expert witness must show special knowledge of very question upon which he promises to express opinion); Kravinsky v. Glover, 263 Pa. Super. 8, 396 A.2d 1349 (1979) (no error in qualifying witness as expert in psychology with special focus on driving phobia similar to plaintiff's); Erschen v. Pennsylvania Independent Oil Co., 259 Pa. Super. 474, 393 A.2d 924 (1978) (witness who had no formal instruction or on-the-job training in origin of gas explosions not qualified as expert, notwithstanding qualifications as fire marshall); Taylor v. Spencer Hospital, 222 Pa. Super. 17, 292 A.2d 449 (1972) (error not to allow nurse experienced in handling psychiatric patients to testify about standards for restraints); Rule 702 of the Rules of Evidence for United States Courts and Magistrates (1979), and Advisory Committee's Note to the rule; 2 Wigmore, Evidence § 555, § 1918 (Chadbourn rev. 1979).

Without doubt, in this case the subject in question called for expert testimony. The jury had to decide whether the presence of a radial tire on the right front wheel of a 1971 Plymouth Fury could, when the other three wheels were mounted with non-radial tires, in some way cause the driver to lose control when driving between 20 and 25 miles per hour. See R.R. at 586a. This was a subject "so distinctively related to some science, profession, business or occupation as to be beyond the ken of the average layman." McCormick, supra. Appellant identifies the "science, profession, business or occupation" as "physics, engineering, and vehicle road dynamics." Brief for Appellant at 39. This formulation, however, is too broad in that it may be understood as suggesting that only someone with a great deal of formal education would be qualified to testify on the causes of the accident. It is nevertheless correct to say that no opinion on the cause of the accident could aid the jury in its search for truth unless offered by someone with knowledge, however gained, whether by formal education, experience, or both, of those principles of physics and engineering pertinent to determining how a vehicle will behave in the conditions in which the accident in which appellee was hurt occurred.

One of the witnesses offered by appellees as an expert, William P. Kelly, had worked for several years as an automobile mechanic and then as a service manager for an automobile dealer. R.R. 241a, 244a, 246a, 247a. In these positions he sometimes road-tested automobiles for inspection purposes, id. at 248a, including the kind of automobile involved in this accident, a 1971 Plymouth Fury, id. at 251a. The fact that Mr. Kelly was an expert mechanic was irrelevant, however, for the subject in question did not involve mechanics but vehicle-road dynamics, and Mr. Kelly had no qualifications in that subject. He "didn't specialize in any mechanical things" whatever in high school, id. at 255a, and did not attend college, id. at 256a; he "never studied engineering of any kind," id., nor mathematics "[b]eyond trigonometry," id., nor had he "taken any courses that involved the design of tires," id., or that "dealt with the subject under whatever name of vehicle dynamics as affected by changes in the types or conditions of tires," id. at 256a, 264a. He had never worked for anyone "whose function was to test automobile performance as affected by change in the types or conditions of tires." Id. at 265a. The only technical literature Mr. Kelly had read on the subject of comparative effects on vehicle performance of radial and non-radial tires were the State Inspection Manual and service manuals put out by automobile makers such as Chrysler, the manufacturer of the Plymouth Fury. Id. at 268a-269a. He had "never read any technical articles published by a scientific or professional or engineering society or journal. . . ." Id. (Even if he had read such articles, it is apparent that neither his educational background nor his experience equipped him to understand them.)

On direct examination Mr. Kelly testified that he would not "pass an automobile for inspection that had on its front axle a radial and a bias belted tire [because of the] erratic steering conditions or erratic handling conditions that it can produce." R.R. at 285a-286a. In support of this opinion Mr. Kelly was permitted to read, without objection, from the Pennsylvania State Inspection Regulations on passing tires for inspection. The Regulations stated that "[r]adial ply tires shall not be used on the same axle with bias or belted-bias tires. . . . Mixing tires, size and type can produce dangerous and erratic steering performance including wander and fishtailing." Id. at 288a. Mr. Kelly was also permitted to read, without objection, from the Chrysler service manual for the 1971 Plymouth Fury. The manual stated that "[t]he use of radial tires is not recommended because of their harsh ride at low speeds and possible unfamiliar stability characteristics. . . . [Intermixing tires] will result in oversteer and could possibly cause spins on wet or icy roads. The safest policy is never intermix radial ply tires with bias belted tires or cross bias tires." Id. at 289a-291a. It is apparent, however, that anyone could have taken the stand to read these statements from the manuals. The manuals were evidence of what the Chrysler Company and Pennsylvania authorities thought about the possible effects of mixing tires, but they in no way qualified Mr. Kelly to say what he thought. One does not become an expert by reading manuals on a subject one is not capable of understanding.

Besides being an automobile mechanic, Mr. Kelly's other claim to being an expert on the effect of mixing tires was founded on his "personal experience in road testing 1971 Plymouth Furys. . . . that had a radial on one side of the front axle and a . . . bias belted tire on the other front axle . . . on wet ground." R.R. 333a-334a. Mr. Kelly said that

[t]he results, when we would road test this type of a car, on braking and turning would be sort of an erratic situation, where there was no predictability at any given time on braking or on turning as to what the car might do.

A test, however, is meaningless unless the person conducting the test knows what he is doing; he must understand the principles involved, and then design and conduct a test such that its results will implicate and depend upon those principles. For example, if the person does something to a liquid, and it turns blue, he must know enough of the principles involved to exclude the possibility that the liquid turned blue not because of what he did but for some other reason, such as the temperature of the room he was working in. Mr. Kelly designed and conducted no such test. The "road tests [he made] . . . would be to determine by driving a car whether there is some mechanical problem with it that cannot be seen by visual inspection." Id. at 252a. The tests were not designed or conducted with the idea of testing the effect of mixing tires, but rather were simply incidental to State inspections. Id. at 253a. Thus, Mr. Kelly did not know whether, or how, the tires on the cars he road-tested were mixed. He could not recall a single instance of "test[ing] a car that in fact had tires not all of the same category." Id. at 254a. Furthermore, when he did road test cars, he "drove them in a normal manner." Id. at 255a. He "did not put them through any unusual or steering or other maneuvers." Id. When asked, "Did you ever undertake as an organized program of an investigation or an inquiry a program by which you made a series of test runs on any given automobile under one set of runs containing all non-radial tires and under the other set of runs some combination of radial and non-radial," he answered: "Not under any laboratory or set up test conditions, no." Id. at 267a. His testimony continued:

Q. And you've never run any series of tests with the same automobile running it eight or ten or twelve times through a particular course with all non-radials and then eight or ten or twelve times through the same course with some combination of radial and non-radial? You've never done that, have you?

A. I could have in the course of doing business.

Q. But do you remember any specific instance where you did it for the purpose of making such a comparison?

A. Names and dates I can't give you. I have made those comparisons, though.

Q. You've never done it on a 1971 Fury, to your knowledge, have you?

A. Again I couldn't give you names and dates and places, but I believe that I have.

Given Mr. Kelly's testimony, it is not surprising that he did not produce any records of any sort. We are unable to escape the conclusion that he was permitted to testify as an expert because he said he was an expert. Nothing, however, justified his opinion of himself; he had never learned, either by education or by experience, the principles of vehicle dynamics. And nothing warrants the belief that what he said could have aided the jury in its search for truth. Neither he nor anyone else knew either when or how many tests he had made; or what mixture of tires was involved in the tests; or what was done in the course of the tests to determine whether the mixture had any effect on the handling of the automobile; or what that effect was. It was therefore clear error and an abuse of discretion to permit Mr. Kelly to testify: he did not have "sufficient skill, knowledge, or experience . . . to . . . aid the trier in his search for truth." McCormick, supra.

The cases cited by appellees are not contrary to but rather support this conclusion. In Kuisis v. Baldwin-Lima-Hamilton Corp., 457 Pa. 321, 319 A.2d 914 (1974), a products liability case, a person was struck by a load of steel pipe that fell when a brake-locking mechanism on a crane malfunctioned. The trial court excluded, as unqualified, expert testimony of a safety engineer as to the safety of the design of the brake-locking mechanism. Reversing, the Supreme Court held that the safety engineer was qualified as an expert. Said the Court:

Appellee points to the fact that Barbe was by his own admission a "safety engineer" rather than a mechanical engineer, and that he was not registered to practice mechanical engineering. From this it is argued that Barbe was not qualified to express an opinion on matters of design. But an engineer need not be registered as such in order to testify as an expert if his education and experience so qualify him. Lance v. Luzerne County Manufacturers Association, 366 Pa. 398, 77 A.2d 386 (1951). The standard of qualification is a liberal one: "If a witness `has any reasonable pretension to specialized knowledge on the subject under investigation he may testify, and the weight to be given to his [testimony] is for the jury: [citations omitted]. . . .' McCullough v. Holland Furnace Co., 293 Pa. 45, 49, 141 A. 631, 632 [(1928)]." Id., 457 Pa. at 338-39, 319 A.2d at 924 (footnotes omitted).

Thus, the "subject under investigation" was whether a crane was safe, and the witness had "specialized knowledge" of the subject not only from experience in inspecting cranes, but also because he had been trained as an engineer and therefore knew, when he inspected a crane, what he was looking at. Rutter v. Northeastern Beaver County School District, 496 Pa. 590, 437 A.2d 1198 (1981) (plurality opinion), presented a similar situation. There, a high school student was injured when playing "jungle football" without protective equipment and while under the supervision of two football coaches. In an action against the School District alleging negligence of the football coaches, the trial court refused to allow a former football coach to testify as an expert. Reversing, the Supreme Court said that "[i]t seems clear than an experienced former coach may have knowledge of the customs and safety standards utilized by coaches of high school teams and of the rules imposed [by authorities] to insure minimum safety. . . ." Id., 496 Pa. at 598, 437 A.2d at 1202. Thus, Kuisis and Rutter show both when a witness should be held qualified, and when he should be held not qualified. In Kuisis the witness was qualified by reason of both experience and education in the subject under investigation; in Rutter, by experience. If — like Mr. Kelly and Mr. Pruyn — the witness has neither experience nor education in the subject under investigation, he should be found not qualified.

The conclusion that the trial court should not have admitted the testimony of Mr. Kelly and Mr. Pruyn is further supported by other cases in which the scope of a witness's experience and education was examined as it bore upon the subject under investigation.

Sometimes it may appear that the scope of the witness's experience and education embraces the subject in question in a logical, or fundamental, sense. In such a case, the witness is qualified to testify even though he has no particularized knowledge of the subject as such; for he will be able to reason from the knowledge he does have. Thus in Whistler Sportswear, Inc. v. Rullo, 289 Pa. Super. 230, 433 A.2d 40 (1981), a civil engineer was held qualified to testify on the causes of the collapse of a roof, even though his area of expertise was not specifically roof design. He was qualified because the "subject under investigation" did not require "knowledge of roofing per se, but rather . . . knowledge of engineering principles of stress and resiliency and an ability to interpret the evidence left in the aftermath of the physical collapse." Id., 289 Pa.Superior Ct. at 238-39, 433 A.2d at 44. And see Jones v. Treegoob, 212 Pa. Super. 482, 243 A.2d 161 (1968), rev'd. on other grounds, 433 Pa. 225, 249 A.2d 352 (1969) (plaintiff injured during windstorm when store window blew out; expert who has B.S. in industrial engineering, twenty-two years experience as safety engineer, and considerable knowledge of wind pressure and its effects, held qualified). See also Dorsey v. Yoder Co., 331 F. Supp. 753 (1971) (witness who lacked familiarity with particular metal slitting machine but who as an engineer, in industry, armed forces, and at University of Pennsylvania where he was professor and Chairman of Graduate Division of Civil Engineering, had designed and worked with many machines that operated by same principles as the slitter, held qualified to testify as to hazards created by design of slitter).

Other times it may appear that the scope of the witness's experience and education may embrace the subject in question in a general way, but the subject may be so specialized that even so, the witness will not be qualified to testify. Thus, every doctor has a general knowledge of the human body. But an ophthalmologist, for example, is not qualified to testify concerning the causes and treatment of heart disease. See Arnold v. Loose, 352 F.2d 959 (3rd Cir. 1965) (striking testimony of orthopedic surgeon that in his opinion defendant's decedent had lapsed into diabetic coma, which was cause of automobile-truck collision, held not beyond trial court's discretion where witness admitted he had never read any text on diabetes or diabetic comas, did not know who was leading authority on diabetes or which was leading treatise, and revealed no other special knowledge in field of diabetes); Hunt v. Bradshaw, 251 F.2d 103 (4th Cir. 1958) (radiologist not qualified to testify as to proper surgical procedure in chest operation); Wesley v. State, 32 Ala. App. 383, 26 So.2d 413 (1946) (toxicologist not qualified to testify wound inflicted with screwdriver or similar instrument); Harris v. Campbell, 2 Ariz. App. 351, 409 P.2d 67 (1965) (discretionary exclusion of general practitioner in malpractice action against gynecologist who performed vaginal hysterectomy); Huffman v. Lindquist, 37 Cal.2d 465, 234 P.2d 34 (1951) (autopsy surgeon not qualified as to treatment for brain injury); Moore v. Belt, 34 Cal.2d 525, 212 P.2d 509 (1950) (autopsy surgeon not qualified as to existing standards in practice or urology); Pearce v. Linde, 113 Cal.App.2d 627, 248 P.2d 506 (1952) (specialist in internal medicine not qualified as to orthopedics); Dolan v. Galluzzo, 77 Ill.2d 279, 32 Ill.Dec. 900, 396 N.E.2d 13 (1979) (physician unlicensed in podiatry not qualified to testify in malpractice action against podiatrist); Swanson v. Chatterton, 281 Minn. 129, 160 N.W.2d 662 (1968) (internist not qualified as to orthopedic surgery); State v. Askin, 90 Mont. 394, 3 P.2d 654 (1931) (general practitioner not qualified to testify as to brain injury); Whitehurst v. Boehm, 41 N.C. App. 670, 255 S.E.2d 761 (1979) (orthopedic surgeon unfamiliar with practice of podiatry not qualified to testify as to standard of care required of podiatrist); Capan v. Divine Providence Hospital, 270 Pa. Super. 127, 410 A.2d 1282 (1980) (anesthesiologist not qualified as to autopsy report); Cf. Kosberg v. Washington Hospital Center, 394 F.2d 947 (D.C. Cir. 1968) (internist held qualified to testify as to effects of electroshock therapy, even though not a psychiatrist or neurologist); Baerman v. Reisinger, 363 F.2d 309 (D.C. Cir. 1966) (general practitioner qualified to testify that cardiologist was negligent in failing to diagnose hypothyroidism in patient over six year period of treatment); State v. Staples, 120 N.H. 278, 415 A.2d 320 (1980) (physician who had been general practitioner for over twenty years, had examined several rape victims, and had recently attended course covering psychological problems of rape victims could give opinion that victim's memory loss was due to mental block rather than intoxication).

Here, Mr. Kelly and Mr. Pruyn had no education on the subject of the effect of mixing radial and non-radial tires. While the scope of the experience they had had embraced the subject of tires in a general way, it did not embrace the specialized subject of the effect of mixing tires, and nothing in their experience, or in such education as they had had, enabled them to reason about what that effect would be.

The trial court's instruction to the jury on appellant's strict liability for a defect in the tire

Having concluded that a new trial is required because of the trial court's error in admitting the testimony of appellees' experts on the cause of the accident, we might refrain from ruling on the other issues that have been argued to us. However, one of those issues, in particular, the issue of how the trial court should instruct the jury on appellant's strict liability, may very well arise at the new trial. Accordingly, some discussion of it is warranted.

The first requirement that appellees will have to meet at the new trial will be to offer qualified expert testimony that the tire mixture was a proximate cause of the accident. If no such testimony is offered, the trial court will have no occasion to instruct the jury regarding appellant's strict liability, for in the absence of proof of causation, appellant's radial tire could not be found defective because it was not embossed with a warning not to mix it with non-radial tires. Suppose, for example, that A sells a bottled drink to B, and B pours the drink into a glass filled with ice-cubes, becomes ill, and sues A for selling a product defective because the bottle's label failed to include a warning not to mix the drink with ice-cubes. B will not make out a case for the jury unless he proves that mixing ice-cubes with the drink did in fact make him ill. If the mixing did not make B ill, there was no need to warn against the mixing, and therefore no defect in the bottle's label. Assuming, however, that appellees are able to present competent evidence of proximate cause at a new trial, that is, that they do offer the testimony of qualified experts, the issue will then arise whether appellant's radial tire was defective because it was not embossed with a warning not to mix it with non-radial tires. In that event, what will be the trial court's responsibility?

-a-

Appellant, relying on Azzarello v. Black Brothers Co., Inc., 480 Pa. 547, 391 A.2d 1020 (1978), argues that it will be the trial court's responsibility to rule, as a matter of law, that the radial tire was not defective. Accordingly, appellant concludes, instead of ordering a new trial, we should enter judgment n.o.v. in appellant's favor. We agree with appellant to this extent: under Azzarello, the trial court will have to rule whether, as a matter of law, the jury could find the radial tire defective. However, we find no merit in appellant's argument that when the trial court makes that ruling, it must conclude that the tire was not defective.

In Azzarello the question was whether in a products liability case the jury should be instructed to determine whether the product in question was "unreasonably dangerous." In considering this question, the Court recognized that before it could be answered, the respective functions of the trial court and the jury had to be defined. Said the Court:

It must be understood that the words, "unreasonably dangerous" have no independent significance and merely represent a label to be used where it is determined that the risk of loss should be placed upon the supplier. It is for this reason that a mere change in terminology does not supply the answer to the basic question as to what instructions should be given to the jury. The answer to the proceeding [ sic] question rests upon the more fundamental question whether the determination as to the risk of loss is a decision to be made by the finder of fact or by the court. While a lay finder of fact is obviously competent in resolving a dispute as to the condition of a product, an entirely different question is presented where a decision as to whether that condition justifies placing liability upon the supplier must be made. 480 Pa. at 556, 391 A.2d at 1025 (footnotes omitted).

The Court then went on to hold that "[i]t is a judicial function to decide whether, under the plaintiff's averment of the facts, recovery would be justified; and only after this judicial determination is made is the cause submitted to the jury to determine whether the facts of the case support the averments of the complaint." Id., 480 Pa. at 558, 391 A.2d at 1026. The Court illustrated its holding by examples of the sort of questions a court is to ask:

Should an ill-conceived design which exposes the user to the risk of harm entitle one injured by the product to recover? Should adequate warnings of the dangerous propensities of an article insulate one who suffers injuries from those propensities? When does the utility of a product outweigh the unavoidable danger it may pose? Id.

These questions, the Court said, are not for the jury but for the court, for they are "questions of law and their resolution depends upon social policy." Id.

In Azzarello the Court did not do what appellant argues the trial court should have done here — the Court did not explicitly formulate or resolve the question whether on the particular facts before it recovery for strict liability would be proper. However, the Court did affirm the trial court's order granting a new trial, and in doing so, it decided what jury instruction should be used on retrial. It is evident, therefore, that the Court regarded the case before it as one in which, on proper jury instruction, and on proof of the plaintiff's averment of the facts, recovery would be justified. The same may be said of this case.

In describing the respective functions of court and jury in a products liability case as compared with some other sorts of cases, Dean Wade has written:

In an action for negligence it is normally the function of the jury to determine whether the defendant was negligent, subject, of course, to the authority of the judge to direct a verdict for the defendant, if he finds that the jury could not reasonably find for the plaintiff. On the other hand, in an action based on strict liability of the Rylands type, for an abnormally dangerous activity, the determination as to whether strict liability will be imposed for the activity is held to be one for the judge, not the jury — for the reason that the decision involves issues of general social policy. In the products cases the courts seem not to have approached the problem in this fashion. Instead, they seem to have assumed that strict products liability is like negligence in this respect, so that a plaintiff, in order to recover, must convince the jury that the product was "defective" or "unreasonably dangerous" or "not duly safe," or whatever test is used. This generally works quite satisfactorily when the question is whether the product was unsafe because of an error in the manufacturing process so that it was not in the condition in which it was intended to be. The issue then seems more factual, of the kind the jury is accustomed to handling. The difficulty comes when it is not just the single article which is to be classed as unsafe because something went wrong in the making of it, but a whole group or class or type which may be unsafe because of the nature of the design. It is here that the policy issues become very important . . . It is here the court — whether trial or appellate — does consider these issues in deciding whether to submit the case to the jury. If a plaintiff sues the manufacturer of a butcher knife because he cut his finger, on the sole ground that the knife was so sharp that it was likely to cut human flesh, the court would probably take the case out of the hands of the jury and not give it the opportunity to find that the knife was unsafe. Similarly with an aspirin manufacturer, when an ordinary tablet stuck to the lining of the plaintiff's stomach and caused a hemorrhage, or the manufacturer of the Pasteur treatment for rabies, when there were untoward reactions. The problem in these cases is likely to be called one of law and decided by the court. Court control of jury action is more extensive here than in the ordinary negligence action. And yet, of course, if the court decides that it would be reasonable to allow the jury to find for the plaintiff, the issue of lack of due safety will be submitted to the jury even in these cases. Wade, On the Nature of Strict Tort Liability for Products, 44 Miss.L.J. 825, 838 (1973) (footnotes omitted).

In addition to there being certain products, there are also certain risks that as a matter of law, or social policy, cannot support imposition of strict liability. In Lobianco v. Property Protection, Inc., 292 Pa. Super. 346, 437 A.2d 417 (1981), the plaintiff sued a burglar alarm company when her alarm system failed to operate properly and jewelry was taken from her home. The trial court held that the plaintiff could not recover in strict liability. This court affirmed, MONTGOMERY, J., joined by HESTER, J., dissenting. Two judges, BROSKY, J., joined by CAVANAUGH, J., affirmed on the ground that "the injury suffered by [the plaintiff; the loss of her jewelry] is not of the type for which strict liability ought to be imposed." Id., 292 Pa. Super. at 363, 437 A.2d at 426. Three judges, SPAETH, J., joined by PRICE, J., with CERCONE, P.J., stating his "agree[ment] with the analysis," affirmed on the ground that under Azzarello the case should not go to the jury because "`[a]s a matter of social adjustment' [quoting Prosser, Law of Torts 495 [4th ed. 1971], the responsibility for protecting against the loss of the jewelry should, so far as the imposition of strict liability is concerned, be appellant's [the plaintiff's], not appellee's [the burglar alarm company's]." 292 Pa.Super. at 362, 437 A.2d at 425. This conclusion followed from the fact that to impose strict liability "would represent a less, not more, equitable allocation of the risk." Id., 292 Pa. Super. at 360, 437 A.2d at 424. As between homeowner and manufacturer, the homeowner was in the better position to protect against the loss, for the homeowner knew the contents of the home sought to be protected by the burglar alarm and so could insure against their loss, whereas the manufacturer could not know the contents and therefore could protect itself against their loss only by increasing its price for its burglar alarms, which as a matter of social adjustment would be undesirable because it would result in "[t]hose of modest means . . . subsidizing the rich." Id., 292 Pa. Super. at 361, 437 A.2d at 425. Azzarello and Lobianco are not unique in holding that there are certain products or risks against which the law of products liability is not meant to protect the user, and as to which the trial court should therefore rule as a matter of law that strict liability does not apply. See, e.g., Daberko v. Heil, Co., 681 F.2d 445 (5th Cir. 1982) (Tex. law) (manufacturer not strictly liable for injuries as a matter of law when truck redesigned after sale); Dubin v. Michael Reese Hospital, 83 Ill.2d 277, 47 Ill.Dec. 345, 415 N.E.2d 350 (1981) (strict liability not applicable in action to recover from hospital for overdose of radiation); Feldman v. Lederle Laboratories, 97 N.J. 429, 479 A.2d 374 (1984) (collecting New Jersey cases: hospitals not liable for injuries from hepatitis-infected blood transfusion; dentist not liable for injuries when hypodermic needle breaks in the jaw); O'Brien v. Muskin Corp., 94 N.J. 169, 186, 463 A.2d 298, 307 (1983) ("If the minds of reasonable men could not differ on whether the risks posed by a product outweigh its utility or vice versa, then the court could make the appropriate determination as a matter of law"); Atkins v. Arlans Department Store of Norman, Inc., 522 P.2d 1020 (Okla. 1974) ("lawndart" not unreasonably dangerous as a matter of law since dangers readily apparent); Silverhart v. Mount Zion Hospital, 20 Cal.App.3d 1022, 98 Cal.Rptr. 187 (1971) (doctrine of strict liability not applicable to injuries sustained when hospital equipment breaks during surgery).

Courts and commentators have identified various factors that a court should consider when making the social policy decision required by Azzarello, and made in Lobianco. Sometimes, no doubt, it will be difficult for a court to decide whether as a matter of social policy a jury should be permitted to impose strict liability. However, where inadequate warnings are alleged, the social policy decision is relatively simple. As has been said, "In the case of an inadequate warning, . . . imposing the requirements of a proper warning will seldom detract from the utility of the product." Freund v. Cellofilm Properties, Inc., 87 N.J. 229, 238 n. 1, 432 A.2d 925, 930 n. 1 (1981). At the same time, the cost of adding a warning, or of making an inadequate warning adequate, will at least in most cases be outweighed by the risk of harm if there is no adequate warning.

Accordingly, the trial court was entitled to rule, at the trial that has been held, and will again be entitled to rule, at the new trial, that as a matter of social policy this is a case in which a jury, on proper instruction and proof, can impose strict liability.

-b-

With the foregoing conclusion on social policy reached, it is in order to consider how the trial court should instruct the jury to decide whether appellant may be found strictly liable because the radial tire was not embossed with a warning not to mix it with non-radial tires. Appellant argues that in instructing the jury the trial court "failed to develop [the] careful distinction between insurer and guarantor" developed in Azzarello. Brief for Appellant at 32. While we agree that the trial court should have explained the difference between an insurer and a guarantor, and believe that its instruction could have been fuller — a matter we shall discuss — nevertheless we find the instruction substantially in accordance with Azzarello and therefore reject appellant's argument that the instruction as given requires a new trial. We are also not persuaded by Judge WIEAND's argument that a new trial is in order not only, as we hold, because of the trial court's error in ruling that appellees' witnesses were qualified to express an opinion on causation, but also because the instruction was not in negligence terms. Rejecting Azzarello in favor of what it he characterizes as "the better approach," Judge WIEAND maintains that the trial court should have instructed the jury that "[t]here can be no liability for a failure to warn without a showing that the defendant supplier failed to give reasonable warning of risks or hazards of which he knew or should have known." Concurring and Dissenting at 437. It is our view that we are not thus free to reject Azzarello, but further, that if we were, we should not, for we believe that the distinction, so emphasized by Azzarello, between products liability cases and negligence cases should be maintained.

We may start our discussion of this point by noting that three sorts of cases should be distinguished.

In what might be called the "traditional" strict liability cases, in which the defendant has engaged in an ultrahazardous activity that cannot be made safe, the defendant is held to be an "insurer" as to injuries that the activity has caused. The point is not that there is anything wrong with the activity, but rather that because of its dangerousness the defendant may be held strictly liable for the injuries the activity caused. See Prosser, supra at 505-16; Restatement (Second) of Torts §§ 519, 520 (1965).

In a products liability case, such as this case, in contrast, the defendant — the manufacturer or supplier of the product — is not an "insurer" but "effectively the guarantor of his product's safety." Salvador v. Atlantic Boiler Co., 457 Pa. 24, 32, 319 A.2d 903, 907 (1974). The plaintiff must therefore prove that the product that caused his injury was defective, or "unsafe for its intended use." Id. If the defendant were the insurer of its product, liability would follow upon a finding that the plaintiff was injured while using the product: the fact that injury occurred would lead to the conclusion that the product was unsafe in some way. But the jury is not to find liability when the product is unsafe in some way; liability may be imposed only on proof that the product lacked an element necessary to make it safe for its intended use. As the Supreme Court said in Berkebile v. Brantly Helicopter Company, 462 Pa. 83, 337 A.2d 893 (1975):

Thus, the plaintiff cannot recover if he proves injury from a product absent proof of defect, such as developing diabetic shock from eating sugar or becoming intoxicated from drinking whiskey. Neither can plaintiff recover by proving a defect in the product absent proof of causation, as where the plaintiff sustains eye injury while not wearing defective safety glasses.

Finally, in a negligence case the plaintiff must prove, not only that the product was defective and that the defect caused his injury, but in addition, that in manufacturing or supplying the product the defendant failed to exercise due care. The defendant is liable neither as an insurer nor guarantor but rather only for failing to act as a reasonable man would have acted. See Morena v. South Hills Health System, 501 Pa. 634, 462 A.2d 680 (1983); Macina v. McAdams, 280 Pa. Super. 115, 421 A.2d 432 (1980).

The adoption of these distinctions in the law of Pennsylvania may be dated from the Supreme Court's decision in Webb v. Zern, 422 Pa. 424, 220 A.2d 853 (1966). There the Court adopted Section 402A of the Restatement (Second) of Torts, which provides:

(1) One who sells any product in a defective condition unreasonably dangerous to the user or consumer or to his property is subject to liability for physical harm thereby caused to the ultimate user or consumer, or to his property, if

(a) the seller is engaged in the business of selling such a product, and

(b) it is expected to and does reach the user or consumer without substantial change in the condition in which it is sold.

(2) The rule stated in Subsection (1) applies although

(a) the seller has exercised all possible care in the preparation and sale of his product, and

(b) the user or consumer has not bought the product from or entered into any contractual relation with the seller.

Restatement (Second) of Torts § 402A (1965).

In Salvador v. Atlantic Steel Boiler Co., supra, the Court reaffirmed its adoption of Section 402A, stating:

Today, . . . a manufacturer by virtue of section 402A is effectively the guarantor of his products' safety. See Webb v. Zern, supra; Kassab v. Central Soya, supra [ 432 Pa. 217, 246 A.2d 848 (1968)]. Our courts have determined that a manufacturer by marketing and advertising his product impliedly represents that it is safe for its intended use. We have decided that no current societal interest is served by permitting the manufacturer to place a defective article in the stream of commerce and then to avoid responsibility for damages caused by the defect. He may not preclude an injured plaintiff's recovery by forcing him to prove negligence in the manufacturing process. Webb v. Zern.

In Berkebile v. Brantly Helicopter Corporation, supra, the Supreme Court was required to distinguish between a case in which the defendant had been sued — as appellant here has been sued — in strict liability for having failed to provide adequate warnings, and a negligence case. The trial court had failed to make the distinction, defining "defective condition" for the jury in negligence terms. Reversing, the Chief Justice, joined by Justice (now Chief Justice) NIX, stated:

The seller is responsible for injury caused by his defective product even if he "has exercised all possible care in the preparation and sale of his product." Restatement (Second) of Torts, § 402A(2)(a). As we declared in Salvador, supra, 457 Pa. at 32, 319 A.2d at 907, the seller "may not preclude an injured plaintiff's recovery by forcing him to prove negligence in the manufacturing process." What the seller is not permitted to do directly, we will not allow him to do indirectly by injecting negligence concepts into strict liability theory.

Examining Section 402A, the Chief Justice noted that it appears to "impose a contradictory burden of proof in that [the product must be both defective and] also be `unreasonably dangerous.'" Id., 462 Pa. at 95, 337 A.2d at 899. However, citing Dean Prosser, the Chief Justice explained that the term "unreasonably dangerous" was included in Section 402A "to foreclose any argument that the seller of a product with inherent possibilities for harm would become `automatically responsible for all the harm that such things do in the world.'" Id. (quoting Prosser, Strict Liability to the Consumer in California, 18 Hast.L.J. 9, 23 (1966)). In other words, the term "unreasonably dangerous", as used in Section 402A, does not import negligence principles into a strict liability case, but rather prevents a supplier sued in strict liability from being held liable as an insurer instead of as a guarantor. Following the California Supreme Court's analysis in Cronin v. J.B.E. Olson Corp., 8 Cal.3d 121, 104 Cal.Rptr. 433, 501 P.2d 1153 (1972), the Chief Justice held that

the "reasonable man" standard has no place in a strict liability case . . . To charge the jury or permit argument concerning the reasonableness of a consumer's or seller's actions and knowledge, even if merely to define "defective condition" undermines the policy considerations that have led us to hold in Salvador that the manufacturer is effectively the guaranter [ sic] of his product's safety. 462 Pa. at 96-97, 337 A.2d at 900 (JONES, C.J.)

The precedential authority of Berkebile was for a time uncertain, for, as noted, only Justice NIX had joined in the Chief Justice's opinion. This uncertainty, however, was removed by the Supreme Court's decision in Azzarello. There the Court, in a unanimous opinion by Justice NIX, cited with approval and adopted the Chief Justice's opinion in Berkebile. In doing so, the Court went one step further than it had in Berkebile. In Berkebile the Chief Justice had explained that the term "unreasonably dangerous", as used in Section 402A, does not import negligence principles into a strict liability case. In Azzarello the Court further decided, as we have already discussed, that what the term "unreasonably dangerous" does do is to impose on the trial court the responsibility of deciding, as a matter of law and by resolving considerations of "social policy", whether "the risk of loss should be placed upon the supplier." 480 Pa. at 556, 391 A.2d at 1025. Emphasizing its previous decisions that the supplier's liability is that of a guarantor, not an insurer, the Court went on to outline a jury instruction for use "in products liability cases in this Commonwealth," id., 480 Pa. at 549, 391 A.2d at 1022, stating that "the jury may find a defect where the product left the supplier's control lacking any element necessary to make it safe for its intended use or possessing any feature that renders it unsafe for the intended use." Id., 480 Pa. at 559, 391 A.2d at 1027 (footnote omitted).

It may well be that as an intermediate appellate court we should end our discussion at this point: the line of authority established by Webb, Salvador, Berkebile, and Azzarello is clear and therefore binding on us. Nevertheless, out of deference to Judge WIEAND's position that the line of authority is not clear, we shall explain why we regard it as sound and deserving to be followed, quite apart from its precedential authority.

It may be noted first that there are difficulties in thinking of an inadequate warnings case as a products liability case. Products liability cases fall into two categories: manufacturing defect cases and defective design cases. In a manufacturing defect case, the question whether the product is defective is relatively simple. Since the allegation is that something went awry in the manufacturing process, so that, for example, the product lacked a component it should have had, the finder of fact need only compare the product that caused the injury with other products that were manufactured according to specifications. In a defective design case, however, the question is whether the product should have been designed more safely. While an inadequate warnings case has been characterized as a kind of defective design case, Incollingo v. Ewing, 444 Pa. 263, 282 A.2d 206 (1971), the term "defective" does not easily fit when applied to warnings. For in a warnings case it is not alleged that there was anything wrong with the product's design as such. Rather, a "defect" is supposed to exist because the user was not adequately instructed on how to use the product as the product was designed.

The way to overcome the possible misunderstanding of the term "defective" in an inadequate warnings case is not, as Judge WIEAND would do, to instruct the jury in negligence terms, but to explain to the jury that it is to consider whether the product was safe in the absence of warnings or in light of the warnings that were given. See Freund v. Cellofilm Properties, Inc., 87 N.J. 229, 243, 432 A.2d 925, 932 (1981) ("adequate warning is one that includes the directions, communications, and information essential to make the use of a product safe."); cf. Little v. P.P.G. Industries, Inc., 92 Wn.2d 118, 122, 594 P.2d 911, 914 (1979) ("The question is: Was the warning sufficient to catch the attention of persons who could be expected to use the product, to apprise them of its dangers and to advise them of the measures to take to avoid those dangers."); Cavers v. Cushman Motor Sales, 95 Cal.App.3d 338, 342, 157 Cal.Rptr. 142, 144 (1979) (approving jury instruction: "An article otherwise appropriately made and maintained is defective . . . if the manufacturer . . . fails to adequately warn of dangerous propensities of such article which in the absence of an adequate warning renders the article substantially dangerous."). Thus, there is no necessity, such as Judge WIEAND suggests, to depart from strict liability principles when inadequate warnings are alleged. The emphasis need only be altered slightly to focus not so much on the product itself as on the safety of the product in light of the warnings that the seller gave, or failed to give.

Judge WIEAND quotes Keeton, that "[a]lthough this ground of recovery [ i.e., recovery in strict liability for failure to give adequate warnings] is sometimes referred to as strict liability, it is really nothing more than a ground of negligence liability better described as the sale of a product in a defective condition." Concurring and Dissenting at 435, quoting Keeton, The Meaning of Defect in Products Liability Law — a Review of Basic Principles, 45 Mo.L.Rev. 579, 586-87 (1980). From this we understand that in Judge WIEAND's opinion, it does not matter in a case such as this one whether negligence or strict liability principles are applied. However, in some cases it will matter, for the proofs will differ according to which principles are applied, and the plaintiff's burden will be greater if negligence rather than strict liability principles are applied. In Cronin v. J.B.E. Olson Corp., 8 Cal.3d 121, 501 P.2d 1153, 104 Cal.Rptr. 433 (1972), which our Supreme Court cited with approval in Azzarello, it was noted:

In fact, it has been observed that the Restatement formulation of strict liability in practice rarely leads to a different conclusion than would have been reached under laws of negligence . . . Yet the very purpose of our pioneering efforts in this field was to relieve the plaintiff from problems of proof inherent in pursuing negligence and warranty . . . remedies, and thereby "to insure that the costs of injuries resulting from defective products are borne by the manufacturers . . ."

8 Cal.3d at 133, 501 P.2d at 1162, 104 Cal.Rptr. at 442, quoting Greenman v. Yuba Power Products Inc., 59 Cal.2d 57, 63, 27 Cal.Rptr. 697, 701, 377 P.2d 897, 901 (1963) (citation omitted).

And the Washington Supreme Court has said that

the objective of the rule of strict liability with respect to dangerous products is defeated if a plaintiff is required to prove that the defendant was negligent, or the latter is allowed to defend upon the ground that he was free of negligence. It is the adequacy of the warning which is given, or the necessity of such a warning which must command the jury's attention, not the defendant's conduct.

Little v. P.P.G. Industries, Inc., supra, 92 Wn.2d at 121, 594 P.2d at 914.

It was on the basis of the distinction between products liability theory, under which the defendant as manufacturer or supplier is guarantor of its product, and negligence theory, under which the defendant is held to the standard of a reasonable man, that the Court in Azzarello rejected the use of the negligence term "unreasonably dangerous" in products liability cases. We see no reason to depart from that reasoning now. To the contrary, it is sound, both as a matter of legal theory and in its practical consequences. Thus, quite independently of the precedential authority of Azzarello, we hold, as the Supreme Court did there, that in a strict liability case, principles of negligence have no place.

This conclusion reached, we may consider how the trial court should on retrial instruct the jury so as to comply with Azzarello.

In Azzarello, after stating that "a jury may find a defect where the product left the supplier's control lacking any element necessary to make it safe for its intended use or possessing any feature that renders it unsafe for its intended use," 480 Pa. at 559, 391 A.2d at 1027 (footnote omitted), the Court stated that an "adequate charge" is the one contained in Pennsylvania Standard Jury Instruction 8.02 (Civil) Subcommittee Draft (June 6, 1976). It was this charge that the trial court followed, and as we have indicated, while we find no error in its instruction, at least, none rising to the level of reversible error, the instruction could, and therefore on retrial should, be improved.

Some commentators have criticized the instruction approved in Azzarello, in a manner supportive of appellant's view that the trial court's instruction here had the effect of making appellant an insurer rather than guarantor of its product:

Is there any product that cannot be made safer in some way? This instruction [ Azzarello's] calls forth fantastic cartoon images of products, both simple and complex, laden with fail-safe mechanism atop fail-safe mechanism. Birnbaum, Defect: From Negligence [to Warranty] to Strict Liability to Negligence, 33 Vand.L.Rev. 593, 637 (1980).

See also Henderson, Renewed Judicial Controversy Over Defective Product Design: Toward the Preservation of an Emerging Consensus, 63 Minn.L.Rev. 773, 800-01 (1980).

This criticism, however, seems to us to have merit only in the abstract. In the abstract, there might be something to its claim that the jury, instructed that "the product . . . must be provided with every element necessary to make it safe for its intended use," will conjure up "fantastic images" of mechanisms that might by some stretch of imagination have been added to the product to make it saf er. But as a practical matter, we believe, the jury will make its decision in the context of the testimony it has heard and the parties' respective contentions. Whatever force the criticism of the Azzarello instruction may have in the abstract can be met by adapting the instruction to the facts of the particular case. Here, such an adaptation might be accomplished somewhat as follows.

The instruction approved by the Court in Azzarello starts with the statement that "[t]he supplier of a product is the guarantor of its safety." This might be expanded by adding a brief explanation of the distinction between a "guarantor" and an "insurer" — a distinction critical to the law of products liability, as we have discussed. Thus the jury might be told: "A guarantor is not an insurer. An insurer of a product is responsible if the user [sometimes "consumer" will be the better word, depending upon the sort of product in question] of the product is injured by the product in some way. But a guarantor of a product is responsible only if the user of the product is injured as a result of a defect in the product."

Next, the jury might be told, "You must therefore decide whether, when the product left the supplier's control, there was a defect in it," the instruction then continuing as required by Azzarello, with the language approved in Azzarello being adapted to this case as a warnings case. For example: "A product otherwise properly made is defective if the supplier does not adequately warn of the dangers of the product. If you find that when the product left the supplier's control, it lacked the warnings necessary to make it safe for its intended use, then the product was defective, and the supplier is liable for all harm caused by the defect."

Finally, these abstract principles should be made concrete by specific reference to the testimony. Thus in this case the jury might be told: "This means that you must decide whether, as the plaintiffs contend, the radial tire supplied by the defendant was unsafe for its intended use because it lacked a warning that it should not be mixed with non-radial tires. If you find that because it lacked such a warning, the tire was unsafe for its intended use, then you should find it defective."

Of course, these suggestions, especially the last, are not offered with the intent that they be repeated verbatim. Exactly what words should be used must depend upon the evidence and the parties' respective contentions as developed at retrial.

Nor should it be assumed that under such an instruction the supplier will always, or necessarily, be found liable — which is perhaps the fear underlying the concurring and dissenting opinion. It may be that either the product or the risk involved is such that as a matter of social policy, strict liability should not be imposed. See, e.g., the plurality opinion in Lobianco v. Property Protection, Inc., supra. Or it may be that the product did not lack what the plaintiff claims it should have had. Or it may be that although the product did lack what the plaintiff claims it should have had, still, that fact was not the cause of the plaintiff's injury. It is this last possibility that seems to us to lie at the heart of this case. As we have said, we have no difficulty in deciding that as a matter of social policy, this is a case in which a jury, on proper instruction, could impose strict liability. And there is no question that appellant's radial tire lacked the warning appellees claim it should have had. But the question remains whether appellees will be able to prove at retrial what they did not prove at trial: that the mixture of radial and non-radial tires caused the accident. As already discussed, there will be no need for the jury on retrial to consider whether the tire, unaccompanied by further warning than was given, was defective unless it is first proved through qualified expert testimony that the mixture of radial and non-radial tires caused the accident. In instructing the jury on retrial, therefore, the trial court should make it clear that only if the jury answers this question affirmatively need it consider the further question whether the tire was defective, that is, whether the radial tire that was placed on the car before the accident lacked the directions and information necessary to make it safe for its intended use.

The judgment is vacated and the case remanded for new trial generally, the trial to be conducted consistent with this opinion.

Jurisdiction is relinquished.

WIEAND, J., files a concurring and dissenting opinion in which ROWLEY, J., joined.