Sandford v. Chevrolet Division of General Motors 292 Or. 590, 642 P.2d 624

*592LINDE, J.

Plaintiff suffered extensive burns when a pickup truck that she was driving overturned and caught fire. She brought an action for damages against a number of defendants in which she alleged, among other things, that the accident was caused by a defective tire manufactured by defendant Uniroyal, Inc., and mounted on the truck by The Tire Factory. The defendants filed answers alleging that plaintiffs own negligence caused her injuries. Over plaintiffs objections, the trial court submitted these allegations to the jury with instructions to reduce or deny plaintiffs damage claim if plaintiffs injuries resulted in part or predominantly from her own fault. The jury found defendants Uniroyal, Inc. and The Tire Factory at fault to the extent of 55 percent and plaintiff to the extent of 45 percent and awarded plaintiff a corresponding fraction of her total damages.

The Court of Appeals reversed. It held that a recovery of damages for injuries caused by a defective product is not barred or reduced by plaintiffs ordinary contributory negligence under Oregon’s proportionate fault statute, ORS 18.470. The court also remanded the case for a new trial because the trial court denied a defense motion to poll the jury. 52 Or App 579, 629 P2d 407 (1981). We allowed review in this case and in Wilson v. B.F. Goodrich, 52 Or App 139, 627 P2d 1280 (1981), also decided today, primarily to decide whether and how the proportionate fault law applies when a dangerously defective product and a plaintiffs negligence together resulted in the plaintiffs injuries. Because it appeared that the question how negligence could be matched against products liability would bear on whether it was meant to be so matched in fixing damages, the Court addressed specific questions on that subject to the parties.1

*593I. Prior Development.

Legal developments before the enactment of the present ORS 18.470 in 1975 can be briefly summarized. This court recognized a tort action for injuries caused by a dangerously defective product in a series of cases beginning with Heaton v. Ford Motor Co., 248 Or 467, 435 P2d 806 (1967). In 1971, the Legislative Assembly enacted the first version of ORS 18.470 as a comparative negligence statute.2 In 1973, the court held that recovery on a products liability theory was not barred by a plaintiffs negligence in failing to discover the defect or to take precautions against its possible existence, as distinct from unreasonably using a product known to be defective. Findlay v. Copeland Lumber Co., 265 Or 300, 509 P2d 28 (1973) (citing Restatement of Torts 2d, § 402 A, comment n).

The question in Findlay was whether contributory negligence, either of the ordinary kind or of the type sometimes characterized as implied assumption of the risk, was a complete defense to a strict products liability claim. There was no occasion to consider ORS 18.470, which by its terms applied only to negligence actions. A later decision enumerated the elements of assumption of the risk that *594would make out such a complete defense. Johnson v. Clark Equipment Co., 274 Or 403, 547 P2d 132 (1976).3 After the decision in Findlay, the 1975 legislature made two significant changes in the relevant law. ORS 18.470 was amended to read:

“Contributory negligence shall not bar recovery in an action by any person or his legal representative to recover damages for death or injury to person or property if the fault attributable to the person seeking recovery was not greater than the combined fault of the person or persons against whom recovery is sought, but any damages allowed shall be diminished in the proportion to the percentage of fault attributable to the person recovering. This section is not intended to create or abolish any defense.”

The same chapter of the 1975 laws also enacted ORS 18.475, which abolished the “doctrines” of “last clear chance” and “implied assumption of the risk.”4 Or Laws 1975, ch 599.

We reviewed the foregoing developments and the legislative history of the 1975 amendment in Baccelleri v. Hyster Co., 287 Or 3, 597 P2d 351 (1979), in which a forklift truck which lacked an automatic warning signal had backed over the legs of a kneeling worker. A verdict for defendant was reversed because the trial court erroneously submitted a defense of assumption of the risk to the jury. This court went on to state that on a possible retrial, the conduct of the injured worker that was charged as an implied assumption of the risk might instead be a form of contributory negligence to be pleaded and compared as “fault” for purposes of the amended proportionate fault statute, ORS 18.470.

*595II. The present dispute.

Uncertainty about the comparison of “fault” in products liability cases was not wholly laid to rest by Baccelleri v. Hyster Co., supra. Some doubts about the reach of ORS 18.475 remained because Baccelleri dealt with alleged contributory negligence of a kind that defendant had characterized as “implied assumption of the risk.” The decision therefore held only that “conduct which was sometimes labeled assumption of the risk but which is a subspecies of contributory negligence can be compared in the apportionment of damages,” 287 Or at 10, and that “comparative fault is applicable to strict liability in tort;” but it did not reach the question whether defendant had shown “that kind of contributory negligence which can qualify as comparative fault in a strict liability case.” 287 Or at 12.

The Court of Appeals, faced with this limited guidance, read Baccelleri as extending proportionate fault in products liability cases only to the kind of conduct by plaintiff that previously had been raised as a defense in Findlay v. Copeland and Johnson v. Clark Equip. Co., supra, under the rubric “assumption of the risk,” Holdsclaw v. Warren & Brewster, 45 Or App 153, 607 P2d 1208 (1980), and. it followed that holding in the present case. Upon careful consideration of the 1975 legislation, however, we conclude that the legislation did not so confine the “fault” on the part of plaintiff to be compared with defendant’s “fault,” when each was a cause of the injury.

A comparison of ORS 18.470 before and after the 1975 amendment, quoted above, shows the following changes.

First, the 1971 version stated that “[contributory negligence, including assumption of the risk,” was not to bar recovery in a negligence action if it “was not as great as the negligence of the person against whom recovery is sought,” but plaintiffs damages should be “diminished in the proportion to the amount of such negligence.” The 1975 amendment removed the words “including assumption of the risk” from the reference to the “contributory negligence” that was no longer to be a bar. Standing alone, the change might suggest that implied assumption of the risk *596was once again to bar recovery, although contributory negligence would not. The context and the legislative history, however, show the contrary. More plausibly, the reference to “assumption of the risk” was deleted from ORS 18.470 because the 1975 act abolished the “doctrine of implied assumption of the risk” altogether. ORS 18.475, supra. Far from reintroducing a distinction between the terms “assumption of the risk” and “contributory negliÍsnce,” the explanations accompanying the bill stated that ¡ontributory negligence” in the statute should be “broadly mstrued” to include assumption of the risk in the form of \ unreasonably proceeding to encounter a known danger, / which henceforth should be treated like any other contributory negligence. See the sources quoted in Baccelleri v. Hyster Co., 287 Or at 10.

Second, the 1975 amendment replaced the reference to an action for “damages for negligence resulting in death or injury to person or property” with one to recover “damages for death or injury to person or property” without limitation to negligence actions. Third, it substituted a comparison of the parties’ relative “fault” for their relative “negligence,” and also substituted the combined fault of several defendants for the previous reference to a single defendant. Fourth, it cut off the defense of contributory negligence when the injured claimant’s fault was “not greater than” that of the defendants’ fault rather than when it was “not as great.” Fifth, it specified that the “proportion” of the claimant’s fault be stated as a “percentage of fault.” It left unexplained, however, of what total entity the claimant’s fault is to be stated as a percentage. Nor did it provide that this percentage was to be compared with a percentage of fault attributable to defendants. Indeed, the words “compare” or “comparison” do not appear in the statute, which speaks only of diminishing a plaintiffs recovery “in the proportion” of plaintiffs fault. Finally, the amendment added the closing sentence that ORS 18.470 “is not intended to create or abolish any defense.”

Removal of the prior reference to negligence actions and substitution of relative “fault” for “negligence” in the allocation of damages extended the principle of proportional fault on both sides to fault other than *597negligence. As we held in Baccelleri, this included products liability, where the defendants’ “fault” lies in putting a dangerously defective product on the market.5 See Phillips v. Kimwood Machine Co., 269 Or 485, 525 P2d 1033 (1974). Nonetheless, the legislative drafters of ORS 18.470 left some textual puzzles.6 One of these is the effect of the final assertion that ORS 18.470 “is not intended to create or abolish any defense.”

As previously stated, prior to the 1975 amendments the Findlay and Johnson opinions, supra, recognized a defense to products liability when the injured party had unreasonably proceeded to encounter a known danger from the defect, a form of conduct which implicitly assumed the risk of injury posed by the known defect. Was ORS 18.470 “not intended to . . . abolish” this existing defense? A contrary conclusion emerges from the 1975 amendments as a whole and the apparent thrust of the somewhat self-contradictory explanations that accompanied its enactment.

As far as the statutory text is concerned, implied assumption of the risk was in fact abolished as a defense, not by the 1975 amendment of ORS 18.470, but by section 4 of the same act, which became ORS 18.475, supra note 3. After that enactment, implied assumption of the risk had no further place in Oregon law as a separate “doctrine” for use in pleadings, motions, jury instructions, or findings. See Hornbeck v. Western States Fire Apparatus, 280 Or 647, 572 P2d 620 (1977), Thompson v. Weaver, 277 Or 299, 560 P2d 620 (1977). But this did not foreclose a claim that, quite apart from that doctrine, the conduct of an injured party who previously might be said to have assumed the *598risk also was negligent under conventional negligence principles. This is how we understand the statement of Representative Frohnmayer, quoted in Baccelleri v. Hyster Co., 287 Or at 9, that “the form of assumption of the risk in which plaintiff voluntarily and unreasonably encounters a known risk ... is unaffected by Section 5 and should be pled as contributory negligence.”7

It is, however, impossible to square this analysis with the conclusion that only such contributory negligence as previously constituted implied assumption of the risk should be considered “fault” in a products liability case, as the Court of Appeals held. That would have the same statute which abolished implied assumption of the risk in one section revive it as a test of proportional fault in another section. We conclude, therefore, that an injured person’s conduct which in fact was a cause of her injury and which constitutes a “fault,” including negligence, is to be considered in a products liability action, unless the user’s alleged negligence consists in the kind of unobservant, inattentive, ignorant, or awkward failure to discover or to guard against the defect that goes toward making the product dangerously defective in the first place.

III. Proportional fault.

A second problem posed by the statute is the question exactly what is to be assessed in determining the “percentage of fault attributable to the person” seeking recovery, and whether that person’s fault was “greater than the combined fault of the person or persons against whom recovery is sought.” The question has puzzled commentators as well as courts. At least three views are possible.

A. Quantifying “fault.” The first is that the formula calls upon the factfinder to assess the relative magnitude of the parties’ respective “fault.” As stated by a leading textbook on these laws: “The process is not allocation of physical causation, which could be scientifically apportioned, but rather of allocating fault, which cannot be *599scientifically measured.” Schwartz, Comparative Negligence 276 (1974). It has been recognized that fault is an evaluation that does not lend itself to quantification, so that a comparison of fault magnifies the subjective elements already intrinsic to the ordinary judgment of negligence.8 This is true even in assigning proportions to two or more distinct types of negligence, but critics have found a greater theoretical obstacle when the responsibility of one party is grounded in fault other than negligence, or in no fault at all.9 The obstacle is greater where strict products liability is explained as a device for spreading losses from economic activity regardless of fault, but this court early disavowed that explanation, at least in the absence of legislation. See Phillips v. Kimwood Machine Co., supra, 269 Or 485, 503, Markle v. Mulholland’s, Inc., 265 Or 259, 264-265, 509 P2d 529 (1973), Wights v. Staff Jennings, 241 Or 301, 309-310, 405 P2d 624 (1965). We have said that the “premise of responsibility has settled on strict liability for marketing the dangerously defective product, a premise stricter than negligence but less than absolute liability.” Russell v. Ford Motor Co., 281 Or 587, 594, 575 P2d 1383 (1978). Whether the “fault” in products liability inheres in the defective product or in the act of placing it on the market, however, difficulties of comparison with the injured party’s fault undeniably remain.10

*600B. “Comparative causation.” Some courts, applying comparative negligence law to products liability under statutes different from ours or under no statute, have tried to escape the difficulty by stating that the allocation of damages is to reflect relative causation, that is to say, an assessment of the proportion in which the plaintiffs injuries were caused by the product defect on the one hand and by plaintiffs own negligence on the other. See, e.g., Murray v. Fairbanks Morse, 610 F2d 149 (3rd Cir 1979) (applying Virgin Island comparative negligence statute) and cases cited at 159-160; Pan-Alaska Fisheries, Inc. v. Marine Const. & Design Co., 565 F2d 1129 (9th Cir 1977) (admiralty law). They have done so in the belief that this is conceptually more logical or pragmatically easier than to compare the defect of a product with the negligence of one whom it has injured. In Murray v. Fairbanks Morse, for instance, Judge Rosenn wrote for the Third Circuit:

“The key conceptual distinction between strict products liability theory and negligence is that the plaintiff need not prove faulty conduct on the part of the defendant in order to recover. The jury is not asked to determine if the defendant deviated from a standard of care in producing *601his product. There is no proven faulty conduct of the defendant to compare with the faulty conduct of the plaintiff in order to apportion the responsibility for an accident. Although we may term a defective product “faulty,” it is qualitatively different from the plaintiff’s conduct that contributes to his injury. A comparison of the two is therefore inappropriate. ...”

“We believe that if the loss for a particular injury is to be apportioned between the product defect and the plaintiffs misconduct, the only conceptual basis for comparison is the causative contribution of each to the particular loss or injury. In apportioning damages we are really asking how much of the injury was caused by the defect in the product versus how much was caused by the plaintiffs own actions.

610 F2d at 159.11 With due respect to these courts, however, we are not persuaded that the concept of “comparative causation” is more cogent or meaningful than comparative fault, if by “causation” is meant some relation of cause and effect in the physical world rather than the very attribution of responsibility for which “causation” is to serve as the premise.

Both the defect and the plaintiffs fault must in fact be causes of one injury before a question of apportionment of fault arises. Although defendants in this case had completed all acts necessary for liability when they manufactured and mounted a dangerously defective tire that might blow out and overturn the Sandford pickup, they obviously would not be liable if the pickup overturned for some unrelated reason. Similarly, it would not matter that a driver operated his car unlawfully or recklessly if he was injured by an explosion due to an electrical defect that would have occurred with the same harmful consequences if the car had been standing still. In less obvious situations where the physical course of events is in doubt, if either party convinces the factfinder that its misconduct in fact *602was not a cause of the injury, there is no occasion for allocating partial damages.

The concept of apportioning causation must be tested on the assumption that both causes had to join to produce the injury for which damages are to be allocated. There are cases in which it may be possible to segregate the harm done by one cause from different or incremental harm done by a second cause, so as to apply proportional allocation to the additional harm only. This might be possible when a quantitative increase in a source of harm causes a corresponding increase in the injury, such as side effects from a negligent overdose of a dangerously defective drug, or if, for instance, Mrs. Sandford had broken a leg in an accident caused by the defective tire and thereafter had been burned by material negligently stored in her vehicle. Alternative solutions of such problems are discussed in Twerski, The Use and Abuse of Comparative Negligence in Products Liability, 10 Ind L Rev 796, 810-823 (1977); as the litigants in today’s cases present no such issues, we do not pursue them here. For the same reason, we do not discuss situations in which either the defect or the injured party’s negligence alone would cause the entire injury, as in a claimant’s neglect properly to use faulty safety equipment which would not have functioned if used, when the parties might debate whether the negligence or the preexisting defect doomed the claimant to suffer his injuries. Once it is assumed, however, that two or more distinct causes had to occur to produce an indivisible injury, we doubt that the purpose of the proportional fault concept is to subject the combined causation to some kind of vector analysis, even in the rare case of simultaneous, physically commensurable forces.12 In most cases, it would be a vain exercise *603to search for a common physical measure for the causative effect of a product defect and of the injured party’s negligent conduct.

C. Mixing “fault” with “proximate” causation. A third view considers it futile to attempt to explain what is to be compared, because it is equally illogical to compare strict liability with negligence and to quantify the relative causative effect of either when it would have caused no harm in the absence of the other. Thus Dean Twerski, in the cited article, describes the technical problems of making the comparison as a “red herring”: “The short answer to the dilemma of how one can compare strict liability and negligence is that one must simply close one’s eyes and accomplish the task.” 10 Ind L Rev 796, 806-808.13 The opinion in Pan-Alaska, supra, similarly questions the significance of theoretical distinctions when it states:

“In any event, whether we use the term comparative fault, contributory negligence, comparative causation, or even comparative blameworthiness, we are merely beating around the semantical bush seeking to achieve an equitable method of allocating the responsibility for an injury or loss.

565 F2d at 1139. In part this second view rests on the assumption that rational analysis in tort cases dissolves in the collegial judgment of juries.14" That is probably an *604unwarranted generalization; ability and effort to decide in accordance with law can be expected to differ from one jury to the next with such variables as the makeup of the particular jury, with the quality of evidence and advocacy, and not least with the rationality of the legal formulations in which the court explains the jury’s task to it.

In any event, the assumption does not let us escape the need to state coherent rules of liability. Some tort cases are tried to the court without a jury, as Pan-Alaska, supra, illustrates. Trial judges must know on what findings an apportionment of damages depends, whether these are to be made by the judge or by a jury. Our system of appeal as well as trial predicates that jurors will conscientiously attempt to apply the law if it is explained in comprehensible terms. Sedillo v. City of Portland, 234 Or 28, 33, 380 P2d 115 (1963), Stage v. St. Pierre, 224 Or 395, 401, 356 P2d 432 (1960); Williams v. Portland Gen. Elec., 195 Or 597, 610, 247 P2d 494 (1952). A juror who wants to know how to treat cause and how to treat fault is entitled to an answer, whatever comes of it in a collective decision. Counsel need to know whether to address the relative gravity of the parties’ fault or to seek expert testimony on the relative impact of their respective fault in causing the asserted harm. We cannot dismiss the question as a distinction without a difference.

The National Conference of Commissioners on Uniform State Laws attempted to overcome the distinction, or perhaps split the difference, in a proposed Uniform Comparative Fault Act, by calling on the factfinder, in determining the percentages of fault, to “consider both the nature of the conduct of each party at fault and the extent of the causal relation between the conduct and the damages claimed.” See Wade, Products Liability and Plaintiffs Fault — The Uniform Comparative Fault Act, 29 Mercer L Rev 373 (1978). This might add the difficulties of comparing causation to those of comparing fault if causation in fact were meant, especially since the act also calls for special findings. But the comments indicate that “the extent of the causal relation” does not mean causation in fact but what has traditionally been labeled “proximate *605cause.”15 In Murray v. Fairbanks Morse, supra, also, “proximate” causation rather than factual causation turns out to be what the court means by “comparative causation.”16

This court has declined to phrase as issues of causation what actually are issues of legal responsibility at least since Justice O’Connell’s concurring opinion in Dewey v. A.F. Klaveness & Co., 233 Or 515, 519-45, 379 P2d 560 *606(1963). See Vetri, Tort Markings: Chief Justice O’Connell’s Contributions to Tort Law, 56 Or L Rev 235, 238-242 (1977). Causation in Oregon law refers to causation in fact, that is to say, whether someone examining the event without regard to legal consequences would conclude that the allegedly faulty conduct or condition in fact played a role in its occurrence. “ ‘Causation in fact’ is unrelated to ‘proximate’ or ‘legal’ cause, concepts which have been discarded by this court.” McEwen v. Ortho Pharmaceutical, 270 Or 375, 385 n. 7, 528 P2d 522 (1974); see also Babler Bros. Inc. v. Pacific Intermountain Express Co., 244 Or 459, 415 P2d 735 (1966). What can be a cause in fact is a person’s behavior, which is a fact, not its faulty or faultless character, which is a legal characterization. In many cases, therefore, there will be no question of causation for a factfinder to decide. Once the test of legal responsibility thus is no longer phrased as a quantum of causation, it would mark a departure if such a phrasing were reintroduced by the proportionate fault statute. An examination of the statute shows that it was not.

D. Proportionate fault under ORS 18.470. ORS 18.470, supra, by its terms applies whenever “the fault attributable to the person seeking recovery was not greater than the combined fault of the person or persons against whom recovery is sought.” If there was such fault, “any damages allowed shall be diminished in the proportion to the percentage of fault attributable to the person recovering.” There is no reference to causation, or to any question how much the fault of each contributed to the injury. Indeed, the reference to negligence “contributing to the injury” in former ORS 18.470 was removed in the 1975 amendment. We do not mean that the allegedly faulty conduct or condition need not have affected the event for which recovery is sought; as we have said, it must have been a cause in fact. But the statute does not call for apportioning damages by quantifying the contribution of several causes that had to coincide to produce the injury.

Rather, ORS 18.470 falls within the first of the different approaches that we have reviewed. It calls upon the factfinder to assess and quantify fault. If the plaintiffs conduct is not faultless, the assessment has two purposes: *607to determine whether her fault is “not greater than” that of defendants, and if it is not, then to reduce the plaintiffs recovery of damages “in the proportion to the percentage of fault attributable to” the plaintiff. The question remains against what standard this “percentage of fault” introduced by the 1975 amendment is to be measured.

The answer is implicit in the two steps found in the statute, one in the amended ORS 18.470 and the second in section 2 of the act, codified in ORS 18.480. First, if the plaintiffs behavior which was one cause of the injury is alleged to have been negligent or otherwise “fault,” it is to be measured against behavior that would have been faultless under the circumstances. The factfinder is to determine the degree to which the plaintiffs behavior fell short of that norm and express this deficit as a numerical percentage, which then is applied to diminish the recoverable damages. There necessarily must be some comparable assessment of the fault attributable to defendants as a departure from the norm invoked against them (which, in products liability, will involve the magnitude of the defect rather than negligence or moral “blameworthiness”) in order to determine which is greater.17 In this comparison, the benchmark for assessing a defendant’s fault, for marketing a product which is dangerously defective in design, manufacture, or warning is what the product should have been without the defect. The benchmark for the injured claimant’s fault is conduct which would not be unlawful or careless in any relevant respect. This corresponds to views expressed by the Supreme Judicial Court of Maine:

*608“[Apportionment is on the basis of fault or blame. This involves a comparison of the culpability of the parties, meaning by culpability not moral blame but the degree of departure from the standard of a reasonable man. . . . [C]omparison is invited between degrees of fault which may range from trivial inadvertence to the grossest recklessness. ... In judging the conduct of an actor it should be considered complete carefulness is at one end, a deliberate intention to bring about the result is at the other. Negligence ranges from the least blameworthy type, namely, inadvertence and negligent errors of judgment up to a state where knowledge or more complete knowledge supervenes and the negligence of obstinacy, self-righteousness or reckless is reached. The factfinder must be told then under our statute, it should give consideration to the relative blameworthiness of the causative fault of the claimant and of the defendant.”

Wing v. Morse, 300 A2d 491, 500 (Me 1973). Quoting this excerpt, one writer suggests:

“One method of conceptualizing the share of the negligence [or “fault”] attributable to each party is to imagine a ‘fault line,’ with the absence of fault at one end having a value of zero and a deliberate wrongdoing at the other having a value of ten. The fact finder would then establish where on this line the conduct of each party falls.”

Pearson, Apportionment of Losses Under Comparative Fault Laws — An Analysis of the Alternatives, 40 La L Rev, 346, 348-9 (1980). Although Professor Pearson acknowledges that this “fault line” method cannot be as precise as it *609sounds, he suggests that “it does provide a means of visualizing the fact finder’s task, which perhaps could be put into a useful form.” Id.

If the claimant’s fault in this sense is greater than the fault of defendants, recovery is barred. Thereafter, ORS 18.470 standing alone does not seem to assign any further role to the magnitude of defendants’ fault in calculating the percentage of plaintiffs damages that she may recover.18 Section 2 of the 1975 act, however, added directions for the second step. ORS 18.480 provides:

“(1) When requested by any party the trier of fact shall answer special questions indicating:

“(a) The amount of damages to which a party seeking recovery would be entitled, assuming that party not to be at fault;

“(b) The degree of each party’s fault expressed as a percentage of the total fault attributable to all parties represented in the action.

“(2) A jury shall be informed of the legal effect of its answer to the questions listed in subsection (1) of this section.”

Accordingly, after determining whether and how far each party’s conduct was at fault, measured against the norm governing that party’s conduct, these respective degrees of fault are to be converted into a percentage which will be applied to the plaintiffs total damages to determine his actual recovery.19

*610 To summarize: When an injured claimant’s misconduct is a cause in fact of the injury, it can defeat a products liability claim if the claimant’s fault is “greater than” the defendants’ combined fault involved in marketing the defective product. If it is not greater, plaintiffs fault proportionately reduces her recoverable damages. “Fault” includes contributory negligence except for such unobservant, inattentive, ignorant, or awkward failure of the injured party to discover the defect or to guard against it as is taken into account in finding the particular product dangerously defective. Accordingly, we reject plaintiffs assertion that ordinary negligence on her part would not be an offset or possible defense against her products liability claim.

IV. The allegations of plaintiff’s negligence.

In view of the foregoing discussion, applicable also to Wilson v. B.F. Goodrich, 292 Or 626, 642 P2d 644 (1982), the trial court in principle did not err in submitting plaintiffs alleged negligence to the jury for consideration in denying or reducing plaintiffs damages. The parties dispute, however, whether there was evidence to support the specifications of negligence alleged by defendants.

The trial court submitted to the jury allegations that plaintiff was negligent in operating the pickup and camper fully loaded when she was not familiar with doing so, in failing to keep proper control, and in operating at an excessive rate of speed under the conditions. The court struck another allegation by defendants that plaintiff was negligent in operating the vehicle with three full gas tanks and six additional five-gallon cans of gas inside the camper.

*611Plaintiff asserted on appeal that there was insufficient evidence on which to submit to the jury those allegations which were submitted, but on review in this court plaintiff waived that objection if this court found evidence sufficient to support any one of the allegations. Plaintiffs major reliance was on the legal argument already discussed that plaintiffs alleged negligence was not the kind that would reduce or deny her damages in a products liability case.

We find sufficient evidence in the record to submit to the jury the issue of plaintiffs familiarity with the operating characteristics of the loaded pickup and camper at the time of the accident. Shortly before that day a power steering unit had been installed in the pickup. Plaintiff had driven the truck with power steering only once before that day, when it had not been loaded with equipment for the camping trip. At the time of the accident, the vehicle began swaying and weaving on the highway before it eventually rolled over and slid on the pavement. It is not impossible that plaintiffs attempts to control the heavily loaded vehicle when the tire failed were adversely affected by her unfamiliarity with the behavior of the power steering system. A jury might so conclude under one of the first two specifications mentioned above.

Defendants contend that the court should have submitted their allegation that plaintiff caused or contributed to her own injuries by operating the vehicle with its numerous extra containers of gasoline. The parties argued this point, in the light of the position earlier taken by the Court of Appeals in Holdsclaw v. Warren & Brewster, supra, largely on the legal issue whether this allegation stated the kind of voluntary assumption of an obvious hazard that would count against the plaintiffs recovery in a products liability claim. Carrying extra gasoline cans is not, of course, a hazard of an unexpectedly defective tire as distinct from anything else that might cause some of the gasoline to escape and ignite, and the trial court did not regard it as an appropriate invocation of the defense left open by Holdsclaw. Although we hold that ordinary negligence can suffice as an offset or defense in a products liability case, it was not error to reject the allegation on the *612basis on which it was pleaded and argued to the trial court.20 These rulings are not grounds for a new trial.

V. The jury poll.

The Court of Appeals reversed the judgment entered on the jury verdict because it found that the manner in which the jury was polled did not adequately assure that three-fourths of the jury had concurred in all parts of the verdict. ORS 17.355.21

The jury had been given a verdict form which called upon it to answer four separate questions. After seven and one-half hours of deliberation, the jury reported a deadlock of eight votes to four. The court instructed it to continue trying. After several more hours, the jury sent word that it had a question, which it was told to put in writing. Instead of the question, however, the jury reported that it had reached a verdict and accordingly was recalled to the courtroom. After the clerk read the four questions and the answers on the verdict form, the following occurred:

*613“THE COURT: Mrs. Horst, was your verdict unanimous? “MRS. HORST: It was nine to three.

“THE COURT: All right, we are going to poll the jury and each of you will be asked this question: Is this your verdict. And your only response to that question should be either ‘yes’, if this was your verdict or ‘no’, if this was not your verdict. Any questions about the polling?”

When the clerk polled the jury, all twelve of the jurors answered “yes.” The court announced that it would receive the verdict and discharge the jury. Counsel for defendants requested that the jury first be polled again as to each of the questions posed in the verdict form, expressing concern that the difference between the poll and the previously reported vote showed confusion on the part of the jurors. This request was denied.

This court has held, and the parties do not dispute, that the same jurors must constitute the three-fourths majority that finds every separate element required for the verdict. Munger v. S.I.A.C., 243 Or 419, 414 P2d 328 (1966), Schultz v. Monterey, 232 Or 421, 375 P2d 829 (1962), Clark v. Strain, 212 Or 357, 319 P2d 940 (1958). It is proper so to instruct the jury. See Aronson v. Fagan, 278 Or 135, 562 P2d 974 (1977). In that case, the trial court had done so, and when polling the jury after a nonunanimous verdict, the court again reminded the jurors: “Of course, those of you that say ‘yes’ have to agree with each of the answers in the several questions.” Under those circumstances, this court held that there was no prejudice in failing also to poll each juror on each separate question. 278 Or at 138. When a trial court failed to give the instruction, on the other hand, the error was said to have been waived because the appellant had not taken advantage of an opportunity to have the jury polled on each separate issue. Whelchel v. Strangways, 275 Or 297, 550 P2d 1228 (1976).

The Court of Appeals concluded that the statutory right to have the jury polled is designed to demonstrate that the announced result represents a valid verdict and therefore includes the right to have the jurors polled whether they concur in each part of the verdict.22 We agree. *614It remains to decide whether this requires reversal of the present judgment, as the Court of Appeals held.

Plaintiff relies on prior decisions of this court that have imposed high evidentiary demands on a party seeking to impeach a jury verdict. See, e.g., Blanton v. Union Pacific Railroad Co., 289 Or 617, 628, 616 P2d 477 (1980) (post-trial letter from juror that verdict was an improper quotient verdict). Blanton cited Carson v. Brauer, 234 Or 333, 382 P2d 79 (1963) (post-trial affidavits concerning comments among jurors), State v. Gardner, 230 Or 569, 371 P2d 558 (1962) (juror’s comments based on acquaintance with defendant), and Schmitz v. Yant, 242 Or 308, 409 P2d 346 (1965) (prospective juror’s comments to other jurors). These decisions are distinguishable insofar as each involved the use of post-trial affidavits or other evidence to show that a verdict was tainted by misconduct. In the present case, we are concerned with the effect of omitting a procedure which is designed to test the numerical validity of a verdict at the time it is rendered.

As the Court of Appeals noted, the jurors earlier had been divided eight to four after more than seven hours of deliberation, and reported that they still had a question just before they returned a verdict. It is not inconceivable that the difficulty lay in finding nine jurors to agree to each of the four answers, and that it may have been met by finding a different ninth vote for one or more of the answers, which would not be a valid verdict. Under the circumstances, the court held that it was impossible to say that the failure correctly to poll the divided jury was harmless error. We cannot disagree with that conclusion. The decision of the Court of Appeals is affirmed.

PETERSON, J.,

concurring.

I concur with the opinion of the court in all respects except one. I disassociate myself from part III of the opinion, not because I necessarily disagree with what is stated therein, but because that part of the opinion is totally unnecessary to the decision of this case, involves questions not raised or considered in either the trial court or the Court of Appeals and not argued or discussed by the parties, and adopts significant rules of law on important controversial questions without the issues being presented, contended for, briefed, or argued.

*615It is clear that the parties, from the beginning of the case, disputed whether the contributory negligence of plaintiff would bar or reduce a recovery when the plaintiffs claim is based on a strict liability theory. But there was no issue in the trial court relating to the quantification of fault, the comparison or apportionment of fault, or the manner by which the plaintiffs comparative fault would be calculated to bar or reduce recovery.

Although the plaintiffs attorney took exception to the giving of any comparative negligence instructions, no other exceptions to the court’s comparative fault instructions were taken.1 On appeal, the plaintiff claimed that the court erred in instructing the jury that the plaintiffs comparative negligence could bar or reduce her recovery. This was one of the issues tendered to the Court of Appeals for decision, and this issue was decided by the Court of Appeals in the plaintiffs favor. There were no other claims of error addressing liability.

The defendants, in their petition for review, claimed only (1) that the Court of Appeals erred in failing to hold that all forms of contributory negligence “can be considered for comparative purposes in a strict liability action” and (2) that the Court of Appeals erred in “* * * * ruling that the jury could not consider the fault of the plaintiff in voluntarily encountering an obvious risk or hazard * * *.” The court’s opinion decides all of the questions presented in the petitions for review. In addition, in part III of the opinion the court adopts rules relating to (1) what conduct determines fault; (2) against what standard the conduct is compared to determine the amount of a *616party’s fault; and (3) the manner by which the “* * * respective degrees of fault are to be converted into a percentage which will be applied to the plaintiffs total damages to determine his actual recovery.”

The question whether the determination of fault is to be made (1) by the extent to which fault caused the injury, (2) by comparing “fault,” or (3) by making both comparisons, is a significant, provocative, controversial, important question.2 I don’t suggest that we back away from provocative and controversial questions. But we should not adopt such significant rules of law when no facts are before us for decision, where the issue has not been presented for decision, when the issue has not been raised by any party, and when a ruling on the issue is not necessary to the disposition of the case.3

*617I am sensitive to the fact that these are important questions. But I would rather make all decisions when the issue is properly before us, rather than in the abstract, without argument from anyone, and without facts to look to in making the decision. All questions, especially important questions, are best decided when the parties raise them, brief them, and argue them. There may be significant reasons not revealed by our own research and consideration which point toward a different conclusion. Even though we are a court of review (Rule of Appellate Procedure 10.05), nonetheless our function is to decide cases that come before us, not to muse about “what the rule of law should be if.”

The advent of strict liability and comparative negligence has created difficult, provocative, challenging and important questions. In the past generation, we have decided a host of cases involving such questions. But to the best of my knowledge, since the enactment of ORS 18.470 in 1971,4 no party has ever claimed error by reason of the instructions of the court relating to the manner in which “fault” is to be determined or apportioned, or the manner in which the plaintiffs damages are to be diminished by reason of such fault. Obviously, the development of new rules of law is necessary if courts are to be responsive to societal needs. But rules of law are best forged on the anvil of actual controversy rather than by taking a massive bite out of the carcass of a field of law and spitting out a rule of law which, in the abstract, tastes good. We have repeatedly refused to give advisory opinions, even though the parties have asked for them. We should not gratuitously render advisory or academic opinions which are not sought. See Oregon Medical Association v. Rawls, 281 Or 293, 296-300, 574 P2d 1103 (1978); Oregon Medical Assn. v. Rawls, 276 Or 1101, 1108, 557 P2d 664 (1977); Smith v. Smith, 209 Or 96, 98, 304 P2d 421 (1956).

I have no quarrel with concurring opinions which point up problems in a given area. Such concurring opinions perform a beneficial function and often lead to beneficial results. See, for example, the concurring opinion of *618Justice Linde in State v. Stroup, 290 Or 185, 620 P2d 1359 (1980), which flagged an issue which is even now being considered by this court in State v. Buttrey, 54 Or App 40, 642 P2d 704, rev allowed 292 Or 108 (1981); the concurring opinion of Chief Justice Denecke in State v. Wolfe, 273 Or 518, 528, 542 P2d 482 (1975), which pointed up a problem which was recently considered by this court in State v. Douglas, 292 Or 516, 641 P2d 561 (1982); and the opinion of Justice Lent in Hendrix v. McKee, 281 Or 123, 125, 575 P2d 134 (1978), which led to our decision in Falk v. Amsberry, 290 Or 839, 626 P2d 362 (1981).

A study of the articles and cases cited in the majority opinion and in this opinion compels the conclusion that the issues with which part III is concerned are difficult, complex, controversial questions which are far from being settled with any substantial uniformity. On the question of allocation of fault under Connecticut’s then new comparative negligence law, Professor Fleming James wrote:

“What is to be compared. A perennial question under comparative negligence statutes is whether apportionment is to be made by comparing fault, or by comparing the extent to which fault contributed to the injury, or by making both comparisons. The statutory language suggests that only faults are to be compared: ‘damages. . . shall be diminished in the proportion of the percentage of negligence attributable to the person recovering.’ And Prosser says bluntly, that ‘once causation is found, the apportionment must be made on the basis of comparative fault, rather than comparative contribution.’

“Courts which have dealt with comparative negligence statutes have not, however, embraced the appealing simplicity of this view. The Wisconsin statute is worded very much like ours. In Kohler v. Dumke defendant urged a view diametrically opposite to Prosser’s, that if plaintiff were found negligent ‘then in resolving the comparative negligence issue only the element of causation should be considered.’ The court disagreed, saying, ‘We deem it clear that the word “negligence” in the comparative negligence statute. . . means causal negligence . . Therefore, in comparing the negligence of two or more persons, the jury is to consider both the elements of negligence and causation.’ The Mississippi statute, which is worded like the Wisconsin statute in this respect, has been construed the same way. The Maine statute, patterned after the British, *619is somewhat differently worded. Under both of these comparison is made both of fault and of causal contribution. Professor Fleming, with an extensive background in the law of the British Commonwealth, takes more or less the same view. The weight of authority, then, inclines heavily to this view and it seems likely that Connecticut courts will follow it.”5 (Footnotes omitted; emphasis in original.)

On the question of how fault is to be determined the answer is anything but clear, as is seen by these commentators.

“On these foundations, we may proceed to examine the theoretical concept of ‘degrees of fault.’ Immediately, we encounter an anomaly, with which we will be required to struggle throughout our study. ‘Fault’ is not a finite object. It is, rather, an evaluative judgment about the quality of conduct or performance. As such, it is not a judgment which proceeds out of sensory observation or measurement in the same direct sense that quantitative evaluations or conclusions do. A judgment of fault, in other words, is only mediately supported by objective evidence; immediately and directly, it is a subjective legal or ethical pronouncement of responsibility for wrongdoing.

* * * *

“It is out of a candid recognition of the essential subjectivity of these judgments that ‘the rule of law’ attempts, by book and by instruction, to superimpose its own standards and definitions on subjective conceptions of fault and nonfault, and solemnly charges its judges and jurors to adopt and apply those standards and definitions in the process of litigation. In this vein, we devise careful legal definitions of such concepts as ‘negligence’ and ‘cause,’ identifying specific standards for determining their presence or absence.

“If, for lack of any alternative, we are compelled to accept the somewhat naive notion that this process of superimposition really works to unseat the studied or instinctive preconceptions of the individuals concerned, it is sheer fantasy to assume that subjective standards of evaluation will not continue to govern the matter where no rational or objective legal standard or definition is possible, or is even suggested. Such is the case with respect to *620the concept of ‘degrees of causative negligence.’ ”6 (Emphasis in original.)

On cause-in-fact apportionment, Professor Twerski writes:

“It is time to consider the use of the comparative fault doctrine to include within its sweep not only fault and proximate cause but cause-in-fact as well. In my opinion, the inevitable effect of instructing a jury that proximate cause is an item for comparison will be that cause-in-fact will be factored into the jury’s consideration. * * *

“As a statement of fact and pure logic, it is clear that cause-in-fact is not subject to apportionment. But in our saner moments, we ought to be ready to admit that cause-in-fact is one of the most intractable items to prove in a law suit. * * *

<<* * * * *

“In reality I believe that this will occur when cases are turned over to juries on a comparative fault instruction. Jurors, not being schooled in the separate pigeon holes created by tort teachers, are still naive enough to believe that an accident does not take place in five stages. Duty, standard of care, cause-in-fact, proximate cause and damages are supposed to be analytical aids. They are not descriptive of the process of accomplishing a tort. Thus, juries will have to be excused if they view the entire injury event as a unitary whole and factor the probability of causation together with fault in arriving at a percentage apportionment. * * *.” (Footnotes omitted.)7

The Uniform Comparative Fault Act recognizes the difficulty in determining fault. Section 2(b) of the Uniform Comparative Fault Act (12 ULA 37, 1982 Pocket) provides:

“(2)(b) In determining the percentages of fault, the trier of fact shall consider both the nature of the conduct of each party at fault and the extent of the causal relation between the conduct and the damages claimed.”

Although I may agree with the holding of part III of the opinion if and when the questions are presented, I have serious reservations as to the practical implications of *621the court’s holding. To implement the holding of part III of the court’s opinion lawyers must formulate requested instructions and trial judges must give instructions to the jury concerning how fault is to be determined, how fault is to be apportioned and how plaintiffs damages are to be diminished by reason of the plaintiffs fault. I have no doubt that this will be done, for now it must be done, but I have reservations whether the desired end product — a verdict based upon objectivity and fairness and in accordance with the law — is more likely to be achieved than is now the case under current Oregon practice. See Oregon Uniform Jury Instructions 11.50-11.60. To .accommodate the rules set forth in Part III of the court’s opinion among the instructions that a trial judge will have to give upon proper request are these:

1. What fault is.

2. The standard against which a party’s fault is to be compared in determining the party’s percentage of fault.

3. That although “[b]oth the defect and the plaintiffs fault must in fact be causes of one injury before a question of apportionment of fault arises” (majority opinion at 601), once the jury finds causation in fact to exist they are not to consider the causal relationship between the fault and the injury in apportioning the party’s fault. Rather, their apportionment of fault must be made by comparing the party’s fault against “* * * behavior that would have been faultless under the circumstances * *

4. How the jury is to calculate the degree of each party’s fault against the norm.

Comment: The majority would require that the initial determination of fault be made against a standard of behavior “that would have been faultless under the circumstances” or by comparing the product to “what the product should have been without the defect.” The result could be expressed numerically, by a percentage, or otherwise. The number is to reflect the degree by which or to which the behavior falls short of the norm, and must then be expressed in terms of a “* * * deficit as a numerical percentage, which then is applied to diminish the recoverable damages.” Majority opinion at 607.

*6225. How the jury should convert the party’s percentage of fault into “* * * a percentage of the total fault attributable to all parties represented in the action,” as required by ORS 18.480.

Comment: Keep in mind that ORS 18.480 requires that, upon request, the jury “* * * shall answer special questions indicating: * * * (b) The degree of each party’s fault expressed as a percentage of the total fault attributable to all parties represented in the action” (Emphasis added). The jury will have to convert the percentage by which the party’s conduct “fell short of [the] norm” into “a percentage of the total fault attributable to all parties represented in the action,” as required by ORS 18.480(1) (b). If the jury decided that a plaintiffs fault, when tested against the norm of faultless conduct, were 20 percent, and that a defendant’s fault, when compared against the appropriate norm, were 25 percent, the jury must then convert those percentages into percentages “of the total fault,” as required by ORS 18.480(1)(b). The result: Plaintiffs percentage of fault would be 44.44 percent (20/45); the defendant’s percentage of fault would be 55.56 percent (25/45), and the plaintiffs recovery would be reduced not 20 percent, but 44.44 percent. If there were three parties, and the respective percentages of fault were plaintiff, 20 percent; defendant No. 1, 25 percent; defendant No. 2, 30 percent, the conversion into percentages “of the total fault attributable to all parties represented in the action” would be: Plaintiff, 26.67 percent (20/75); defendant No. 1, 33.33 percent (25/75); and defendant No. 2, 40 percent (30/75), and the plaintiffs recovery would be reduced by 26.67 percent.

6. Finally, the jury is required to “* * * be informed of the legal effect of its answer to the questions ***.” ORS 18.480(2).

The majority opinion itself reflects the potential for confusion which may arise in applying the rules set forth in the opinion. Footnote 18 states:

“* * * This approach would not necessarily deny all damages even to a plaintiff guilty of extreme misconduct (say ‘80 percent’ below the ‘fault line’) if the extent that defendant fell short of the norm is even greater; it would *623simply reduce plaintiffs recovery to 20 percent of her damages.”

I gather that the quoted statement applies to the theory discussed in 43 Mo L Rev 431, 449-450, for the statement cannot be correct as applied to ORS 18.470 and ORS 18.480. If plaintiffs misconduct were 80 percent, before there could be any recovery, the fault of the defendants) must equal or exceed that of the plaintiff. Thus, if there were but one defendant, whose fault was also 80 percent, the plaintiffs recoverable damages would be 50 percent of the total damages, not 20 percent. If the fault of the defendant were 90 percent, plaintiffs recoverable damages would be reduced by 47.06 percent (80/170).

The mathematics can be difficult. Regarding the mathematical calculations, David L. Nixon wrote:

“Entrusting the entire process of dollar damages apportionment to the jury under general verdict procedures will strike some as an act of faith bordering on irresponsibility. The express prohibition of special interrogatories was not contained in the bill as drafted when first introduced, although the bill did require that ‘the jury. . . diminish the damages’. As the hearings and discussions on the subject of the comparative negligence bill progressed through the Legislative Session and as knowledge was gained of Maine’s experience with its 1965 statute, the realization grew that the rough and basic justice of unfettered jury deliberation was probably preferable (and at least worth a good try) as an alternative to the horrendous mathematical processes described by opponents of the legislation as necessarily a part of the jury’s function under special verdict statutes. Most of those involved in the jury trial process, it is submitted, would agree on reflection that New Hampshire juries are populated in the main by neither misers nor spendthrifts. Others, with much in the way of research and experience to draw from, agree on the fundamental fairness and good sense of the average juror.” (Footnotes omitted.)8

The practices which have been followed in Oregon trial courts (which generally utilize the Uniform Jury Instructions) appear to me to have been generally satisfactory. Whether or not I am correct, there is no reason for *624this court to unilaterally impose far-reaching rules of procedure and substantive law on the citizenry without input from anyone.

As a young lawyer, it took me months to understand and apply the “Lamb-Weston rule,” as set forth in Lamb-Weston et al v. Ore. Auto. Ins. Co., 219 Or 110, 137-138, 341 P2d 110, 346 P2d 643, 76 ALR 2d 485 (1959), which states the principle that when there are two or more policies of valid and collectible insurance which are applicable to a loss, the total loss will be apportioned among the insurers in the ratio which the limits of each policy bear to the total available coverage. I fear that many juries will find it difficult if not impossible to understand and apply the similar rules set forth in part III of the opinion.

The doctrine of comparative negligence intends to avoid the harsh result of the common law contributory negligence rule that any negligence by the plaintiff, however small, barred recovery. Comparative fault statutes were designed to temper that rule by either granting plaintiff recovery against a negligent defendant whatever the degree of the plaintiffs fault (pure comparative fault) or by permitting the plaintiff to recover from negligent defendants if the plaintiffs fault was “not greater than” or was less than the defendant’s or the combined fault of multiple defendants. The superimposed declaration under statutes similar to Oregon’s comparative fault statute, is that the plaintiffs right to recover turns on whether the plaintiffs fault was not greater than or less than the fault of the person against whom claim is made. Inherent in the doctrine is the comparison of fault — the plaintiffs fault is compared to the defendant’s. That premise leads me to conclude that it is likely that the framers of comparative fault legislation contemplated that the fault of a party be determined by directly comparing his or her fault against the fault of other party or parties against whom claims were made, rather than by the injection of an intermediate step whereby fault is determined by comparing conduct against a separate standard and then comparing the numerical results to determine who wins or loses and by how much. I am not convinced that the drafters of ORS 18.470 intended this type of comparison.

*625The advent of strict liability has made it more difficult to compare fault based on negligence against fault based on strict liability. But jurors have been equal to that task, possibly without benefit of rational definition or standard, and until the issue arises, I would do nothing and say nothing.

Part III of the opinion of this court is unnecessary to the decision, does not involve the correction of any error, does not involve a question likely to arise on remand (except for our gratuitous reference to the matter), and involves issues which have concerned the bench and bar of this state, until now, very little or not at all. On a comparative fault basis, our fault is a 99 for publishing part III of the opinion.

Denecke, C. J., joins in this opinion.