Carl O. Akermanis, Plaintiff-Appellee-Cross-Appellant v. Sea-Land Service, Inc., Defendant-Appellant-Cross-Appellee

NEWMAN, Circuit Judge: This appeal concerns primarily the issue, apparently one of first impression in the federal courts, whether a trial judge may use the device of a new trial order conditioned on a remittitur to increase a jury’s determination of the percent of responsibility for an injury that is attributable to a plaintiff’s contributory negligence. That device was employed in this suit under the Jones Act, 46 U.S.C. § 688 (1976), brought by plaintiff Carl 0. Akermanis, an injured seaman, against defendant shipowner Sea-Land Service, Inc. After plaintiff agreed to the remittitur, the District Court for the Southern District of New York (Charles S. Haight, Jr., Judge) entered judgment on November 17,1981, in favor of the plaintiff. 521 F.Supp. 44. Defendant appeals, contending that it is entitled to an unconditional order for a new trial. Plaintiff cross-appeals, seeking an increased judgment based on the jury’s initial determination as to the share of responsibility attributable to his negligence. For reasons that follow, we conclude that a remittitur may not be used to adjust a jury’s contributory negligence percentage, and we therefore reverse and remand for further proceedings. In the spring of 1977, Akermanis was a third assistant day engineer aboard Sea-Land’s vessel S/S Los Angeles during a voyage from Greece to Rotterdam. One of the plaintiff’s duties was to replace deteriorated angle irons on pedestals used to secure the vessel’s deck cranes. At trial, Akermanis contended that the defendant’s negligence caused him to injure himself while working on these pedestals. Although there was conflicting evidence, Ak *901 ermanis testified that on June 4, 1977, his superiors instructed him to work on the deck burning off a rusted pedestal bracket, despite rolling seas and a deck slippery from the ocean’s spray. According to the plaintiff, as a result of these unsafe working conditions, he lost his balance,- hit his head against the pedestal, and hurt his back when his head was snapped backwards as he fell. The accident was said to have severely injured Akermanis’ cervical spine and forced his early retirement from the merchant marine. Sea-Land Service contested almost all of Akermanis’ allegations. First, Sea-Land claimed that Akermanis’ injuries stemmed not from an accident aboard the Los Angeles, but rather from several preexisting ailments. Second, Sea-Land contended that it was in no way to blame for the condition of the Los Angeles on June 4,1977, that Akermanis had not been ordered to work on the pedestal bracket at any particular time, and that any risks Akermanis faced while working on the deck were inherent in the life of a seaman. In addition, Sea-Land raised the defense of contributory negligence, arguing that, if working on the pedestal was unsafe, Akermanis, who was 59 years old and had 30 years of maritime experience at the time of the accident, should have been able to appreciate the danger and rearrange his schedule to work below deck until the weather cleared. According to the defendant, Akermanis had considerable discretion in scheduling his work. Moreover, counsel for Sea-Land suggested that the alleged accident was more likely the result of Akermanis’ carelessness in doing his job than any fault of the defendant. Finally, Sea-Land introduced evidence through a series of expert witnesses that contradicted the testimony of Akermanis’ doctors as to the extent and cause of his spinal injuries. In response to a special verdict form, the jury found that Sea-Land was negligent, that its negligence was a proximate cause of Akermanis’ June 4, 1977, accident, and that as a result of the accident, Akermanis suffered damages totaling $528,000. The jury further found that Akermanis also was negligent and that the share of responsibility attributable to his negligence was four percent. Following the jury’s verdict, Sea-Land moved for judgment notwithstanding the verdict or, in the alternative, for a new trial, pursuant to Fed. R. Civ. P. 50(b). After reviewing the record, Judge Haight concluded that there was sufficient evidence to support the jury’s findings that negligence on the part of the defendant caused the plaintiff’s injuries. Because it was not “ ‘clear that the jury had reached a seriously erroneous result,’ ” Bevevino v. MS. Saydjari, 574 F.2d 676, 684 (2d Cir. 1978) (quoting 6A Moore’s Federal Practice K59.08[5], at 59-160 to -161 (1973)), the District Court denied defendant’s motion for judgment notwithstanding the verdict, see Berner v. British Commonwealth Pacific Airlines, Ltd., 346 F.2d 532, 538 (2d Cir. 1965) (evidence must be viewed most favorably for the defendant on j.n.o.v. motion), cert. denied, 382 U.S. 983, 86 S.Ct. 559, 15 L.Ed.2d 472 (1966), and declined to order a new trial of all the issues. Judge Haight then considered the jury’s finding that plaintiff’s contributory negligence was a four percent cause of the accident, and concluded that the selection of this percentage was against the weight of the evidence, in his view a “clear and serious error.” While he agreed that it was reasonable for the jury to determine that Sea-Land’s negligence was the major cause of the accident, he réjected four percent as the measure of plaintiff’s share of responsibility. His reasoning is set forth in his memorandum opinion. He first assumed that the jury’s finding of plaintiff’s contributory negligence was based on their acceptance of defendant’s evidence that Akermanis had some discretion to determine when he would work on the ship’s pedestals and had exercised that discretion without using reasonable care. Judge Haight then determined that a factor of only four percent for contributory negligence of this sort was against the weight of the evidence and so substantially below jury determinations in similar Jones Act cases as to warrant a new trial. After referring to cases cited at 44A *902 Modern Federal Practice Digest Seamen § 29(4)(H) (West 1968), he concluded that the lowest contributory negligence factor the evidence would support was 25 percent. He therefore ordered a new trial on liability issues, but with the condition that the defendant’s motion for new trial would be denied if the plaintiff would accept a “remittitur” of damages based on an increase of the contributory negligence factor from four to 25 percent. The plaintiff accepted, and judgment was entered in favor of the plaintiff for 75% of the jury’s determination of the total amount of damages suffered. I. We consider first Sea-Land’s appeal, which challenges the District Court’s authority to use the device of a remittitur to adjust the jury’s determination of the contributory negligence percentage. Remittitur is a limited exception to the sanctity of jury fact-finding. It allows trial judges to reduce damages, but only when an award is grossly excessive. As the Supreme Court has stated, this exception is justified because, “[wjhere the verdict is excessive, the practice of substituting a remission of the excess for a new trial is not without plausible support in the view that what remains is included in the verdict along with the unlawful excess — in that sense that it has been found by the jury — and that the remittitur has the effect of merely lopping off an excrescence,” Dimick v. Schiedt, 293 U.S. 474, 486, 55 S.Ct. 296, 301, 79 L.Ed. 603 (1935). But in Dimick, the Supreme Court made clear that remittitur was not an expansive doctrine: “Maintenance of the jury as a fact-finding body is of such importance and occupies so firm a place in our history and jurisprudence that any seeming curtailment of the right to a jury trial should be scrutinized with the utmost care.” Id. In giving that scrutiny to the use of remittitur in this case, we note a fundamental difference between the use of remittitur to decrease a determination of damages and its use here to increase a contributory negligence percentage. Though both have the same ultimate consequence of reducing the amount of the judgment that the plaintiff is invited to accept as the price of avoiding a new trial, the means by which the reduction is accomplished differ in a way that is critical to determining the lawfulness of the technique. A conditional reduction of a damage calculation leaves in the judgment a portion of what the jury awarded, a circumstance that the Supreme Court considered crucial to its willingness to permit remittitur while rejecting additur in Dimick. In this case, however, the conditional adjustment of the contributory negligence percentage inserts into the judgment something beyond what the jury found: a conclusion that the plaintiff’s negligence was responsible for a greater share of the accident than the jury had thought. In Dimick, the four dissenters thought it was needlessly artificial to deny a trial judge the authority to condition a new trial order on payment of an additur while permitting him to use the device of remittitur. In either circumstance, they argued, the judge is simply conditioning the new trial order on the minimum adjustment necessary to render the verdict within the bounds of reasonableness. But their view did not prevail. We are therefore obliged to apply the rationale of the Dimick majority, which, as we understand it, precludes any adjustment that extends a jury’s finding, even if that extension results in a reduced monetary judgment. 1 We have located only one case in which a trial court attempted to use the remittitur device to adjust a jury’s contributory negligence assessment, and that decision did not *903 survive appellate review. In Ferguson v. Chester A. Poling, Inc., 285 N.Y.S. 340, 247 A.D. 727 (2d Dep’t 1936) (per curiam), also a Jones Act case, the jury found that the plaintiff had suffered damages of $25,000, and then reduced this sum by $5,000 because the plaintiff’s contributory negligence was a 20 percent cause of the accident. The trial court thought the lowest reasonable contributory negligence factor was 60 percent and therefore ordered a new trial unless the plaintiff agreed to a reduction in the judgment from $20,000 to $10,000. The Appellate Division reversed because “the question of apportioning the negligence was peculiarly within the province of the jury.” Id. at 341, 247 A.D. at 728. We agree with the New York court that a jury’s apportionment of responsibility in a Jones Act case is not subject to adjustment by the device of a remittitur. In Ferguson the Appellate Division ordered judgment upon the jury’s fact-finding. Whether that result or a new trial is warranted here requires consideration of plaintiff’s cross-appeal. II. A. In considering the plaintiff’s cross-appeal, we face the threshold issue of whether we have appellate jurisdiction. This Circuit has followed the majority rule that a plaintiff who accepts a remittitur may not appeal from a judgment entered upon the reduced award. Donovan v. Penn Shipping Co., 536 F.2d 536 (2d Cir. 1976), aff’d per curiam, 429 U.S. 648, 97 S.Ct. 835, 51 L.Ed.2d 112 (1977); Evans v. Calmar Steamship Co., 534 F.2d 519 (2d Cir. 1976). And we have also dismissed the cross-appeal of a plaintiff who accepted a remittitur, even though we adjudicated the merits of the defendant’s appeal from the judgment entered upon the reduced award. Mattox v. News Syndicate Co., 176 F.2d 897, 904 (2d Cir.), cert. denied, 338 U.S. 858, 70 S.Ct. 100, 94 L.Ed. 325 (1949). But see Burris v. American Chicle Co., 120 F.2d 218, 223 (2d Cir. 1941) (questioning appealability but reaching merits of cross-appeal). That rule, whatever its merit, see Donovan v. Penn Shipping Co., supra, 536 F.2d at 538 (Feinberg J. (now C.J.), dissenting); Reinertsen v. George W. Rogers Construction Corp., 519 F.2d 531 (2d Cir. 1975), does not necessarily preclude our exercise of jurisdiction over Akermanis’ cross-appeal. The rationale of Donovan v. Penn Shipping Co., supra, appears to be two-fold: waiver and efficiency. The plaintiff who has accepted a remittitur, even if the acceptance was under protest, will not be heard to complain, and allowance of such appeals would add to the burdens of courts. Neither rationale applies here. Now that we have adjudicated Sea-Land’s appeal and agreed with appellant that the District Judge lacked the power to condition the new trial order on a remittitur adjusting the contributory negligence percentage, the first step in ordering relief with respect to the appeal is, as Sea-Land requests, to vacate the judgment entered upon the reduced award on the ground that its entry was erroneous. Once we do that, it is conceptually difficult and practically unfair to think of the plaintiff as having waived a cross-appeal by consenting to a judgment that no longer exists. The rationale of court efficiency does not disappear, but it becomes less persuasive. Normally, when an appellate court dismisses the appeal or cross-appeal of a plaintiff who has accepted a remittitur, the burden on the court system is reduced because there will be neither an appeal nor a new trial. That is not true in this case. If we dismiss the cross-appeal and simply return the case to the District Court for a new trial, now that the conditional aspect of the new trial order has been excised, there would clearly be less of a burden on this Court. But that course would add considerably to the burden on the District Court, which would have to conduct a trial that the cross-appellant is prepared now to show us should never have been ordered. What we really face is an issue that has nothing to do with acceptance of a remittitur: whether a cross-appeal will lie as to a non-final order for a new trial when the main appeal from a final judgment has already brought the case to an appellate court. *904 That is similar to the situation that arises under Fed. R. Civ. P. 50(c) when a trial court grants a motion for judgment n.o.v. and also, as required by Rule 50(c), conditionally grants a motion for new trial. Upon appeal of the judgment by the party that wants judgment entered on the verdict, Rule 50(c) provides that if the judgment is reversed, “the new trial shall proceed unless the appellate court has otherwise ordered.” This provision, which the Advisory Committee explains was not intended to alter the scope of appellate review, see Fed. R. Civ. P. 50 advisory committee note, contemplates that an appellate court, after vacating the judgment n.o.v., may then review the trial court’s ruling conditionally granting a new trial and “otherwise order[]” that the new trial not be held. The result is review of a new trial order that would not have been appealable, if not accompanied by the granting of a motion for judgment n.o.v., until entry of judgment after retrial. See Taylor v. Washington Terminal Co., 409 F.2d 145 (D.C. Cir.), cert. denied, 396 U.S. 835, 90 S.Ct. 93, 24 L.Ed.2d 85 (1969). Even more analogous is the procedure authorized by Rule 50(d), which applies when appeal is taken by a party whose motion for judgment n.o.v. was denied. If that party prevails on its appeal from the judgment entered upon the verdict, the appellee is entitled to urge the court of appeals to order a new trial, rather than order the entry of judgment n.o.v. See Neely v. Martin K. Eby Construction Co., 386 U.S. 317, 329, 87 S.Ct. 1072, 1080, 18 L.Ed.2d 75 (1967) (making clear that court of appeals can consider the new trial request apart from the authority of Rule 50(d)). Our case is an amalgam of the situations contemplated by Rules 50(c) and 50(d): like the 50(c) appellant, Akermanis wants to be heard in opposition to the trial court’s grant of a new trial, but he is not the party who appealed the judgment; like the 50(d) appellee, he is the adversary of the party who appealed the judgment, but he is opposing, rather than supporting, an order for a new trial. The situations contemplated by Rules 50(c) and 50(d) are variations of the general authority of a court of appeals, upon vacating a judgment, to “require such further proceedings to be had as may be just under the circumstances.” 28 U.S.C. § 2106 (1976). The circumstances of this case are unusual, to say the least. We need not and do not consider the full range of otherwise interlocutory rulings that an appellee would like to urge an appellate court to consider upon his adversary’s successful appeal from a final judgment. But we think it is entirely appropriate to entertain the cross-appeal in this case. Doing so is in keeping with the spirit of Rules 50(c) and 50(d) and, as will be seen in Part 11(B), infra, affords us the opportunity to focus the District Court’s attention on matters that may obviate the need for a new trial or at least limit its scope. We therefore turn to the cross-appeal. B. In granting the defendant’s motion for new trial, Judge Haight reasoned that the jury’s finding of contributory negligence was based on a conclusion that Akermanis had the authority to schedule his own work and that he exercised poor judgment by agreeing to work on the Los Angeles’ deck on June 4, 1977, when he might have deferred the task to another day. 2 If that was the jury’s conclusion, then, based on the evidence presented at trial, Judge Haight was acting within his broad discretion in ruling that a finding of only four percent contributory negligence was against the weight of the evidence and that a new trial should be granted. 3 *905 It is possible, however, that the jury’s finding of contributory negligence was not based on a conclusion that Akermanis chose to work on the pedestal that day. The jury might have concluded that Akermanis was directed to work on the pedestal at the time of the accident, but that he was negligent in the manner in which he performed his work. During cross-examination, Akermanis admitted that his own preoccupation with burning the bracket might have contributed to the accident. 4 During his summation, counsel for the defendant reviewed the events surrounding Akermanis’ accident and suggested to the jury that the accident “was something due to, perhaps, a momentary lapse of carefulness on his part.” If the jury was focusing on the plaintiff’s manner of working at the time of the accident, it is arguable that they considered Akermanis’ work performance to be negligent but that such negligence represented only a slight percentage of the cause of the accident. In his memorandum opinion, Judge Haight gave no explicit consideration to this possible explanation for the jury’s contributory negligence factor of four percent. Ordinarily, we might not be so concerned that a trial judge explicitly consider every possible justification for a jury’s verdict before concluding that a verdict is against the weight of the evidence and ordering a new trial. But this case arises under the Jones Act, and jury findings under that Act, which incorporates standards of the FELA, are particularly resistant to being overturned. See Morgan v. Consolidated Rail Corp., 509 F.Supp. 281, 285 (S.D.N.Y. 1980). Because we recognize that a trial judge, with a “ ‘feel’ ” of the case, Neely v. Martin K. Eby Construction Co., supra, 386 U.S. at 325, 87 S.Ct. at 1078, is normally in the best position to determine whether a verdict is against the weight of the evidence, we prefer to have Judge Haight consider whether the “work performance” theory of contributory negligence finds enough support in the evidence, assessed in the context of the entire trial, to justify the four percent factor. If he concludes it does not, then his order for a new trial will stand. Even if he concludes that the “work performance” theory alone would support a four percent factor, then Judge Haight will have the option of either entering judgment for the plaintiff on the jury’s damage computation, reduced by only four percent, or ordering a new trial. This latter option remains a possibility because of an intimation in Judge Haight’s opinion that he thinks the jury would have committed serious error if they had not attributed at least part of the responsibility for the accident to Akermanis for continuing to work on deck despite the inclement weather and unsafe conditions. In other words, Judge Haight may find that the “work-performance” theory justifies a four percent factor, but that Akermanis’ negligence in continuing to work is so strongly supported by the evidence that a verdict assessing his share of fault at only four percent is against the weight of the evidence, whether or not the jury based its finding of contributory negligence on the “continuing-to-work” theory. In short, the District Judge, upon remand, *906 has discretion whether or not to order a new trial. If the District Court decides that a new trial is necessary to determine the extent of Akermanis’ contributory negligence, there will remain a question as to the proper scope of the retrial. Rule 59(a) permits partial retrial of distinct issues, but the trial court must examine whether a jury’s award of damages and its finding of liability are sufficiently separate to allow a partial new trial as to liability issues. See Gasoline Products Co. v. Champlin Refining Co., 283 U.S. 494, 500, 51 S.Ct. 513, 515, 75 L.Ed. 1188 (1931) (Partial retrial “may not properly be resorted to unless it clearly appears that the issue to be retried is so distinct and separable from others that a trial of it alone may be had without injustice.”). In its memorandum opinion, the District Court indicated that, if a new trial were necessary, it would be limited to questions of liability. Defendant contends that any retrial should include redetermination of damages since the jury might have perceived comparative negligence as closely related to damages and might have allowed its determination of damages to be influenced by its erroneous finding of relative fault. 5 Courts have been reluctant to hold a new trial on contributory negligence without also retrying damages. See Rivera v. Farrell Lines, Inc., 474 F.2d 255, 259 (2d Cir.) (Jones Act case), cert. denied, 414 U.S. 822, 94 S.Ct. 122, 38 L.Ed.2d 55 (1973); cf. Norfolk Southern Railroad Co. v. Ferebee, 238 U.S. 269, 273, 35 S.Ct. 781, 782, 59 L.Ed. 1303 (1915) (it would rarely be proper to allow a jury to consider a question of damages without also submitting the issue of contributory negligence). But this reluctance has been expressed in cases like Rivera, where the form of the jury verdict left the court “in the dark as to what the total award would have been absent the finding of contributory negligence and the extent to which that finding affected the verdict.” Rivera, supra, 474 F.2d at 259. The jury in that case had reported a finding of contributory negligence, but had not been asked to report the percent of responsibility it attributed to plaintiff’s contributory negligence. However, when a jury arrives at its decision by detailed special verdicts, enabling a trial or a reviewing court to be reasonably certain that an erroneous verdict was reached independent of another verdict, a partial retrial may be in order. See, e.g., Ferebee, supra. In this case, a new trial, if one is held, need not reconsider damages. Because Judge Haight submitted detailed interrogatories to the jury, following substantially the form suggested in Rivera, supra, 474 F.2d at 259 n. 5, we know the jury’s determination of aggregate damages suffered by the plaintiff, as well as their view of how much of a discount should be applied because of contributory negligence. Armed with these specific findings, the District Court acted within its discretion in excluding the issue of damages from retrial. Cf. Landry v. Two R. Drilling Co., 511 F.2d 138, 143 n. 4 (5th Cir. 1975) (court accepted 20% contributory negligence factor from first trial and limited retrial to question of damages). Although it is possible that the jury’s special verdicts encompassed some undisclosed compromise, absent obvious inconsistencies we will not presume that the jury’s findings represent anything other than good faith responses to the questions presented. As Judge Haight has already determined that the jury’s determination of damages is supported by the evidence, we conclude that damages need not be reconsidered at retrial. In his opinion conditionally ordering a new trial, Judge Haight indicated that *907 the defendant is entitled to a new trial “on the question of contributory negligence,” but the operative language of his order grants a new trial “on liability issues.” The authority provided by Rule 59(a) for a new trial on “part” of the issues has been used by some courts with precision. E.g., Higginbotham v. Ford Motor Co., 540 F.2d 762 (5th Cir. 1976). If Judge Haight determines that a new trial, if held, should be limited to the question of contributory negligence — both its existence and its percentage — he has discretion to narrow the retrial to that extent. 6 He might well conclude that since the retrial is occasioned because the first jury assigned too great a share of the blame to the defendant, Akermanis should not have to persuade the second jury that the defendant was negligent simply because he had made that demonstration too convincingly to the first jury. 7 On the other hand, the District Judge also has discretion to retry all the liability issues, if he concludes that a trial limited to contributory negligence would not, on balance, be fair to the parties. We have considered the other issues raised by the defendant concerning evidentiary rulings and the instructions to the jury and find them to be without merit. Accordingly, the judgment is reversed, and the case remanded for further consideration consistent with this opinion. Footnotes [1] It could be argued that the type of remittitur permitted by Judge Haight should be allowed since the increase in the contributory negligence factor can be accomplished only with the consent of the plaintiff, whose share of fault is being enlarged. However, it was equally true in Dimick that the increase in the amount of damages sought to be achieved by the use of additur could have been accomplished only with the consent of the defendant, who would have paid the increase. The Supreme Court’s rejection of additur, indeed, its grudging acceptance of remittitur, 293 U.S. at 484, 55 S.Ct. at 300, suggests that jury determinations are not to be enlarged upon, even with the consent of the party thereby disadvantaged. [2] As the District Court correctly noted, under the Jones Act, negligence requires more than knowledge of a hazard; it entails failure to adopt safer alternative courses of action. See Tolar v. Kinsman Marine Transit Co., 618 F.2d 1193, 1196 (6th Cir. 1980); Rivera v. Farrell Lines, Ltd., 474 F.2d 255, 257-58 (2d Cir.), cert. denied, 414 U.S. 822, 94 S.Ct. 122, 38 L.Ed.2d 55 (1973). Inquiry at the trial properly centered on what choices were available to Akermanis and how he exercised those choices. [3] Indeed, if the jury thought that Akermanis was negligent in electing to do his work at the time of the accident, there is at least a superficial inconsistency with the finding that the defendant was negligent in ordering him to do the *905 work at that time. But inconsistency is not inevitable. The jury may have thought that the defendant had negligently instructed him to work at that time, but not in the form of such an absolute order as to preclude Akermanis from exercising care in deciding whether to do the work or at least whether to call to his superiors’ attention the risks of doing the work at that time. In any event, it is a function of the comparative negligence doctrine to permit the jury to reconcile conflicting versions of an episode. [4] Akermanis’ testimony on cross-examination included the following: Q. Sir, in fact, do you know why your foot came off the pedestal? A. Because it was wet or the sea — the ship slightly rolling just plain slip. Q. You don’t know why? A. Why? It could be the ship’s rolling. I don’t know why. Q. Sir — - A. I was concentrating on my work, right. I was interested to cut off that thing. I didn’t watch my foot. The foot was standing there. I slipped while I working and the ship moved, the foot slipped. That’s why I say I don’t know why [emphasis added]. [5] While it is understandable that the defendant would prefer a second opportunity to persuade a jury to return a modest damage award, the logic of the argument is unclear: if the jury had determined the amount of damages in relation to their contributory negligence percentage, the setting of a low percentage would indicate that, if anything, damages had been scaled down from what they might have been had the contributory negligence factor been high. In this circumstance, it would normally be the plaintiff, not the defendant, who might have a claim that damages should be retried in the event of a retrial as to contributory negligence. [6] If this procedure were used, the jury would be told that defendant’s negligence has already been determined and that their task is only to determine whether the plaintiff was contributorily negligent and, if so, to determine what percentage of responsibility for the accident is attributable to the plaintiff’s contributory negligence. The parties would, of course, be entitled to present all evidence relevant to the fault of both parties in order for the jury to make an apportionment of fault. [7] In some situations, it might be inappropriate to order such a limited partial new trial. For instance, if in this case the jury’s contributory negligence factor had been impermissibly high, then any retrial might have to encompass all aspects of liability: it might be unclear from such a verdict whether the jury really thought the defendant was liable or whether their finding of liability coupled with a high assessment of contributory negligence represented a compromise between factions. Similarly, if in this case the jury’s award of damages as well as its assessment of contributory negligence had been low, then it might be necessary to afford the plaintiff the option to retry damages as well as liability: the jury might have discounted the damages to compensate for the plaintiff’s contributory negligence.