CSX Transportation, Inc., a railroad corporation, appeals the jury verdict in favor of George Kavorkian in his personal injury suit brought against CSX under the Federal Employers' Liability Act ("FELA") and the Federal Safety Appliance Act ("FSAA"). The main issue in CSX's appeal is whether, as a defense to a law suit under Sec. 2 of the FSAA, a railroad may present evidence of the alignment of a railroad car's drawbar, the mechanism that connects the car to other cars. The district court in this case declined to admit such evidence, although it received it in a separate record, and declined to give a jury instruction that the existence of an improperly aligned, but non-defective, drawbar was a defense to the FSAA claim here. Kavorkian, in turn, appeals the jury's award of damages, arguing that the evidence requires a larger award. For the reasons discussed herein, we reverse the jury's verdict on the liability issue and remand for a new trial.

* Kavorkian was employed by CSX as a brakeman/conductor. His duties included "coupling" railroad cars. A railroad car's coupling mechanism consists of a "knuckle" connected to a drawbar, which in turn is connected to a housing mechanism on the car. The knuckle of the drawbar of one car engages the knuckle of the drawbar on the other car, and the cars are coupled automatically when they come together. Once the cars are coupled, a worker secures the knuckle by moving a lever on the side of the car. The knuckle may also be opened by the lever.

For coupling to occur, the drawbars of the two cars must be aligned, which can be done only manually, by someone going between the cars on the track to move the drawbars. On a straight track, the drawbars must be aligned in the center of the housing mechanism, which acts as a spring to cushion the shock when the cars meet. On a curve, however, the drawbars may need to be moved to one side or the other for coupling to occur. Kavorkian's duties included aligning the drawbars, and on October 16, 1987, he was injured in attempting to center the drawbar of a railroad car.

Two cars had failed to couple because the drawbars bypassed each other; that is, they were not aligned in the center of the housing mechanism and so the knuckles missed each other when the cars came together in an attempt to connect. Kavorkian attempted to center the drawbars so coupling could be accomplished. In pushing with his back and hands on the drawbar, which weighs three to four hundred pounds, Kavorkian injured his back. The railroad car was stationary when Kavorkian attempted to align the drawbar, and the car to be coupled was a safe distance away. Despite the pain in his back, Kavorkian aligned the drawbars and coupling was completed successfully thereafter.

Kavorkian testified at trial that at the time he moved the drawbar he did not see any defect in the coupling mechanism, but he stated that there could have been an unseen defect underneath the car. Kavorkian did not know what caused the bypass, as he was watching the speed of the cars as they approached each other and not the coupling mechanisms, but he testified that the track was straight, the knuckles open, and the drawbars aligned for coupling.

The district court granted Kavorkian's pretrial motion to exclude testimony that the coupling mechanisms were not defective or in bad condition and to preclude CSX from arguing that the drawbar was misaligned due to normal operations and that such a misalignment was not a violation of the FSAA. Similarly, the district court rejected CSX's proposed jury instructions that would have allowed CSX to argue--and the jury to find--that there was no violation of the FSAA inasmuch as the evidence showed that the drawbars were not aligned for coupling to occur.

The district court felt constrained by its reading of the Sixth Circuit's interpretation of the FSAA, although it stated that it disagreed with that interpretation. The court allowed CSX to present evidence in a separate record that drawbars can become misaligned, through normal railroad operations and without any defect, so that they will not couple automatically. Another brakeman, Gus Vitale, testified for the separate record that misaligned drawbars do not necessarily indicate defect, that normal operation can misalign a drawbar through jarring or vibration, and that a worker can misjudge the positions of the drawbars such that coupling does not occur as expected. Other evidence presented in the separate record showed that a long drawbar, such as the one involved in this case, has more lateral play in the housing than a short drawbar, because it is longer; drawbars must move laterally to permit railroad cars to go around curves without derailing. Thus, for example, if two cars are uncoupled on a curve, their drawbars would not necessarily be aligned for subsequent coupling on a straight track. In brief, the evidence in the separate record indicated that the lateral movement of the drawbars means that the drawbars may become aligned improperly for coupling without any defect in the drawbars.

The district court denied CSX's motion for directed verdict, made at the close of Kavorkian's case. The district court reasoned that, under Sixth Circuit law, a failure to couple, even due to misaligned drawbars, is automatically a violation of the FSAA. The district court urged this court to review the issue and adopt a different rule.

The jury found that CSX was not negligent, but that the coupling device violated the FSAA's strict liability provisions and that this violation caused Kavorkian's injuries. The jury determined damages in the amount of $65,000, upon which verdict the district court entered judgment.

CSX's motion for j.n.o.v. or new trial was denied. CSX appeals the judgment of the district court and the denial of its motion for j.n.o.v. or new trial on the issue of liability only under the FSAA. Kavorkian cross-appeals, arguing that the evidence indicates a permanent back injury, pain and suffering, and lost wages, together totalling more than $65,000.

The FELA renders railroads liable for damages resulting from violations of the FSAA. 45 U.S.C. Sec. 51; San Antonio & Aransas Pass Ry. Co. v. Wagner, 241 U.S. 476, 484, 36 S. Ct. 626, 630, 60 L. Ed. 1110 (1916). Section 2 of the FSAA prohibits railroads from using cars that lack "couplers coupling automatically by impact, and which can be uncoupled without the necessity of men going between the ends of the cars." 45 U.S.C. Sec. 2. Congress enacted this provision "in an effort to reduce injuries that were occurring as ... railroad workers were standing between rail cars during the coupling process." Goedel v. Norfolk & Western Ry. Co., 13 F.3d 807, 812 (4th Cir.1994); see also United Transp. Union v. Lewis, 711 F.2d 233, 24548 (D.C.Cir.1983) (describing danger that FSAA addresses).

a failure of equipment to perform as required by the Safety Appliance Act is in itself an actionable wrong, in no way dependent upon negligence and for the proximate results of which there is liability--a liability that cannot be escaped by proof of care or diligence.... "The statutory liability is not based upon the carrier's negligence. The duty imposed is an absolute one, and the carrier is not excused by any showing of care, however assiduous."

O'Donnell v. Elgin, J. & E. Ry. Co., 338 U.S. 384, 390-91, 70 S. Ct. 200, 204, 94 L. Ed. 187 (1949) (quoting Brady v. Terminal Railroad Ass'n of St. Louis, 303 U.S. 10, 15, 58 S. Ct. 426, 429, 82 L. Ed. 614 (1938)). "Once the violation is established, only causal relation is in issue." Carter v. Atlanta & St. Andrews Bay Ry. Co., 338 U.S. 430, 434, 70 S. Ct. 226, 229, 94 L. Ed. 236 (1949). The injury must proximately result from the violation. Id. at 435, 70 S.Ct. at 229.

However, for the railroad to be liable for injuries resulting from the failure to couple, the coupler must have been set properly. Id. at 434, 70 S.Ct. at 229; Affolder v. New York, C. & St. L. R. Co., 339 U.S. 96, 99, 70 S. Ct. 509, 511, 94 L. Ed. 683 (1950) ("Of course [Sec. 2 liability] assumes that the coupler was placed in a position to operate on impact."). Dissenting in Affolder, Justice Jackson summarized the law: "Before a failure to couple establishes a defective coupler, it must be found that it was properly set so it could couple. If it was not adjusted as such automatic couplers must be, of course the failure is not that of the device." Affolder, 339 U.S. at 101, 70 S.Ct. at 512 (Justice Jackson dissented on the issue of whether the jury instructions accurately conveyed the law).

Affolder stated that the issue of whether the coupler was set properly was for the jury to determine; specifically, the factual question there was whether the knuckle of the coupler had been open when coupling was attempted. 339 U.S. at 99, 70 S.Ct. at 511. In Carter, the Court also noted that whether the coupler was set properly is a jury question. 338 U.S. at 434 n. 3, 70 S. Ct. at 229 n. 3. Similarly, Atlantic City Railroad Co. v. Parker, 242 U.S. 56, 59, 37 S. Ct. 69, 70, 61 L. Ed. 150 (1916), suggests that whether a coupler failed to couple because the drawbar was misaligned is a question for the jury. The Court there stated that "a jury would be warranted in finding that a lateral play so great as to prevent coupling [on a straight track] was not needed, and that, in the absence of any explanation believed by them, the failure indicated that the railroad had not fully complied with the law." Ibid . See also San Antonio & Aransas Pass Ry., 241 U.S. at 483-84, 36 S.Ct. at 629 (drawbar misalignment as cause of failure to couple is question for jury). Thus, a key question for the jury under these cases is whether there has been a failure "of the device"; if so, the railroad is liable even if that failure is not due to the railroad's negligence. However, if the railroad can persuade the jury that the coupler was not set properly for coupling--whether the knuckles were closed or drawbars not aligned--then Sec. 2 has not been violated.

Cobb v. Union Ry. Co., 318 F.2d 33 (6th Cir.), cert. denied, 375 U.S. 945, 84 S. Ct. 352, 11 L. Ed. 2d 275 (1963), follows Affolder and the other Supreme Court cases and indicates that the defendant railroad has the burden of proving that the couplers were not set properly at the time that they failed to couple automatically. Id. at 37. Whether the couplers were set properly is a question for the jury. Id. at 36, 37. Cobb, however, does not indicate what constitutes a properly-set coupler.

In Clark v. Kentucky & Indiana Terminal Railroad, 728 F.2d 307 (6th Cir.1984), we stated that a FSAA plaintiff:

makes a sufficient showing of a violation of Sec. 2 by establishing that train couplers failed to couple automatically on the single impact in question or by establishing that cars failed to remain coupled until purposely released once a coupling was effected. If any probative evidence supports finding either of these violations of Sec. 2, a jury verdict of absolute statutory liability is supportable and the railroad is responsible for resulting injuries.

Id. at 310 (citations omitted). In Clark, the evidence indicated that an impact had been intended and supported the jury's finding that there had been an impact but no coupling. Id. at 310-11. The plaintiff returned to the uncoupled cars and found a misaligned coupler with a closed knuckle. He then stepped between the cars and injured himself when, bracing himself against the track, he attempted to push the drawbar into its proper position for coupling. With respect to the railroad's argument that the coupler needs to have been set properly before a violation of Sec. 2 can be found, we stated: "[The railroad] reads Affolder and Cobb to hold that no violation of Sec. 2 can be found unless the couplers were aligned and at least one knuckle open before the failed impact. Assuming this is a proper reading of those cases, the jury was permitted to find that the couplers were properly set on this record." Id. at 311. We held that there was sufficient evidence to support the jury's finding that the couplers had been properly set: testimony indicated that workers inspected the condition of the cars and realigned couplers when necessary, and that the knuckle of one of the couplers was defective, suggesting that a defective knuckle, and not a misaligned coupler, was the cause of the failure to couple. Ibid.

Thus, in Clark, the jury had the opportunity to determine if the couplers had been set properly. We held that the evidence supported a finding by the jury that they were set properly and that one of the knuckles was defective. In other words, the jury could find that there had been a failure "of the device," even though the railroad had an opportunity to persuade the jury that the coupling failed because of an improperly set coupling mechanism. In our case, by contrast, the district court--believing itself bound by an incorrect reading of the law in this circuit (albeit recognizing that that reading was not a correct interpretation of Sec. 2 and was inconsistent with the modern trend in the federal courts of appeals)--did not allow the jury the opportunity to determine if the couplers had been set properly, i.e., at least one knuckle open and the drawbars aligned. Had CSX been allowed to persuade the jury that the couplers had not been set properly, the jury could have found that there was no failure "of the device," the key question from Affolder. On the other hand, as in Carter, the jury could have found that the couplers were set properly, in which case CSX would be held liable for Kavorkian's injuries under the FSAA's strict liability provisions.

However, the Clark court also stated that, even if there were no evidence to support the jury's finding that the couplers had been set properly--which we deemed to mean "aligned and at least one knuckle open"--there was an alternative basis for imposing liability under Sec. 2 on the railroad. Citing an Eighth Circuit case from 1942 and one from 1933, we stated that:

railroads have also been held responsible for injuries suffered by workers who passed between cars to correct a misaligned drawbar preparatory to coupling even though no attempt at coupling had yet been made. In such cases, it was enough that the worker saw the coupler sufficiently out-of-line to make coupling impossible, passed between the cars to straighten the coupler and was injured as the cars came together.

728 F.2d at 312 (citations omitted). "More recent cases have held railroads liable for injuries suffered by workers attempting to straighten misaligned couplers even though the plaintiff's injuries were not caused by the coming together of the cars but simply by the strain of aligning the coupler." Ibid. (citing Kansas City Southern Ry. Co. v. Cagle, 229 F.2d 12 (10th Cir.1955); Schaaf v. Chesapeake & O. Ry. Co., 113 Mich.App. 544, 317 N.W.2d 679 (1982); Buskirk v. Burlington Northern, Inc., 103 Ill.App.3d 414, 59 Ill. Dec. 125, 431 N.E.2d 410 (1982); Coleman v. Burlington Northern, Inc., 681 F.2d 542 (8th Cir.1982)).

Clark, therefore, could be read to limit the Affolder and Cobb requirement of properly-set couplers to mean only that at least one knuckle be open. See, e.g., Shuff v. Consolidated Rail Corp., 818 F. Supp. 219, 222-23 (N.D.Ill.1993). However, we believe that that is an incorrect reading of Clark (and of Affolder and Cobb ). We first note that the language in Clark limiting the requirement of properly set couplers is obiter dicta, raised in a discussion of an alternative basis for imposing FSAA liability in that case. More important, the conclusion in Clark rested upon the fact of railroad workers going between cars to correct for what was perceived as defective equipment--aging and worn drawbars. We stated that:

A coupler typically has a lateral play of one to one and one-half inches on either side that permits movement of coupled cars around curves. But a coupler's lateral motion may be increased by wear.... Unlike a misalignment, the knuckle's position is not attributable to aging equipment.... In order to correct a coupler so out-of-line that a successful coupling is impossible a railroad worker is required to step between the cars and make the adjustment.... [W]e hold that a person injured while attempting to align a coupler in order to effect a coupling may recover through Sec. 2 of the FSAA.

728 F.2d at 312-13. Thus, we allowed the imposition of FSAA liability on the (alternative) basis of moving a drawbar not merely because a worker went between the cars, but because the injury resulted when the worker went between the cars to handle defective equipment.

The problem in the Clark alternative holding, as suggested by the evidence developed in the separate record in the trial court in this case, is that a misaligned drawbar does not necessarily indicate a defect in the equipment. Rather, a drawbar frequently becomes misaligned by the normal jarring and vibrations of the railroad car or when the car is uncoupled on a different track, to such a degree that coupling can not occur without realignment. Clark assumed a drawbar's typical lateral movement to be typically one to one and one-half inches "on either side" and therefore reasoned that workers going between the cars to realign drawbars did so because of defective equipment, which the FSAA prohibits. In other words, the alternative basis for liability in Clark gave effect to the FSAA as an equipment safety statute. This alternative basis does not allow for the possibility of having to realign a non-defective drawbar--i.e., to set the drawbar in a position proper for coupling to occur, as required by Affolder, 339 U.S. at 99, 70 S. Ct. at 511. Although workers must go between the cars to realign even non-defective drawbars, the danger in doing so does not come from the situation against which the FSAA seeks to safeguard (workers having to go between railroad cars moving together); the danger here is outside Sec. 2's contemplated scope. Thus, we read Affolder's requirement of "properly set" to include properly aligned drawbars.

This reading of the requirement of "properly set" enjoys support in other federal courts of appeals. In Reed v. Philadelphia, Bethlehem & New England Railroad Co., 939 F.2d 128 (3d Cir.1991), for example, the Third Circuit addressed the issue of Sec. 2 liability arising out of misaligned drawbars. After examining the conflicting cases on the issue and stressing that the FSAA is an equipment safety statute designed to impose liability for defective equipment, the Third Circuit stated:

We think the rationale of Affolder, which allows a defense of closed knuckles, applies equally to a situation where non-defective equipment is misaligned. In both instances, the causative factor which prevents the coupling may result from jarring or vibration of the car while moving, or from human error in failing to open the knuckle or align the drawbar. In these situations, the failure to couple might not be caused by defective equipment and consequently is not a violation of the Safety Appliance Act.

Id. at 132. The court also noted that the defendant railroad has the burden of proving that the couplers were not aligned properly. Ibid.; see also Maldonado v. Missouri Pacific Ry. Co., 798 F.2d 764, 768 (5th Cir.1986), cert. denied, 480 U.S. 932, 107 S. Ct. 1571, 94 L. Ed. 2d 762 (1987) ("drawbar misalignment should be a defense to failure to couple in instances where the misalignment which causes the noncoupling does not reflect defective or failed equipment"); Lewis, 711 F.2d 233 (Sec. 2 does not impose an independent prohibition against procedures that require workers to go between cars; the FSAA must allow workers to go between cars or else "licit coupling" would be impossible). Thus, in Reed, the defendant railroad had the opportunity to show that a factor other than equipment failure caused the failure to couple. In our case, the district court erred in not giving CSX such an opportunity. On remand, the district court will be able to fashion its jury instructions accordingly.1

The Fourth Circuit recently reached a conclusion similar to the Reed, Maldonado, and Lewis decisions. In Goedel, the court held that "a misaligned coupler, absent a defect in the mechanism, is not a violation of Section Two of the Safety Appliance Act." 13 F.3d at 812. The Fourth Circuit "perceive[d]" the FSAA's requirement that all coupling mechanisms perform automatically "to mean that the coupling knuckles must couple automatically once they are properly in the 'mating' position," which includes proper alignment. Ibid. Examining the legislative history of the FSAA and the mechanics and practical requirements of coupling, the court reasoned that the FSAA does not "require that drawbars be aligned perfectly at all times." Ibid.; see also Lisek v. Norfolk & Ry. Co., 30 F.3d 823, 829 (7th Cir.1994) ("Affolder defense" extends to misaligned drawbars).

Consistent with the reasoning and conclusions in Reed, Maldonado, Lewis, Goedel, and Lisek, we hold that for a drawbar to be "properly set" under Affolder's interpretation of Sec. 2 of the FSAA, so as to impose strict liability, it must be aligned properly and at least one of the knuckles must be open. The defendant railroad has the burden of proving improper alignment, but if it can persuade the jury that the drawbars were non-defective and aligned improperly, then the railroad is not liable under Sec. 2 of the FSAA (at least for injuries that proceed from a failed coupling due to misaligned drawbars). Because the district court in this case did not permit CSX to prove to the jury that the drawbars in question were non-defective and aligned improperly, we remand for a new trial.

The judgment of the district court is REVERSED and this case REMANDED.