Anthony Moretto, an employee of the electrical contractor Delta Electric Inc., was injured when an electrical switch he was helping to operate exploded on November 30, 1989. Moretto and his wife sued the switch's manufacturer, G & W Electric Company, and the owner of the switch, American Cyanamid Company (ACC).1 ACC and G & W sought indemnification or contribution from each other and from Delta, which they brought in as a third-party defendant, and the entire action went to trial in January 1993.

In order to accommodate the plaintiffs' scheduling difficulties, ACC presented its liability witnesses before the close of the plaintiffs' case. When the Morettos rested, various motions were made for judgments as a matter of law. The court first granted G & W's motion, before then granting the Morettos' motion against ACC. Both decisions relied on letters sent by G & W to purchasers of the type of switch that had exploded, with the court apparently finding that G & W had provided adequate warning and that ACC had acted negligently by allowing the switch to be used in an "improper" manner. Id. at 746. ACC's claim against Delta went to the jury, as did the issue of damages. The jury awarded the Morettos $7.4 million, split evenly between compensatory and punitive damages. It also found that between ACC and Delta, ACC was entirely responsible. The Morettos later agreed to reduce the damages to $4.5 million.

ACC appeals, arguing that the district court erred in granting the Morettos' motion for judgment as a matter of law. It asks that we order a new trial on all issues and raises a variety of other errors. Because we agree that the district court erred in granting the Morettos' motion, we need only address that issue, the scope of the new trial to be held on remand, and an issue involving punitive damages that is likely to recur in the next trial.

* This is a rather complicated case, both in terms of the various parties' theories of causation and the sequence of events that led up to the explosion. Our recitation of those facts is not made necessary by the failure of both the Morettos and the district court to describe in detail either the facts or the legal theories on which the motion for judgment as a matter of law, and the granting of that motion, were based. We shall discuss these omissions in greater detail below, but first we shall outline the evidence on liability presented at trial.

* Anthony Moretto was employed as an electrician by Delta, which had been hired by ACC to undertake a two-year project to upgrade the power distribution system at its Lederle Laboratories plant. This project was supervised on a daily basis by Lederle employees, and it involved, in relevant part, the replacement of old cable with new, the addition of new cable, and, consequently, the splicing of old cable to new. One such high-voltage splice had just been completed when Delta's general foreman on this project, Lawrence Pelletier, asked Moretto to accompany a Lederle employee, Ken Baisley, to activate a switch that was connected to the newly spliced cable. Activation of these switches required the participation of employees of both Delta and ACC as each had control over separate locks on the switch. When the switch was turned on, or "closed," it exploded, and both Moretto and Baisley were injured.

It seems that unbeknownst to anyone at the time, Delta had misconnected one of the cables in such a fashion that when the switch was closed, a "fault" or short-circuit was created. Delta was contractually obligated to "megger" test or "ring out" the line, which would have uncovered this mistake, but there is evidence that ACC directed Delta not to undertake that testing on this occasion because of the delay that would result. In any event, Delta concedes both that the mistake was made and that the testing was not done.

If this were the only evidence of causation, we would have a much simpler case, but it is not. In the pleadings and early stages of the trial, the Morettos argued that G & W was liable for this accident in that the switch ACC purchased from it was defective in design and manufacture. The Morettos introduced testimony against G & W that it had been aware of problems with these switches. Specifically, the Morettos relied on testimony from Delta foreman Pelletier that he had spoken with G & W just after the accident and had learned of two warning letters that G & W had sent out to users of these switches in 1983 and 1985. Jt.App. at 239-43. When this testimony was introduced, the court, without the benefit of objection by counsel, directed that this evidence could only come in against G & W "because the only person on the phone was somebody from G & W." Id. at 241.2 Moments later the letters themselves were introduced into evidence without objection. Id. at 243.

The two letters differ in material respects. The first letter, dated August 1, 1983, was sent by G & W's vice president of marketing to all those who had purchased these RA-style, oil-based switches from G & W since the 1920s (the switch in question was thirty-five to forty years old). After first speaking of the wonders that G & W's RA switches had accomplished over the past sixty years, it noted that these switches do not have "an assigned close into fault rating." Id. at 1148. The letter went on to suggest that G & W's switches of more recent design, those with a "spring charged mechanism," do have such a rating and, in addition, require less skill for successful operation on the part of operating personnel. Id. Based on their many years of experience building and servicing such switches, G & W then made several recommendations, including that those who had purchased the RA switches over the past sixty years replace those switches with "modern switches (such as G & W's RA20, RA32, RA40, MR, IFLDR, PMO, LDR) that utilize a spring-assisted operator." Id. The letter also recommends that users establish a maintenance program for their oil-based switches, operate all their existing RA switches with "the correct remote operation tool" (catalog and price list enclosed), and contact their local sales representatives to receive "[c]aution labels" for the existing switches that recommend remote operation of the switches. Id.

The letter of nearly two years later, dated July 10, 1985, came from G & W's president and took a markedly different tack. Rather than beginning with a discussion of the wonders of G & W switches and closing with recommendations that one either buy new switches from G & W or buy the remote operation tool it was now offering, this letter begins by noting that, as a result of the failure of switch users to follow G & W's earlier recommendations, there have been "serious injuries and, in some cases, fatalities." Id. at 1149. The letter then went on to again recommend replacement of the existing switches and elaborated on the need to instruct personnel who operate the existing switches in the proper method of operation. Id. at 1149-50.

Although Pelletier's testimony regarding the letters, and the letters themselves, were only admitted into evidence against G & W (a matter we discuss in greater detail in part II.B.1), evidence was introduced against ACC indicating that it did receive at least one of the letters. The deposition of Edward Roeloffs, a principal electrical engineer in Lederle's plant engineering department, was read into the record. In his testimony, Roeloffs recounts a discussion he had with the former head of Lederle's power department, Ramon Martello. See id. at 1331-33. Roeloffs indicates that he had several discussions with the power department about his understanding that "they were replacing a couple of switches a year." Id. at 1330. He wondered why they were doing this, and concluded that it was because the switches "were old and possibly just unreliable as far as service continuity was concerned." Id. at 1331. He was not "aware there was a safety problem with them," but he knew, because Martello had told him, that the department had received a letter from G & W recommending replacement of the old switches. Id. at 1331-32. Martello "dismissed [the recommendation] at the time by saying, oh well, what they're trying to do is sell more switches." Id. at 1345. Roeloffs was unaware whether Martello had received more than one letter, but indicated that Martello "obviously made reference to at least one of the letters in that conversation," id. at 1346, which took place one to two years prior to the accident. Id. at 1333; see also id. at 1346.

In addition to the evidence that G & W had recommended that ACC replace the existing switches, the Morettos also introduced evidence that ACC had deactivated the circuit breakers that might, if operating correctly, have prevented the explosion. Once again this evidence came in through Pelletier, who indicated he had been told by John Orr, who worked in Lederle's plant engineering department, that the safety trips had been removed from the circuit breakers, which meant that they "would not trip." Id. at 251. Pelletier asked why this "absolutely crazy" course of action would be undertaken, and he was told that Lederle could not afford any power outages or down time because its production was worth too much. Id. Pelletier testified further that the switch exploded because the safety trip from the "upstream" circuit breaker had been removed. Id. at 265, 267.

Pelletier is also the source of the evidence that the so-called "meggar" testing was not done, that such testing would have uncovered the existence of the fault, and that despite Delta's request to conduct such testing (and its apparent obligation to do so), ACC refused to allow it to be done. Id. at 291. The explanation for this alleged refusal is less than clear, but apparently any such testing would have required that the other end of the cable, which was connected to another switch, be disconnected, and ACC did not want that done. Id. at 292-93. Pelletier also explained that such testing would have taken half a day, id. at 302, and one might infer, based on Pelletier's testimony regarding the circuit breakers, that such delays would be unacceptable to ACC.

Pelletier's testimony on these various matters is not uncontested. John Orr testified that he did not tell Pelletier that the safety trips had been removed, and that, in fact, he "would never even know what the status of those trips would be" because he "wasn't involved in the power department" at the time of the accident. Id. at 655. Roeloffs testified that the circuit breaker simply did not open fast enough, id. at 1317, but he was not certain that even if the breaker had opened sooner, that would necessarily have prevented the explosion from occurring. Id. at 1320.

Juan Serna, an electrical engineer who headed Lederle's power department at the time of the accident, also indicated that the circuit breakers in this case simply did not open fast enough to prevent the explosion. Id. at 646. He went further than Roeloffs on the cause of the accident, however, testifying that the misconnection by Delta caused the explosion. Id. at 610. He also testified that "megger" or continuity testing would have uncovered the misconnection, id. at 613, and that he most certainly did not tell Delta that it could not conduct such testing. Id. at 615-16; see also id. at 630-31. In fact, he expected that such testing would be done and a half-day delay would be of no significance given the amount of time Delta had to work on getting the cable in place. Id. at 616.

Serna added that prior to the accident he believed that the RA switch had been designed to be operated under a load, or "while power is flowing through" it. Id. at 621-22. Only after the accident did Serna learn that the switches did not have a "close into fault" rating; if Martello, the former head of the department, knew of this, he never told Serna. Id. at 622-24. Serna agreed that if Martello knew, he "probably" should have told Serna, and that had he known it "probably" would have affected the way in which the switches were operated. Id. at 624. He later admitted that if one of Lederle's employees knew about the necessity of operating the switch only with the remote operating tool, it would have been a "departure from good engineering practice" not to operate it in that manner. Id. at 632.

After the evidence we have just outlined had been presented, G & W moved for judgment as a matter of law. The Morettos agreed that G & W had "established that notice was given to Lederle, and Lederle ha[d] not in any fashion argued that that notice was insufficient to warn them of inherent dangers in this project," and so they concurred in G & W's motion, as did Delta. Id. at 742, 743. ACC objected, arguing that G & W had not taken sufficient measures to assure notice and that perhaps the letter should have been sent by certified mail or a national recall should have been undertaken. Id. at 742. The district court found "no evidence" to support ACC's view and granted the motion. Id.

The Morettos then moved "for a directed verdict on the evidence against Lederle under the [labor] law on the grounds we have expressed previously to the Court, but not on the record." Id. Under any set of circumstances, they argued, ACC was liable under New York's Labor Law, though it might then have some claim for miswiring against Delta. ACC did not ask for any clarification of the motion--that it, for example, comply with Federal Rule of Civil Procedure 50(a)(2) and "specify the ... law and the facts on which the moving party is entitled to the judgment"--but simply argued that the proximate cause of the explosion was "the misconnection of the wires" by Delta. Jt.App. at 744.

The district court found this argument unpersuasive and stated two potentially distinct reasons. The judge first said that if Delta were bankrupt, it seemed to him that ACC would be liable. Id. The court added that ACC might "well have a claim that Delta Electric is liable to you, but it seems to me, in the first [instance], you have got undisputed evidence that Lederle was put on notice that this switch is defective, [and] cannot be used for the purpose for which they used it." Id. at 745. In the court's view, that evidence supported a holding that ACC violated New York's Labor Law. Although, as ACC points out on appeal, the court did not indicate which section of the labor law was implicated, both sections 200 and 241(6) rely on a finding of negligence, at least for direct liability, so the court's failure is of minimal significance.

In response to ACC's further argument that Delta was responsible, the court said that each of them was "liable for the damage that occurred," and that if a third party had been injured (not an employee of either, presumably), "[h]e would have had the right to sue each of you and could probably recover against each of you because the negligence of each of you caused his injury." Id. at 746. When Delta moved for judgment as matter of law, the court simply stated that "[i]f Delta had made the proper connection, it would not have mattered that this switch was used without the remote device." Id. at 748. The court then denied Delta's motion, finding the issue to be one for the jury or for later apportionment. Id. at 749.

We review the granting of a judgment as a matter of law de novo, applying the same standard as does the district court. Sir Speedy, Inc. v. L & P Graphics, Inc., 957 F.2d 1033, 1039 (2d Cir.1992). We may only affirm such a judgment if, "viewed in the light most favorable to the nonmoving party, 'the evidence is such that, without weighing the credibility of the witnesses or otherwise considering the weight of the evidence, there can be but one conclusion as to the verdict that reasonable men could have reached.' " Id. at 1038-39 (quoting Simblest v. Maynard, 427 F.2d 1, 4 (2d Cir.1970)). "If, drawing all reasonable inferences in favor of the nonmoving party and making all credibility assessments in [its] favor, there is sufficient evidence to permit a rational juror to find in [its] favor, we must overturn the directed verdict...." Id. at 1039.

The record is not as clear as we might like on either the factual or legal bases for the motion or for the granting of the motion. Although Rule 50(a)(2) requires, as mentioned above, that such motions be made with specificity, ACC did not complain at the time, so it may not now be heard to complain. See 9 Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure: Civil Sec. 2533, at 582 (1971). This does not fully address our problem, however, for when the factual and legal bases of a decision are not adequately articulated, we are left to speculate on the basis of those decisions. It appears in this case that there are only two possible legal theories that would support the district court's judgment, so our task is not as difficult as it might be, but we do note, if only in passing, that the Third Circuit has exercised its supervisory authority to require that each district court entering a directed verdict "set forth 'an explanation sufficient to permit [the appellate] court to understand the legal premise for the court's order.' " Sowell v. Butcher & Singer, Inc., 926 F.2d 289, 295 (3d Cir.1991) (citation omitted). It did so "[i]n the interest of the parties' ability to present relevant appellate argument, efficient allocation of judicial resources, and effective appellate review." Id. Although we do not choose at this time to follow the Third Circuit's lead, we do note that our task would have been made simpler, and scarce judicial resources would have been conserved, if the basis of the decision had been clearly set forth.

* The first possible basis for the judgment, raised by the district court's hypothetical discussion of Delta's bankruptcy, is that either Delta or ACC is liable, and if it is simply Delta, then ACC is vicariously liable under New York Labor Law section 241(6), so either way ACC faces liability. There is no record evidence that the Morettos made this a basis of their motion, nor do they argue it on appeal. ACC argues that this could not have been the basis of the judgment because the district court later instructed the jury that ACC was directly liable for the explosion. ACC Br. at 30 n. 7; see Jt.App. at 874-75. So, despite the implicit suggestion in the district court's remarks, there is no reason to believe that this served as the basis for the judgment, and, because no party asserts this as a basis for affirmance, we shall not address it.

The second possible basis for the judgment is that ACC had notice that it was improper to use the switch as it did, that ACC introduced no evidence to dispute this point, that ACC is therefore negligent, and that ACC's negligence is a proximate cause of the explosion. Both of the Morettos' causes of action under New York Labor Law require a showing of negligence, at least for purposes of ACC's direct liability. See Zimmer v. Chemung County Performing Arts, Inc., 65 N.Y.2d 513, 523, 493 N.Y.S.2d 102, 482 N.E.2d 898 (1985) ("Neither section 241(6), nor section 200, ... [are] self-executing, but rather, both require reference to outside sources to determine the standard by which a defendant's conduct must be measured."). The Morettos have made this notice argument to the lower court and to this court, and the district court asserted that "undisputed evidence" exists that ACC had such notice. Id. at 745. Our task is to analyze this "undisputed evidence" and to determine whether a juror looking at that evidence could reasonably find for ACC.

* The threshold question presented is, what undisputed evidence do we have before us? The Morettos, G & W, and Delta premise all of their arguments in this regard on the presence of both of the warning letters in the record. They appear to base their argument that both letters were admitted against ACC on the failure of ACC to object at the moment the letters were received into evidence. Such reliance is misplaced, however, when the testimony that served as the foundation for the introduction of these letters was admitted only against G & W. While it would have been helpful for the court or ACC to clarify when these letters were introduced that they were only being received against G & W, it does not follow from the failure to so clarify that the letters were admitted generally. When foundation testimony is expressly admitted only against one party, it follows that the exhibits for which this foundation was laid are also only introduced against that same party, absent a clear statement to the contrary.

The appellees also argue that by failing to object to the district court's statement that undisputed evidence indicated that ACC had notice, ACC is therefore bound. The problem with this argument is that we are reviewing a motion for judgment as a matter of law. We must read the district court's statement against ACC as narrowly as possible. By failing to object to that statement, ACC merely acknowledged that there was testimony in the record to suggest that ACC had received one of the letters. That testimony came from Edward Roeloffs.

Roeloffs's testimony could correctly be labeled "undisputed." Roeloffs testified that the former head of the power department at Lederle, Ramon Martello, told him that he had received a letter from G & W recommending that the switches be replaced, but that he considered the letter nothing more than an attempt to sell more switches. In granting all inferences in ACC's favor, we must assume that Martello was referring only to the least damaging of the two letters, the one sent in 1983. Although the 1983 letter was never formally introduced against ACC, its substance was discussed during Roeloffs's testimony, and we shall therefore consider it, with his testimony, as the "undisputed evidence" of notice referred to by the district court.3

We have finally reached the crux of the matter; that is, given the testimony of Roeloffs and the 1983 letter, must any rational juror find ACC liable for the switch's explosion? An affirmative conclusion to that question would require both a finding that ACC was negligent and that its negligence was a proximate cause of the explosion. Because we find that this is not one of those rare cases in which the evidence merits a finding of negligence as a matter of law, see Andre v. Pomeroy, 35 N.Y.2d 361, 364, 362 N.Y.S.2d 131, 320 N.E.2d 853 (1974), we therefore need not reach the issue of proximate cause.

Roeloffs simply testified that Martello had told him that he had received a letter from G & W recommending that the old switches be replaced. He testified that he was not aware, from this discussion or from any other source, of any safety problems involving these switches. Our analysis must necessarily focus on the text of the letter itself, then, informed by Roeloffs's statement that Martello simply considered this an attempt to sell more switches.

The Morettos would have us focus on one factual statement in a dense one-page letter. As discussed above, the letter began with a description of the successful use G & W's switches had been put to over the past sixty years. At the end of this paragraph, which mentioned that "thousands of these switches have given satisfactory service throughout the United States and around the world," was the following statement: "They did not have an assigned close into fault rating." Jt.App. at 1148. The letter then went on recommend new switches that are easier to operate and that do have a close-into-fault rating, and recommended that the old switches be replaced. Until the switches could be replaced, the letter recommended, inter alia, that they be operated with remote operating tools that could be purchased from G & W.

One could reasonably consider this letter an attempt to sell more switches (or, in the alternative, remote operating tools). In fact, the Morettos' opening statement made just such an argument. See id. at 162-63. Counsel noted that the letter came from the vice president of marketing; that it described the old switches as "hunkydory"; that not one of the 403 words in the letter said, "Do not operate that switch under power"; and that the letter in fact said the opposite, that the switch could be operated under power. Id. Although counsel noted that, as he put it, ten of the 403 words did telegraph a potential problem (the "no assigned close-into-fault rating" language), this was "the only thing they're saying in there that in any way dances around danger." Id. at 163 (emphasis added).

Counsel's opening is, of course, not evidence, but unless the Morettos would have us believe that this opening could not be believed by a rational juror, it does lend credence to our conclusion. The 1985 letter, which is properly before us as evidence against G & W, further indicates G & W's ability to write a letter that makes the point more clearly. The necessity of sending out such a letter, as G & W admittedly did, further indicates that other users received the 1983 letter without jumping to the conclusion that they must change their practices immediately at the risk of serious injury to those operating these switches. The significance of the "no assigned close-into-fault" language in the 1983 letter is simply that G & W had not warranted that these switches would survive being closed into a fault. The switch in this case obviously did not survive such an event, and a jury could reasonably find that this statement was sufficient to put ACC on notice. It need not do so, however, and so we must reverse the district court's decision to grant the Morettos judgment as a matter of law.

To reiterate, the district court erred in the first instance to the extent that it relied on the 1985 letter as "undisputed evidence." Once the 1985 letter is removed from consideration,4 we simply do not believe that a jury would be unreasonable if it found that the 1983 letter failed to place ACC on notice of the dangers of operating the switch under a load. Any other finding of ACC's negligence would necessarily rely on disputed evidence, so we must therefore send this issue back to the district court so that it may be addressed by a jury.

Having determined that a new trial is warranted, we feel it necessary to add a few words of direction regarding that trial; first, on the scope of that trial, and second, on the propriety of instructing the jury on the so-called "complicity rule" in awarding punitive damages against an employer for actions taken by employees.A

ACC has asked that we grant a new trial on all issues. It has named G & W and Delta as appellees (in addition to the Morettos), and it has alleged claims against both for indemnification or contribution. We believe that ACC should be allowed to pursue both of those claims again on retrial, though the Morettos' direct claim against G & W has been effectively waived, both by failure to appeal that judgment and by counsel's virtual stipulation to the judgment when it was requested.

* Delta places great weight on the fact that a jury found it zero-percent liable for the explosion vis-a-vis ACC. While true, that jury verdict was predicated on a direction that ACC was liable for the accident. Once that direction is taken away, we can have no certainty that the jury would reach the same verdict. But for the direction that ACC was liable, the jury might well have found that Delta's miswiring caused the accident. Any argument about the circuit breakers or to the effect that Delta was ordered not to do the testing are of no consequence to us, for as indicated earlier, the record includes conflicting testimony on those issues. As we review the record evidence, we are not at all certain that a jury untainted by the improper direction of the verdict against ACC would necessarily find Delta blameless, and we shall therefore reverse the judgment in that regard as well.

G & W presents a closer question, but that question is controlled by our ruling that if one grants ACC all reasonable inferences, the record only supports a finding that it received the 1983 letter. The issue of G & W's possible liability to ACC turns on the adequacy of the notice given to ACC. G & W may have sent out dozens of explicit warning letters about this switch, but absent record evidence that any of them were received by ACC, they only support the argument that G & W knew of the possible problems with these switches. Although, as we stated above, a jury could reasonably find that the 1983 letter put ACC on notice, it need not do so.

In discussing warnings included with prescription drugs, the New York Court of Appeals recently stated that whether a warning presents a question for the jury "requires a careful analysis of the warning's language. The court must examine not only the meaning and informational content of the language but also its form and manner of expression." Martin v. Hacker, 83 N.Y.2d 1, 10, 607 N.Y.S.2d 598, 598, 628 N.E.2d 1308, 1312 (N.Y.1993). Bearing in mind that the warning is to be read by professionals, "the factors to be considered in resolving this question include whether the warning is accurate, clear, consistent on its face, and whether it portrays with sufficient intensity the risk involved...." Id. The risk involved here was, as the jury's award of damages indicates, quite serious. The ten-word warning in a 403-word letter in no way portrayed with sufficient intensity the risk involved to warrant taking this question from the jury.

The final question involves ACC's assertion that the district court committed plain error by failing to instruct the jury that in order to award punitive damages against an employer for its employee's conduct, it must find that "persons in a position of authority ... in some way authorized, ratified or fostered the acts complained of." Williams v. City of New York, 508 F.2d 356, 361 (2d Cir.1974). See Loughry v. Lincoln First Bank, N.A., 67 N.Y.2d 369, 380, 494 N.E.2d 70, 76, 502 N.Y.S.2d 965, 971 (1986) (requiring "a high level of general managerial authority in relation to the nature and operation of the employer's business"). In Williams we found that the district court's failure to instruct the jury on this "complicity rule" was plain error. 508 F.2d at 362. ACC raised this argument to the district court in its post-verdict motions, and the district court properly rejected the argument, noting that in Williams there was no evidence whatsoever of complicity by persons in positions of authority, while in this case there was evidence that "senior officials" at ACC were involved in making the decisions that led to Morettos' injuries. Jt.App. at 938.

Although we do not believe that it was plain error not to give this instruction, see Parkin v. Cornell University, Inc., 182 A.D.2d 850, 851, 581 N.Y.S.2d 914, appeal dismissed, 80 N.Y.2d 914, 588 N.Y.S.2d 821, 602 N.E.2d 229 (1992), it would nonetheless be appropriate, now that it has been called to the court's attention, to give such an instruction at the next trial.

To reiterate, we find that the district court erred in granting the Morettos and G & W judgments as a matter of law against ACC, and we therefore vacate those judgments. We also find that the erroneous judgment in the Morettos' favor casts sufficient doubt on the jury verdict for Delta that we must vacate that judgment as well. Consequently, all damage awards to the Morettos, both compensatory and punitive, are vacated, and the case is remanded to the district court for a new trial consistent with this opinion.