This is an appeal by Chevron Chemical Company from a judgment rendered against it after a jury trial in a suit brought by the minor children and the estate of Richard Ferebee. Ferebee, an agricultural worker at the Beltsville Agricultural Research Center (BARC), an installation of the United States Department of Agriculture located in Beltsville, Maryland, allegedly contracted pulmonary fibrosis as a result of long-term skin exposure to dilute solutions of paraquat, a herbicide distributed in the United States solely by Chevron. When Ferebee died before trial, his estate continued with a survival action and a wrongful death count was added on behalf of his minor children. After a first trial ended in a mistrial, a second jury returned a verdict on the wrongful death count for $60,000 against Chevron. That verdict was based on the theory that Chevron's failure to label paraquat in a manner which adequately warned that long-term skin exposure to paraquat could cause serious lung disease made Chevron strictly liable for Ferebee's injuries.

After unsuccessfully moving for a directed verdict and a judgment notwithstanding the verdict, 552 F. Supp. 1293, Chevron now asks this court to overturn the jury's verdict. Chevron's appeal offers a plethora of bases upon which it asserts that the jury's verdict was in error. Essentially, these claims divide into two groups, the first of which attacks the jury's verdict as inconsistent with the evidence and the second of which asserts that federal law precludes the tort action upon which Ferebee's children recovered. We find all of Chevron's claims to be without merit and affirm the district court's judgment upholding the jury verdict.

Paraquat is an important agricultural herbicide that has been sold in the United States since 1966. Paraquat is known to be toxic and to cause acute injury if directly absorbed into the body. For this reason, the sale and labelling of paraquat has been extensively regulated since 1966 by the federal government, first by the Department of Agriculture and currently by the Environmental Protection Agency (EPA).

At trial, the jury was presented with a complicated set of facts. In 1967 Richard Ferebee began work as an agricultural worker at BARC. His job frequently required him to spray various chemicals, including insecticides and herbicides, on greenhouse plants and, in the summer, on plants outside in the fields. Mr. Ferebee began spraying paraquat in the summer of 1977. He ordinarily sprayed six or seven times a month for between one and three hours. He used the product regularly during the outdoor growing seasons of 1977, 1978, and 1979.

When Mr. Ferebee sprayed paraquat in the fields, he frequently got the dilute spray on his skin, typically when he used his hands to shield plants while he sprayed weeds growing around them. In addition, in a videotape deposition taken before his death, Mr. Ferebee described two incidents of more extensive exposure to paraquat. The first occurred soon after he began spraying the compound. On that day, Ferebee spent several hours walking behind a tractor that was spraying paraquat. His head and bare arms became drenched with spray. At the end of the day, he began to feel dizzy and exhausted. When he went home, he was too tired to wash or change his clothes and fell asleep instantly. The dizziness and other symptoms did not persist, however, and he later returned to work.

Mr. Ferebee's second major exposure to paraquat also occurred during the 1977 growing season. On that occasion, he was spraying paraquat with a hand-held sprayer for some time when he noticed that the sprayer was defective and had leaked paraquat solution all over his pants. He stopped spraying and cleaned up as much as possible, but was not able to change his clothes until he went home.

Even before 1977, Mr. Ferebee was not a picture of perfect health. He was overweight, suffered from high blood pressure, and had a life-long sinus problem. Nonetheless, in late 1977, according to Mr. Ferebee's testimony, he began to notice a marked change in his physical condition, most notably increasing shortness of breath. Over the next several years, Mr. Ferebee's condition progressively deteriorated. In November of 1979 he checked into Capital Hill Hospital, where Dr. Muhammad Yusuf, a pulmonary specialist, diagnosed Ferebee's disease as pulmonary fibrosis. Dr. Yusuf referred Mr. Ferebee to the National Institutes of Health, where he was treated during 1981 and 1982 by Dr. Ronald G. Crystal, Chief of the Pulmonary Branch of the Heart, Lung, and Blood Institute. After several consultations and tests, both Drs. Yusuf and Crystal concluded that Ferebee's pulmonary fibrosis was caused by paraquat poisoning. Mr. Ferebee's lung condition continued to degenerate, and on March 18, 1982 he died.

In the legal action prosecuted by Ferebee's estate and minor children, appellees presented both of Mr. Ferebee's treating physicians as expert witnesses. Both Dr. Yusuf and Dr. Crystal testified that, in their opinion, paraquat had caused Mr. Ferebee's pulmonary fibrosis. To support this view, they relied not only upon their own observation of Mr. Ferebee and the medical tests performed on him, but also upon medical studies which, they asserted, suggested that dermal absorption of paraquat can lead to chronic lung abnormalities of the sort characterized as pulmonary fibrosis. Appellees then argued to the jury that Chevron had not adequately labelled paraquat to warn against the possibility that chronic skin exposure could lead to lung disease and death and that this failure was a proximate cause of Mr. Ferebee's illness and death. Chevron appeals the jury's verdict which was necessarily based on acceptance of this theory.

We begin by determining the source and content of the cause of action upon which this suit is founded. Ferebee's exposure to paraquat occurred entirely at the Beltsville Agricultural Research Center (BARC). BARC is located on federal land, over which the federal government exercises "exclusive Legislation," U.S. Const. art. I, Sec. 8, cl. 17. Accordingly, any right of appellees to recover for their alleged injuries must initially find its source in either federal common law or federal statutory law. Since 1928, a federal statute has provided just such a right in an act which creates a cause of action for wrongful death occurring on federal property:

In the case of the death of any person by the neglect or wrongful act of another within a national park or other place subject to the exclusive jurisdiction of the United States, within the exterior boundaries of any state, such right of action shall exist as though the place were under the jurisdiction of the State within whose exterior boundaries such place may be; and in any action brought to recover on account of injuries sustained in any such place the rights of the parties shall be governed by the laws of the State within the exterior boundaries of which it may be.

The substantive elements of the cause of action created by the federal wrongful death act are thus defined by reference to state law. There is no dispute in this case that the relevant state law by which the federal cause of action must be defined is that of Maryland. There is a dispute, however, as to the content of that law. At the time of Ferebee's injury, Maryland law recognized a cause of action in strict products liability for failure to warn adequately of a product's inherently dangerous condition; it was according to this law that the jury was instructed. At the time BARC became federal property in 1938, in contrast, Maryland tort law recognized a cause of action only for negligence. Chevron argues that the federal statute incorporates only that state law which was in effect at the time the federal government acquired the relevant property and that, given the state of Maryland law in 1938, it was error to allow plaintiffs, who had waived all claims sounding in negligence, to take their case to the jury on a theory of strict products liability.

We reject Chevron's view, which amounts to the position that Section 457 requires the wrongful death law on federal enclaves to be frozen at the date of cession. "The natural reading of the statutory language is that the wrongful-death law of a federal enclave should be identical to that of the surrounding state, whatever that law might be and however it might change over time." Vasina v. Grumman Corp., 644 F.2d 112, 117 (2d Cir.1981). In addition to the plain language of the statute, its history and logic belie the argument that Congress intended a static, rather than a dynamic, incorporation of state law. The federal act was passed in an effort to put tort victims on federal land on an equal footing with those injured just outside the boundaries of the federal enclave, see 69 CONG.REC. 1486 (1928) (statement of Senator Walsh); it is implausible to ascribe to a Congress so intended a contrary aim to saddle tort victims on federal land with the consequences of outdated legislation. We have no trouble agreeing with the numerous lower courts which have considered this question: Congress intended to allow the federal action to evolve concomitantly with changes in state law. See, e.g., Mathis v. General Electric Corp., 580 F.2d 192, 194 (5th Cir.1978) (implicitly interpreting Sec. 457 to apply current state law); Greene v. Vantage Steamship Corp., 466 F.2d 159, 166 n. 9 (4th Cir.1972) (same); Muniz v. United States, 280 F. Supp. 542, 546 (S.D.N.Y.1968) (same). We hold that it is the state law in effect at the time of the injury that governs in suits brought under Sec. 457.

Under that law as interpreted by the trial judge, appellees had the burden of proving, by a preponderance of the evidence, the following elements:

(1) That paraquat proximately caused Mr. Ferebee's illness and death;

(2) That paraquat is inherently dangerous;

(3) That Chevron knew, or should have known, at the time it sold the paraquat used by Mr. Ferebee, that the chemical was inherently dangerous;

(4) That the resulting duty to provide an adequate warning of the danger was not met;

(5) That the inadequacy of the warning proximately caused Mr. Ferebee's illness and death.

Chevron does not dispute that this is a correct statement of the elements necessary to recover under Maryland strict liability principles in a failure-to-warn suit. Nonetheless, Chevron seeks to overturn the jury's verdict on the theory that, with the exception of the second element, the jury could not reasonably have inferred from the conflicting testimony the existence of any of these elements. A review of the conflicting testimony with respect to each of these elements makes it clear that Chevron cannot sustain the heavy burden it must carry to overturn a jury's verdict. We also reject Chevron's legal claim that other federal law preempts this tort action.

An appellate court's function in reviewing the denial of a judgment notwithstanding the verdict is very limited. "The jury's verdict must stand unless 'the evidence, together with all inferences that can reasonably be drawn therefrom is so one-sided that reasonable men could not disagree on the verdict.' " Carter v. Duncan-Huggins, Ltd., 727 F.2d 1225 at 1227 (D.C.Cir.1984) (cited cases omitted). The appellate court does not assess witness credibility nor weigh the evidence, but rather seeks to verify only that fair-minded jurors could reach the verdict rendered. Coburn v. Pan American World Airways, Inc., 711 F.2d 339, 342 (D.C.Cir.1983). Chevron's burden on this appeal is thus "extremely heavy." Boutros v. Riggs National Bank, 655 F.2d 1257, 1260 (D.C.Cir.1981).

These admonitions apply with special force in the context of the present action, in which an admittedly dangerous chemical is alleged through long-term exposure to have caused disease. Judges, both trial and appellate, have no special competence to resolve the complex and refractory causal issues raised by the attempt to link low-level exposure to toxic chemicals with human disease. On questions such as these, which stand at the frontier of current medical and epidemiological inquiry, if experts are willing to testify that such a link exists, it is for the jury to decide whether to credit such testimony. Against this background, we turn to Chevron's claims that appellees failed to establish the elements necessary to sustain a jury verdict in their favor.

Chevron first argues that the jury was obligated to reject appellee's theory that long-term exposure to paraquat caused Ferebee's illness and death. Chevron acknowledges that paraquat is known to be toxic, but argues that it is only acutely toxic--that is, that any injuries resulting from exposure to paraquat occur within a very short time of exposure, such as days or weeks, and that when exposure ceases, so too does the injury. In this case, Ferebee did not experience any of the symptoms of pulmonary fibrosis until late 1978, at which point it had been ten months since he last sprayed paraquat, and his chronic inflammatory lung disease continued to worsen long after his final use of paraquat in August of 1979. Plaintiffs' theory of recovery was thus that paraquat, when absorbed through the skin, can attack the lungs in such a way as to cause chronic and self-perpetuating inflammation. Chevron argues that there has never been any evidence nor any suggestion that paraquat can cause chronic injury of this sort and that, in any event, Ferebee could not have been exposed to enough paraquat to injure him in this fashion.

The short answer to Chevron's argument is that two expert witnesses refuted it and that the jury was entitled to believe those experts. Both Drs. Crystal and Yusuf, who are eminent specialists in pulmonary medicine and who were Ferebee's treating physicians, testified that paraquat poisoning was the cause of Ferebee's illness and death. Both admitted that cases like Ferebee's were rare, but Dr. Crystal identified three other cases he felt were similar to that of Mr. Ferebee. Chevron argues that these cases can be distinguished from Mr. Ferebee's, but it is not our role to decide the merits of Chevron's attempted distinctions; Dr. Crystal thought the cases were similar, and the jury was entitled to believe him. Chevron of course introduced its own experts who were of the view that Ferebee's illness was not caused by paraquat, but the testimony of those witnesses, who did not treat Mr. Ferebee or examine him, can hardly be deemed so substantial that the jury had no choice but to accept it. The experts on both sides relied on essentially the same diagnostic methodology; they differed solely on the conclusions they drew from test results and other information. The case was thus a classic battle of the experts, a battle in which the jury must decide the victor. Jenkins v. United States, 307 F.2d 637, 646 (D.C.Cir.1962).

Chevron seeks to avoid this conclusion by asserting that expert opinion testimony must be generally accepted in the scientific community before it can be introduced as evidence and that the views of Drs. Crystal and Yusuf, while not rejected by the medical community, are sufficiently novel at this point as to be inadmissible. As support for this proposition, Chevron cites the Maryland Court of Appeal's statement in the criminal case of Reed v. State, 283 Md. 374, 381, 391 A.2d 364, 368 (1978): "before a scientific opinion will be received as evidence at trial, the basis of that opinion must be shown to be generally accepted as reliable within the expert's particular scientific field." The court in Reed, however, was referring to the introduction of evidence based on novel scientific techniques or methodologies, see id. 391 A.2d at 367-68, and carefully distinguished the test applied to such evidence from that involved in the admission of scientific opinion testimony that, while controversial in its conclusions, is based on well-founded methodologies. ("The question of the reliability of a scientific technique or process is unlike the question, for example, of the helpfulness of particular expert testimony to the trier of facts in a specific case."). Id. Thus, a cause-effect relationship need not be clearly established by animal or epidemiological studies before a doctor can testify that, in his opinion, such a relationship exists. As long as the basic methodology employed to reach such a conclusion is sound, such as use of tissue samples, standard tests, and patient examination, products liability law does not preclude recovery until a "statistically significant" number of people have been injured or until science has had the time and resources to complete sophisticated laboratory studies of the chemical. In a courtroom, the test for allowing a plaintiff to recover in a tort suit of this type is not scientific certainty but legal sufficiency; if reasonable jurors could conclude from the expert testimony that paraquat more likely than not caused Ferebee's injury, the fact that another jury might reach the opposite conclusion or that science would require more evidence before conclusively considering the causation question resolved is irrelevant. That Ferebee's case may have been the first of its exact type, or that his doctors may have been the first alert enough to recognize such a case, does not mean that the testimony of those doctors, who are concededly well qualified in their fields, should not have been admitted.

Finally, Chevron argues that its expert, Dr. Fisher, proved that it was physically impossible for Ferebee to have been exposed to enough paraquat to cause him any injury. Dr. Fisher's testimony, however, went only to the amount of paraquat necessary to cause the short-term illnesses that have long been recognized to follow from paraquat exposure. Accepting Dr. Fisher's testimony as true, as plaintiffs did not at trial, it still would not necessarily follow that the same amount is needed to trigger a chronic case like that which Ferebee allegedly contracted. The jury could therefore have concluded that Ferebee had been exposed to sufficient amounts of paraquat to cause the chronic disease from which he suffered--even if that exposure was not substantial enough to produce acute symptoms. The dose-response relationship at low levels of exposure for admittedly toxic chemicals like paraquat is one of the most sharply contested questions currently being debated in the medical community, see generally Leape, Quantitative Risk Assessment in Regulation of Environmental Carcinogens, 4 HARVARD ENVT'L L. REV. 86, 100-103 (1980); surely it would be rash for a court to declare as a matter of law that, below a certain threshold level of exposure, dermal absorption of paraquat has no detrimental effect. We therefore conclude that there was sufficient evidence of causation to justify submission of that issue to the jury.

Chevron next argues that, even if Ferebee's pulmonary fibrosis was caused by paraquat exposure, Chevron had no knowledge until this case that long-term dermal absorption of paraquat could induce pulmonary fibrosis; accordingly, Chevron claims, it cannot have been charged with a duty to warn against a danger that was not foreseeable at the time Ferebee was exposed to paraquat. Chevron is correct that its duty to warn is limited to dangers that it knew or should have known about during the time Ferebee was exposed to paraquat. See Chambers v. G.D. Searle & Co., 441 F. Supp. 377, 381 (D.Md.1975), aff'd, 567 F.2d 269 (4th Cir.1977). The conclusion which Chevron would have us draw from this principle, however, is not correct. The jury could have found that, as of July 1979, the last date on which Ferebee sprayed paraquat, Chevron's knowledge of the link between dermal paraquat exposure and lung disease was sufficient to require a more detailed label than that which the company provided.

In assessing Chevron's argument, it is necessary to consider the label which Chevron did provide. That label contains several warnings about the danger of skin exposure to paraquat. For example, the label in large bold letters states:

The label also states that, in case of skin contact, the area exposed should be washed immediately and that contaminated clothing may have to be removed. The label further notes that "prolonged contact" will cause severe irritation and that "repeated contact" may increase the danger of absorption.

The label thus does inform the user that spilling paraquat on the skin may cause immediate and perhaps severe skin irritation. Yet nowhere does the label persuasively suggest that users whose skin comes into contact with the herbicide should be concerned about other possible consequences of skin exposure--particularly the specter of long-term lung disease culminating, perhaps, in death. (The label does note that repeated contact may increase the danger of absorption, but there is no hint of the dangers that may attend absorption). The jury certainly could have concluded that, absent a more specific warning about the relationship between skin exposure and lung disease, the average reader of the label who got paraquat onto his skin would presume that, once no immediate and acute injury occurred, the user was safe and need not take further precautions--such as seeing a doctor, having x-rays taken, or ceasing use of the pesticide.

Moreover, there can be no doubt that Chevron did by 1979 have sufficient information regarding the general link between dermal paraquat exposure and lung disease that the company could have been charged with a duty to provide this more specific warning. First, Chevron does not dispute the fact that, since the early 1960's, it has been known that paraquat exposure can lead to fibrotic lung disease. Second, expert witnesses for both sides agreed at trial that, once paraquat enters the body, it selectively attacks the lungs. Finally, the medical literature and the company's own incident reports catalogued cases in which dermal exposure to paraquat in some cases caused almost immediate death and in other cases caused rather immediate lung problems. For example, by 1976 Chevron had learned of an incident in Israel in which a man was accidentally sprayed with a jet of paraquat on his face, nose, mouth and hands; the man recovered after showing some signs of pulmonary involvement. In April of 1976 Chevron learned of a New Guinea man who had his back painted with paraquat and died two days later. That same year, Chevron was informed of a 29-year-old worker from Malaysia who spilled paraquat on his lap and scrotal area and died ten days later. These are but a few of the examples submitted to the jury of cases in which dermal paraquat exposure caused serious injury, including lung disease and death.

Chevron seeks to distinguish these incidents and to establish compliance with its duty to warn by arguing that these incidents all involved immediate internal complications resulting from dermal paraquat exposure, whereas Ferebee suffered from delayed and prolonged pulmonary fibrosis. Chevron argues that it had no information that prolonged exposure to paraquat could cause a chronic illness like Ferebee's or that such illness could continue long after exposure to paraquat had ceased. But the fact that the injuries of which Chevron knew occurred much more quickly than the prolonged illness through which Ferebee suffered is no answer to Chevron's complete failure to warn that any such injuries, whether immediate or latent, could result from dermal exposure to paraquat. Had Chevron warned that dermal exposure could induce lung disease, the company conceivably might not have had the more focused duty to warn against the particular form and nature of the disease that Ferebee contracted. Absent a general warning about the relationship between such exposure and lung complications, a relationship about which the jury could reasonably have concluded Chevron had substantial knowledge, Chevron cannot escape its duty to warn by asserting that it had insufficient knowledge of the particular way in which Mr. Ferebee's lung disease came about. The label gave no indication at all that skin exposure could produce lung disease of any sort, a failing upon which the jury could ground liability.

Chevron next argues that, if Ferebee indeed died from chronic paraquat exposure of his skin, his case is so unique as to make him a hypersensitive plaintiff against whose injuries Chevron is not obligated to guard, see Chambers, supra. Once again, however, Chevron has attempted to define its duty at too specific a level. If paraquat caused Mr. Ferebee's illness, and the jury could reasonably conclude that it did, there is nothing to suggest that another individual in Ferebee's position would not also have been likely to contract the disease. The jury was entitled to find it reasonably foreseeable that illness of the general type from which Mr. Ferebee suffered--fibrotic lung disease leading to death--could be caused by dermal exposure to paraquat. That cases exactly like Mr. Ferebee's, involving dilute solutions of the herbicide and chronic rather than acute injury, had not previously been called to Chevron's attention does not attenuate the company's duty to warn against the general danger of lung disease and death attributable to dermal exposure to paraquat. See generally Billiar v. Minnesota Mining and Manufacturing Co., 623 F.2d 240, 246 (2d Cir.1980) ("If the injury is reasonably foreseeable, however, even if rare, the seller cannot rely on its history of good fortune to exempt itself from liability."). As a matter of law, we cannot hold it so unforeseeable that dermal paraquat exposure could lead to serious lung disease and death that Chevron was insulated from any duty to warn against such a possibility.

Chevron also claims that all cases of dermal paraquat exposure producing serious injury were accompanied by skin lesions which were present either before or after the injury. Ferebee's skin was admittedly intact before exposure and he testified in deposition that he never experienced a rash or any form of cracking or burning sensation. It is unclear whether Chevron raises this point to negate causation or foreseeability, but in either case, the point is unpersuasive. If the argument goes to causation, it is again answered by credible expert testimony that paraquat did cause Mr. Ferebee's injuries. If Chevron's argument instead is that it had no notice that paraquat exposure to intact skin could lead to serious injury, that argument is belied by the numerous cases in which paraquat absorption through previously intact skin led to nearly immediate death. See supra pp. 1537-1538. Accordingly, Chevron cannot, by pointing to the fact that Ferebee's skin was intact at the time of exposure, avoid the jury's conclusion that the company knew or should have known that dermal exposure to paraquat could cause serious lung disease or death.

D. Inadequate Labelling as the Proximate Cause of Mr. Ferebee's Pulmonary Fibrosis

Given that Chevron had a duty to warn that dermal paraquat exposure could produce serious lung disease and death, Mr. Ferebee was still required to show that Chevron's failure to perform that duty was a proximate cause of his illness and death. On appeal Chevron argues that the evidence overwhelmingly suggests that Mr. Ferebee did not read the label that was provided and that a more detailed label thus would have done nothing to prevent Ferebee's injuries.

It is true that the plaintiff's failure to read a warning actually provided may at times absolve a manufacturer of liability. See, e.g., Johnson v. Niagra Machines & Tool Works, 666 F.2d 1223, 1225 (8th Cir.1981). It is also true in this case that the district court expressed "some difficulty in agreeing that a rational jury could have found either that Mr. Ferebee read the label or that the label was not adequately designed to attract attention." We need not reexamine the evidence to decide this issue, however, for we hold that satisfaction of the proximate cause requirement does not depend entirely on whether Mr. Ferebee read the label. Instead, if the jury could reasonably have found that the information on an adequately labelled paraquat bottle would have been communicated to Mr. Ferebee--even if he personally did not read the warning--the failure to provide such a warning could validly be treated as a proximate cause of Mr. Ferebee's pulmonary fibrosis.

We live in an organizational society in which traditional common-law limitations on an actor's duty must give way to the realities of society. See, e.g., MacPherson v. Buick Motor Co., 217 N.Y. 382, 111 N.E. 1050 (1916) (eliminating "privity" requirement for tort suits). In this case, Mr. Ferebee did not purchase the paraquat for his personal use, and there was substantial evidence that workplace communication about the dangers associated with various chemicals usually took the form of oral instructions from supervisors to workers, the latter of whom then retransmitted the information to co-workers. This, rather than individual reading of product warnings, is a typical method by which information is disseminated in the modern workplace. See Schwartz & Driver, Warnings in the Workplace: The Need for a Synthesis of Law and Communication Theory, 52 U.CINN.L.REV. 38, 66-83 (1983). The requirement that an improper warning proximately "cause" the injury should be elaborated against this background. We believe Maryland would construe its tort law in this case to require only that someone in the workplace have read the label, not that Mr. Ferebee personally have read it. Because there is no dispute that one or more employees at BARC did read the label, we hold that the jury could properly have inferred that, had a warning about the danger of disease from dermal exposure been included on the label, that warning would have been communicated to Mr. Ferebee and that he would as a result have acted differently. Alternatively, the jury could have inferred that an adequate warning would have led Ferebee's employers to undertake steps that would have protected him from paraquat poisoning--for example, provision of showers for use after spraying.

Finally, Ferebee's alleged failure to read the label does not constitute misuse as a matter of law, see Levin v. Walter Kidde & Co., 251 Md. 560, 248 A.2d 151 (1968). While a manufacturer is entitled to assume that warnings given will be heeded, see RESTATEMENT (SECOND) OF TORTS Sec. 402(A) comment j, in this case Chevron failed to warn Ferebee adequately that dermal exposure to paraquat could seriously injure him. As a result, Ferebee could not have known the results that might follow from his failure to heed a warning that was not given; he therefore cannot be charged with misuse.

In sum, the narrow standard of review of a district court's failure to grant a judgment notwithstanding the verdict provides a very small window for appellant through which it has failed to pass. Sufficient evidence supports the jury's finding that plaintiffs established all the elements of their case by a preponderance of the evidence. We turn now to appellant's claim that the jury verdict should nonetheless be overturned because other federal law demands that Chevron be immune from tort liability of the sort levied against it in this case.

Chevron contends that, because paraquat is sold in the United States only when accompanied by a label approved by the federal EPA, a state jury is not allowed in a tort suit to find that label inadequate. Chevron's contention on this point appears to be twofold: first, that EPA approval of the label requires a jury to find that label adequate, and second, that federal law preempts state common law actions against Chevron which are based on the theory that paraquat was inadequately labelled. We reject both of these contentions.

Under the Federal Insecticide, Fungicide, and Rodenticide Act, 7 U.S.C. Sec. 136 et seq. (1982) ("FIFRA"), EPA extensively regulates the sale and labelling of paraquat. The statute precludes EPA from authorizing the sale of paraquat unless the product, as labelled, will not cause "unreasonable adverse effects on the environment." 7 U.S.C. Sec. 136a(c)(5)(C). Such effects are in turn defined as

any unreasonable risk to man or the environment, taking into account the economic, social, and environmental costs and benefits of the use of [the] pesticide.

7 U.S.C. Sec. 136(bb). The Act further requires that the label be "adequate to protect health and the environment" and that it be "likely to be read and understood ...." 7 U.S.C. Sec. 136(q)(1)(E).

After extensive scientific testimony, EPA approved the sale of paraquat with the label which is at issue in this case. According to Chevron, this approval constitutes an expert, federal determination that paraquat as labelled does not pose an unreasonable risk to the normal user--a determination that a jury is not authorized to question. Because the trial court, applying state law, instructed the jury that it could impose liability only if paraquat was "unreasonably dangerous" and the label "inadequate," Chevron concludes that the verdict must be reversed.

Chevron's argument misunderstands the nature of the determination made by EPA and misconceives the relation between federal and state law. The fact that EPA has determined that Chevron's label is adequate for purposes of FIFRA does not compel a jury to find that the label is also adequate for purposes of state tort law as well. The purposes of FIFRA and those of state tort law may be quite distinct. FIFRA aims at ensuring that, from a cost-benefit point of view, paraquat as labelled does not produce "unreasonable adverse effects on the environment." State tort law, in contrast, may have broader compensatory goals; conceivably, a label may be inadequate under state law if that label, while sufficient under a cost-benefit standard, nonetheless fails to warn against any significant risk. In addition, even if the ultimate purposes of federal and state law in this area are the same, a state (acting through its jurors) may assign distinct weight to the elements which go into determining whether a substance as labelled is of sufficient net benefit as to warrant its use. To approve use and sale of paraquat, for example, EPA was required to assess inter alia the health risks from use of paraquat, to assign a cost value to those risks, and to estimate the benefit to society at large from use of the chemical. Unless Congress intended to preempt states from considering these issues, a question we address below, there is no reason a state need strike the same balance on these difficult questions as EPA. Assignment of values to such "soft" variables as human health is among the most difficult tasks faced in a regulatory society, see NATIONAL ACADEMY OF SCIENCES, DECISION MAKING FOR REGULATING CHEMICALS IN THE ENVIRONMENT 41 (1975), and a state may choose to tip the scales more heavily in favor of the health of its citizens than EPA is permitted to by FIFRA. Similarly, a state heavily dependent on agriculture may consider use of paraquat more of a benefit than a state whose economy is primarily industrial, and both states may therefore reach a conclusion different from EPA's regarding the net benefit of paraquat use. Unless FIFRA preempts a state from making these choices, a state jury may find a product inadequately labeled despite EPA's determination that, for purposes of FIFRA, the label is adequate. EPA's determination may be taken into account by the jury, and the jury was instructed in this case that it was permitted to do so, but absent preemption the jury need not give that determination conclusive weight.

That brings us to the heart of Chevron's preemption claim: that FIFRA does preempt states from reconsidering such questions and that state tort suits of the sort at issue here are completely preempted by the Act. Chevron's position is grounded upon a section of FIFRA which provides that a state "shall not impose or continue in effect any requirements for labeling ... in addition to or different from those required under this subchapter." 7 U.S.C. Sec. 136v(b). Chevron argues that a damage action based on the inadequacy of a label has a regulatory aim--to assure that adequate labels are used--and that it is precisely this regulatory aim that FIFRA explicitly preempts.

Damage actions typically, however, can have both regulatory and compensatory aims. Moreover, these aims can be distinct; it need not be the case, as Chevron apparently assumes, that the company can be held liable for failure to warn only if the company could actually have altered its warning. (In any event, as we discuss below, Chevron can take steps to alter its label). In this case, a Maryland jury found that the EPA-approved label did not sufficiently guard against certain injuries. Even if Chevron could not alter the label, Maryland could decide that, as between a manufacturer and an injured party, the manufacturer ought to bear the cost of compensating for those injuries that could have been prevented with a more detailed label than that approved by the EPA. That is, Maryland can be conceived of as having decided that, if it must abide by EPA's determination that a label is adequate, Maryland will nonetheless require manufacturers to bear the risk of any injuries that could have been prevented had Maryland been allowed to require a more detailed label or had Chevron persuaded EPA that a more comprehensive label was needed. The verdict itself does not command Chevron to alter its label--the verdict merely tells Chevron that, if it chooses to continue selling paraquat in Maryland, it may have to compensate for some of the resulting injuries. That may in some sense impose a burden on the sale of paraquat in Maryland, but it is not equivalent to a direct regulatory command that Chevron change its label. Chevron can comply with both federal and state law by continuing to use the EPA-approved label and by simultaneously paying damages to successful tort plaintiffs such as Mr. Ferebee.

Imposition of such a dual obligation upon a manufacturer is permissible under the Act. While FIFRA does not allow states directly to impose additional labelling requirements, the Act clearly allows states to impose more stringent constraints on the use of EPA-approved pesticides than those imposed by the EPA: "A State may regulate the sale or use of any federally registered pesticide or device in the State, but only if and to the extent the regulation does not permit any sale or use prohibited by this subchapter." 7 U.S.C. Sec. 136v(a). See also SEN.REP. NO. 838 92d Cong., 2d Sess. 30 (1982) reprinted in 1972 U.S.CODE CONG. & ADMIN.NEWS 4021 ("Generally, the intent of the provision is to leave to the States the authority to impose stricter regulation on pesticides uses than that required under the Act."); SEN.REP. NO. 970, 92d Cong., 2d Sess. 44 (1972) reprinted in 1972 U.S.CODE CONG. & ADMIN.NEWS 4128 (same); see generally National Agricultural Chemicals Association v. Rominger, 500 F. Supp. 465 (E.D.Cal.1978) (state may require additional data on EPA-registered pesticides). Given this provision, Maryland might well have the power to ban paraquat entirely. We need not decide that issue, however, to hold that, if a state chooses to restrict pesticide use by requiring that the manufacturer compensate for all injuries or for some of these injuries resulting from use of a pesticide, federal law stands as no barrier. The fact that Congress has authorized this form of state action also disposes of any argument that such state tort remedies impose undue burdens on interstate commerce. Southern Pacific Co. v. Arizona, 325 U.S. 761, 769, 65 S. Ct. 1515, 1520, 89 L. Ed. 1915 (1945). As a result, Maryland is entitled to control the use of paraquat for compensatory aims by holding Chevron liable for injuries that could have been prevented by a more adequate label.

Moreover, tort recovery in a case such as this one may also promote legitimate regulatory aims. By encouraging plaintiffs to bring suit for injuries not previously recognized as traceable to pesticides such as paraquat, a state tort action of the kind under review may aid in the exposure of new dangers associated with pesticides. Successful actions of this sort may lead manufacturers to petition EPA to allow more detailed labelling of their products; alternatively, EPA itself may decide that revised labels are required in light of the new information that has been brought to its attention through common law suits. In addition, the specter of damage actions may provide manufacturers with added dynamic incentives to continue to keep abreast of all possible injuries stemming from use of their product so as to forestall such actions through product improvement. That Maryland cannot directly order a change in the way in which paraquat is labelled thus does not deprive the state of legitimate aims which it is entitled to further through the imposition of traditional tort liability.

Moreover, in assessing Chevron's preemption argument and in defining the scope of the Act's preemption provision, it is necessary to bear in mind, as Chevron forgets, the circumspect view courts must take of a claim that Congress has preempted states from exercising their traditional police powers on behalf of their citizens. The provision of tort remedies to compensate for personal injuries "is a subject matter of the kind [the] Court has traditionally regarded as properly within the scope of state superintendence," Florida Lime & Avocado Growers, Inc. v. Paul, 373 U.S. 132, 144, 83 S. Ct. 1210, 1218, 10 L. Ed. 2d 248 (1963). In such a case, we have been admonished to "start with the assumption that the historic police powers of the States were not to be superseded by the Federal Act unless that was the clear and manifest purpose of Congress." Rice v. Santa Fe Elevator Corp., 331 U.S. 218, 230, 67 S. Ct. 1146, 1152, 91 L. Ed. 1447 (1947). That clear and manifest purpose will exist when Congress has explicitly preempted the precise and specific state action, when compliance with both federal and state law would be impossible, or when the state's law stands as an obstacle to accomplishment of the full purposes and objectives of Congress. Silkwood v. Kerr-McGee Corp., --- U.S. ----, 104 S. Ct. 615, 78 L. Ed. 2d 443 (1984).

As we have indicated, none of these conditions is present in the instant case. First, Congress has not explicitly preempted state damage actions; it has merely precluded states from directly ordering changes in the EPA-approved labels. As many state courts have recognized, in general the "mere compliance with [federal or state] regulatory labeling requirements does not preclude a [jury from] finding that additional warnings should have been given." Burch v. Amsterdam Corp., 366 A.2d 1079, 1086 (D.C.1976) (compliance with labelling requirements of Federal Hazardous Substances Act does not immunize manufacturer from liability for defective warning); see also, e.g., Stevens v. Parke, Davis & Co., 9 Cal. 3d 51, 65, 107 Cal. Rptr. 45, 53, 507 P.2d 653, 661 (1973) (compliance with warnings required by Food and Drug Administration not sufficient to immunize manufacturer from liability); Blasing v. P.R.L. Hardenbergh Co., 303 Minn. 41, 49, 226 N.W.2d 110, 115 (1975) (compliance with federal and local labelling requirements does not preclude finding of negligence); Maize v. Atlantic Refining Co., 352 Pa. 51, 56, 41 A.2d 850, 853 (1945) (label approved by Surgeon General not adequate as a matter of law); see generally Restatement (Second) of Torts Sec. 288c (1965) (Generally, "[c]ompliance with a legislative enactment or an administrative regulation does not prevent a finding of negligence where a reasonable man would take additional precautions."). Given this general background of interpretative decisions against which Congress acted, and given the clear authority of states to regulate the "use" of paraquat, the Act cannot be said to express with the requisite clarity an intent to bar damage actions based on the inadequacy of an EPA-approved label.

Second, compliance with both federal and state law cannot be said to be impossible: Chevron can continue to use the EPA-approved label and can at the same time pay damages to successful tort plaintiffs such as Mr. Ferebee; alternatively, Chevron can petition the EPA to allow the label to be made more comprehensive.

Third, state damages actions of the sort at issue here do not stand as an obstacle to the accomplishment of FIFRA's purposes. Such a conflict would exist only if FIFRA were viewed not as a regulatory statute aimed at protecting citizens from the hazards of modern pesticides, but rather as an affirmative subsidization of the pesticide industry that commanded states to accept the use of EPA-registered pesticides. That interpretation of FIFRA, however, is precluded by both the explicit savings clause at 7 U.S.C. Sec. 136v(b) and by the entire legislative history of the Act. Of equal importance, federal legislation has traditionally occupied a limited role as the floor of safe conduct; before transforming such legislation into a ceiling on the ability of states to protect their citizens, and thereby radically adjusting the historic federal-state balance, United States v. Bass, 404 U.S. 336, 349, 92 S. Ct. 515, 523, 30 L. Ed. 2d 488 (1971), courts should wait for a clear statement of congressional intent to work such an alteration. The Supreme Court has often counselled such hesitance. Thus, in Nader v. Allegheny Airlines, 426 U.S. 290, 96 S. Ct. 1978, 48 L. Ed. 2d 643 (1975), the Court held that, even were the Civil Aeronautics Board to find that an action was not a "deceptive" one within the meaning of the Federal Aviation Act of 1958, a state jury remained entitled to find that action fraudulent. That the action was permissible under federal law did not mean, in the absence of a clear congressional intent to promote the action, that states were required to tolerate it. See also Silkwood v. Kerr-McGee Corp., --- U.S. ----, 104 S. Ct. 615, 78 L. Ed. 2d 443 (1984). So too in this case the fact that Congress has chosen to allow the states to regulate the use of pesticides approved by the EPA means that states retain the lesser power to control the use of such pesticides by requiring that at least some of the resulting injuries be compensated. In response, Chevron perhaps will choose not to send paraquat into Maryland; perhaps the company will distribute additional information on paraquat to Maryland users; or Chevron may petition the EPA to be allowed to use a more detailed label. What Chevron cannot do, however, is to force states, under the purported aegis of a statute aimed at protecting against the hazards of modern pesticides, to accept the use of paraquat and to tolerate uncompensated injuries to that state's citizens. Congress has not expressed a "clear and manifest purpose" to achieve such a result; on the contrary, protection of pesticide users and victims by both federal and state law lies at the center of the Act's design. Accordingly, we affirm the decision of the district court and allow the jury's verdict to stand.

It is so ordered.