RODOWSKY, Judge.
Presented here are product liability claims arising out of the deaths of three individuals from mesothelioma, as a result of inhaling asbestos fibers. A consolidated trial in the Circuit Court for Baltimore City resulted in jury verdicts against the petitioners based in legal theory on strict liability and negligence and based factually on failure to warn. The jury awarded punitive damages in addition to compensatory damages. Judgments on the verdicts were affirmed by the Court of Special Appeals. ACandS, Inc. v. Asner, 104 Md.App. 608, 657 A.2d 379 (1995). We granted the writ of certiorari, primarily to review issues concerning the exclusion of evidence relating to threshold limit values (TLVs) and relating to substantial factor causation, as well as the issue concerning the sufficiency of the evidence to support punitive damages. For the reasons set forth below we shall reverse and remand for a new trial on liability for compensatory damages only.
Respondent Ida Sara Masket Asner is the widow and personal representative of Zalma Asner (Asner) who worked as an outside machinist at Bethlehem Steel’s Key Highway Shipyard from 1941 to 1982. Asner died on December 6,1988, the year in which he was diagnosed with mesothelioma. Respondent Mary Matilda Wilson is the widow and personal representative of Charles F. Wilson (Wilson) who died on [162]*162August 23, 1992, of mesothelioma that had been diagnosed that year. Wilson worked from 1946 to 1975 as a supervisor of the tin shop at Allegany Ballistics Laboratory in Pinto, Maryland. Respondent Jean Payne is the widow of Milton W. Payne (Payne). Jean Payne and respondent Harriet Hess are the personal representatives of Payne’s estate. Payne was an electrician at the Fairfield Shipyard in Baltimore during World War II. He died on March 24, 1992, of mesothelioma that had been diagnosed in January of that year. We shall refer to Asner, Wilson, and Payne collectively as “Claimants.”
All defendants in the consolidated actions either settled, filed for protection under the Bankruptcy Code, or were dismissed prior to trial, with the exceptions of the petitioners, ACandS, Inc. (ACandS) and Porter Hayden Company, Inc. (PH). ACandS is an insulation supplier/installation contractor. It was incorporated in November 1957 under the name Armstrong Contracting and Supply Co. as a wholly owned subsidiary of Armstrong Cork Co., later known as Armstrong World Industries. The parent corporation spun off the subsidiary in 1969 when the latter’s name was changed to ACandS. Petitioner PH is also an insulation supplier/installation contractor. It is the surviving corporation in the 1966 or 1967 merger of Reid-Hayden, Inc., a Maryland corporation, and H.W. Porter & Company, Inc., a New Jersey corporation.
In the consolidated action ACandS and PH obtained judgments on cross-claims against four cross-claim defendants, none of whom appealed. The determination of joint liability on the part of the cross-claim defendants resulted in a diminution of the amounts awarded in the jury verdicts for compensatory damages. As so adjusted, the judgments entered by the trial court against ACandS and PH were as follows:
In the Asner case $842,887.88 in compensatory damages against both defendants and $250,000 in punitive damages against each defendant;
In the Wilson case $717,168.40 in compensatory damages against both defendants and $250,000 in punitive damages against each defendant; and
[163]*163In the Payne case $637,366.12 in compensatory damages against both defendants and $250,000 in punitive damages against PH only.
The facts necessary for an understanding of the questions presented for our review will be stated in the parts of this opinion dealing with the respective issues. The issues are:
I. Did the trial court erroneously and prejudicially exclude evidence relating to TLVs?;
II. Did the trial court erroneously and prejudicially exclude evidence of exposure of the Claimants to the asbestos products of non-parties?;
III. In the Payne case, did the trial court erroneously exclude evidence offered by PH to contradict the plaintiffs’ theory of PH’s exclusive distributorship for a certain manufacturer?;
IV. In the Wilson case, was there sufficient evidence to support finding that Wilson’s exposure to ACandS’s products was a substantial factor in causing Wilson’s illness?; and
V. Was there sufficient evidence to support punitive damages?
I
Prior to trial the plaintiffs moved in limine for an order excluding evidence of TLVs. According to a proffer at trial by defendants, TLVs are values that had been set for a large number of injurious, occupational substances by the American Conference of Governmental Industrial Hygienists (ACGIH). The concept is that, if the measurement of the injurious substance in the workplace is below the threshold value, an individual would be working in an adequately controlled environment, but if the measurement is above that value there would be an increased risk of developing disease. In their memorandum supporting the motion in limine, the plaintiffs argued that a TLV for asbestos was irrelevant to the plaintiffs’ claims based on strict liability in tort. They argued:
[164]*164“[T]he law imputes to manufacturers knowledge of the harmful character of their products regardless of their actual knowledge. Thus, TLVs and other evidence related thereto is irrelevant to the central issue under a strict liability analysis: whether the product itself was unreasonably dangerous and defective.”
(Citation omitted). The circuit court granted the plaintiffs’ motion by a longhand notation on the filing.
On the fifth day of trial, prior to plaintiffs’ calling one of the medical witnesses, Dr. Herbert Abrams, defendants sought to clarify that the court’s ruling in limine was limited to plaintiffs’ strict liability theory of the case. The court, however, stated that the prior ruling “applies to the entire case,” and that evidence of TLVs would “not be permitted under any circumstances.”
The circuit court submitted the case to the jury on special interrogatories, and the jury found the defendants liable for failure to warn under both the negligence and the strict liability theories. The Court of Special Appeals, in affirming the exclusion of TLVs, viewed the defendants’ contention in terms of the sufficiency of the plaintiffs’ evidence. That court’s opinion first reviewed the evidence that related to punitive damages before considering the exclusion of TLV evidence. 104 Md.App. at 619-37, 657 A.2d at 385-94. Believing that the evidence relevant to punitive damages was legally sufficient to show that the defendants in fact had known that asbestos was unreasonably dangerous, the Court of Special Appeals concluded that it was unnecessary for the plaintiffs also to show what the defendants should have known. Id. at 638-39, 657 A.2d at 394. That analysis missed the mark. The error complained of lies in excluding from the jury’s consideration evidence that is relevant because it tends to counter or rebut plaintiffs’ evidence as to negligence, strict liability, and punitive damages.
A
TLV evidence is an important subset of state of the art evidence concerning asbestos-containing products and consum[165]*165er health. See Owens-Corning Fiberglas Corp. v. Garrett, 343 Md. 500, 549, 682 A.2d 1143, 1167 (1996); ACandS, Inc. v. Godwin, 340 Md. 334, 365-67, 667 A.2d 116, 130-32 (1995).
“State of the art includes all of the available knowledge on a subject at a given time, and this includes scientific, medical, engineering, and any other knowledge that may be available. State of the art includes the element of time: What is known and when was this knowledge available.”
Lohrmann v. Pittsburgh Corning Corp., 782 F.2d 1156, 1164 (4th Cir.1986).
State of the art is not synonymous with industry standards.
“Industry standards are the practices common to a given industry. They are often set forth in some type of code, such as a building code or electrical code, or they may be adopted by the trade organization of a given industry. State of the art is a higher standard because scientific knowledge expands much more rapidly than industry can assimilate the knowledge and adopt it as a standard.”
Id.
The proffered TLV evidence, reviewed in Part I.B, infra, was clearly relevant to the plaintiffs’ negligence theory, inasmuch as the proffer at least tended to prove industry standards. Indeed, plaintiffs do not argue to the contrary.
The role of state of the art evidence under Maryland law in strict liability, failure to warn cases was explained in Owens-Illinois, Inc. v. Zenobia, 325 Md. 420, 432-38, 601 A.2d 633, 638-41, reh’g denied, 325 Md. 665, 602 A.2d 1182 (1992). Acknowledging some division of opinion in the authorities, we recognized in Zenobia that a majority of courts hold, expressly or implicitly, “that a manufacturer of a product, which is defective only because of the lack of an adequate warning, is not liable when the failure to warn resulted from an absence of knowledge of the dangerous quality of that product.” Id. at 433, 601 A.2d at 639. But, “the required knowledge can be [166]*166established by evidence that the dangerous quality of the product should have been known by a manufacturer because it was known in the scientific or expert' community.” Id. “[E]vidence concerning the presence or absence of knowledge in the expert community is called ‘state of the art’ evidence.” Id. at 435, 601 A.2d at 640 (emphasis added).
We further explained in Zenobia that our adoption of the rule of strict liability in tort, as articulated in § 402A of the Restatement (Second) of Torts (1965), had included Comment j and its knowledge component. Id. at 436, 601 A.2d at 641. Comment j states that, where consumers either do not know of the presence of a dangerous ingredient in a product, or do not know of the danger of a known ingredient, “the seller is required to give warning against it, if he has knowledge, or by the application of reasonable, developed human skill and foresight should have knowledge, of the presence of the ingredient and the danger.” We ended our discussion in Zenobia by leaving open the question of “whether the knowledge component is an element of the plaintiffs case or an affirmative defense.... ” Id. at 438 n. 8, 601 A.2d at 641 n. 8.
Lohrmann, supra, was an asbestosis case governed by Maryland law. The Fourth Circuit there concluded that in Phipps v. General Motors Corp., 278 Md. 337, 363 A.2d 955 (1976), this Court had included Official Comment j in the adoption of strict liability in tort as expressed in § 402A of Restatement (Second) of Torts. Lohrmann, 782 F.2d at 1164-65. Zenobia confirmed that conclusion. 325 Md. at 436-37, 601 A.2d at 641.
In Lohrmann the plaintiff-appellants contended that the trial court erred by including consideration of state of the art evidence in the instructions on strict liability. 782 F.2d at 1164. The Fourth Circuit affirmed. It reasoned, based on Comment j to § 402A, “that in Maryland, state of the art can be considered in a strict liability tort case where the claimed defect is a failure to warn.” Id. The court explained the nexus as follows:
[167]*167“State of the art in the present case would determine the manufacturer’s duty to warn of the danger of asbestos to persons such as Lohrmann, who did not work directly with asbestos, but worked in close proximity to insulators using asbestos-containing products. The question was when the state of the art included knowledge that people such as Lohrmann were at risk because they were working near insulators and others using asbestos-containing products. There were many state-of-the-art witnesses called, and each tried to answer the question of what was known and when it was known.”
This Court in Zenobia favorably reiterated Lohrmann’s holding “that in a strict liability failure to warn case, ‘state of the art’ is relevant with regard to the defendant’s liability.” 325 Md. at 436, 601 A.2d at 640. The Zenobia holding on the effect of Comment j under Maryland law was presaged, as to asbestos products liability, in Aumiller v. Raybestos-Manhattan, Inc., Asbestos Litig. Rep. (Andrews Pub., Inc.) 6338 (U.S.D.C.D.Md. Mar. 11, 1983), aff'd, Asbestos Litig. Rep. (Andrews Pub., Inc.) 9452 (4th Cir. Dec. 21, 1984); Luby v. Johns-Manville Sales Corp., Asbestos Litig. Rep. (Andrews Pub., Inc.) 6692 (Cir. Ct., Balto. City May 27, 1983) (Greenfeld, J.).
Two of the elements of a § 402A cause of action are that the product was in a defective condition when it left the possession or control of the seller and that the product was unreasonably dangerous to the consumer. § 402A(1). In the instant matter TLV evidence was relevant whether or not the defendants knew what the experts knew. TLV evidence was relevant to the issue of when experts in the field of occupational health knew that the asbestos products to which the plaintiffs were exposed were dangerous to persons who were not direct and regular users of the products. The later in time that the jury placed the date when the experts knew, the more that finding ordinarily would operate to reduce, or possibly eliminate, the duration of a claimant’s exposure after unrea[168]*168sonable dangerousness to bystanders was known. “Consequently, in a failure to warn case governed by the Restatement § 402A and Comment j, negligence concepts to some extent have been grafted onto strict liability.” Zenobia, 325 Md. at 435, 601 A.2d at 640. Otherwise, strict liability would be absolute liability, or that of an insurer.
The instant plaintiffs filed their motion in limine as to TLVs in October 1993 and supported their contention principally by citing Kisor v. Johns-Manville Corp., 783 F.2d 1337 (9th Cir.1986). That was a diversity case applying Washington law. Id. at 1340. Kisor held that, when the plaintiffs claim was solely in strict liability, an asbestos manufacturer was not “entitled to introduce evidence of then current medical knowledge to rebut ‘state of the art’ testimony presented by Kisor in her case-in-chief.” Id. at 1340 n. 7. We cited Kisor and rejected its analysis in Zenobia, decided in February 1992. 325 Md. at 435-38, 601 A.2d at 640-41.
In this Court the plaintiffs contend that TLV evidence was irrelevant under the facts because the defendants did not prove that the workplace environments at which their insulators, and the bystander plaintiffs, were exposed were within the TLV. The requirements for admissibility are not so stringent. “[F]or purposes of the ‘should have knowledge’ component of [C]omment j, a manufacturer of a product is held to the knowledge of an expert in the field.”1 Id. at 437, 601 A.2d at 641. On the issue of when warnings should have been affixed to the asbestos products sold by the defendants in order to render those products not unreasonably dangerous, TLV evidence bears on what was known to the experts and when. The relevance of that evidence is not dependent on the defendant’s having taken asbestos dust measurements. In addition, for the court to admit only the plaintiffs’ state of the art evidence would leave the jury with an incomplete picture of what was known to the experts and when.
[169]*169B
The Court of Special Appeals also agreed with the plaintiffs’ contention that any error in the exclusion of TLV evidence was not prejudicial to the defendants, because substantially the same proof that the defendants proffered was in evidence before the jury. ACandS, Inc. v. Asner, 104 Md. App. at 639-41, 657 A.2d at 395-96. The record does not support that conclusion.
Prior to the plaintiffs’ direct examination of Dr. Abrams, and after the court had made explicit that its in limine ruling applied under all circumstances, the defendants proffered what they anticipated, based on prior depositions, to prove through cross-examination of Dr. Abrams. The proffer included the Dreessen Report of 1938 placing the TLV for asbestos at five million particles per cubic foot (5 MPPCF); that the ACGIH adopted that TLV in the 1940s and retained that value until the late 1960s or early 1970s; that the United States Government in 1953 utilized that TLV under the Walsh-Healey Public Contracts Act; and that it was adopted in California in the late 1940s or early 1950s.
Later in the trial, the defendants presented the videotape deposition of Dr. H. Corwin Hinshaw. Pursuant to the court’s prior ruling, the portion of that deposition dealing with TLVs was not presented to the jury. The defendants proffered Dr. Hinshaw’s excluded testimony by means of the typewritten transcript of the videotape deposition. The proffer included that Dr. Dreessen’s report was made for the United States Public Health Service, that his suggested TLV of 5 MPPCF was for total dust in textile mills, that when the ACGIH adopted the asbestos TLV of 5 MPPCF it was for particles of asbestos and not total dust as Dreessen had proposed, that this standard was retained until approximately 1968, and that the 5 MPPCF TLV for asbestos was adopted in a “number” of states and in eighteen foreign countries.
A third proffer, made by the defendants just prior to plaintiffs’ direct examination of Dr. Barry Castleman, included [170]*170some but not all of what was proffered to be proven through Dr. Abrams.
Plaintiffs have directed us to considerable evidence that was before the jury concerning health studies over the years, but the only evidence of governmental recognition and attempted enforcement of a TLV for asbestos came in the cross-examination of Dr. Abrams. He had been the head of the Bureau of Adult Health in California. Dr. Abrams referred to Dreessen’s having “suggested” 5 MPPCF as a tentative level, and he put the suggestion in the following context:
“I, myself, administered these levels in California, and we always thought, well, it’s simply a guideline, it’s not a mathematically precise figure, you see. And we simply tried to get the factories and all to reduce down to that level or below, and to reduce it to nothing, if possible, you see, because we knew that any amount might be dangerous.”
(Emphasis added).
The jury never heard that the 5 MPPCF TLV for asbestos was recommended throughout most of the relevant period by the national association of public health officers. The jury never heard that the TLV for asbestos of 5 MPPCF was recognized by a law of the United States with respect to federal contracts and by the laws of a “number” of states of the United States and of foreign countries. There is a considerable difference between Dr. Abrams’s nod in California to Dreessen’s suggestion, while Dr. Abrams was trying to achieve zero asbestos contamination, and the much more comprehensive recognition of the 5 MPPCF TLV embraced in the proffers. The defendants were prejudiced by the court’s exclusion of the proffered evidence, and we therefore reverse on the liability issues.
II
Prior to trial the circuit court granted another motion in limine by the plaintiffs and excluded evidence of the Claimants’ exposures to the asbestos-containing products of manufacturers other than those who were parties at the time of [171]*171trial. Defendants attack, and plaintiffs defend, this ruling. Because of our holding in Part I, supra, it is not necessary that we resolve these contentions on the present record. Nevertheless, we shall express some general views, both for guidance on retrial in the instant action and because the arguments of the parties address the scope of our decision in Eagle-Picher Indus., Inc. v. Balbos, 326 Md. 179, 604 A.2d 445 (1992), with respect to causation in asbestos exposure cases.
Balbos rejected rules for determining causation that respectively lie near opposite ends of a causation continuum. At the defense extreme we rejected a “but for” rule under which there would be no liability based on substantial factor exposure to a particular defendant’s product if the plaintiff would have suffered the disease even without that exposure. Id. at 208, 604 A.2d at 459. In so doing we quoted favorably from Restatement (Second) of Torts § 431 the following rule:
“ ‘The actor’s negligent conduct is a legal cause of harm to another if
(a) his conduct is a substantial factor in bringing about the harm, and
(b) there is no rule of law relieving the actor from liability because of the manner in which his negligence has resulted in the harm.’ ”
Id. at 208-09, 604 A.2d at 459.
Balbos also rejected, at the plaintiffs end of the spectrum, the “fiber drift theory,” as therein defined. Id. at 216-17, 604 A.2d at 463. Instead, for asbestos exposure claims of bystanders we adopted a multiple factor test that is shorthandedly described as the frequency, proximity, and regularity test. Id. at 210-13, 604 A.2d at 460-61.2
[172]*172In the case before us plaintiffs expressly recognized in their motion in limine that this Court in Balbos “was not presented with the question at issue.” To support their motion in limine, plaintiffs relied primarily on two decisions applying Illinois law. The first was Tragarz v. Keene Corp., 980 F.2d 411 (7th Cir.1992), where the court said:
“[Defendants’] argument implies that the so-called, substantial factor test is a comparative test in which the jury assesses all contributing causes and determines which ones are substantial. This seems to misstate the nature of the substantial factor inquiry under Illinois law because the Illinois courts in applying the substantial factor test do not seem concerned with which of the many contributing causes are most substantial. Rather, they seem concerned with whether each contributing cause, standing alone, is a substantial factor in causing the alleged injury.”
Id. at 424.
In Tragarz, the Seventh Circuit recognized the problems that would result from adopting a comparative analysis. Id. at 425. If a plaintiff were exposed to the asbestos-containing products of only one defendant, and could prove that that exposure alone was sufficient to cause mesothelioma, the plaintiff would be able to prove that but for that exposure he would not have been injured. Id. If, on the other hand, a plaintiff were exposed to many asbestos products, the same exposure might not be considered substantial when compared to the exposures to other products. Id. The Tragarz court reasoned that this anomalous result would not serve the [173]*173purpose of the substantial factor test and would create inequities.
The second case relied upon by plaintiffs in their motion was Kochan v. Owens-Corning Fiberglass Corp., 242 Ill.App.3d 781, 182 Ill.Dec. 814, 610 N.E.2d 683 (1993), cert. denied, 510 U.S. 1177, 114 S.Ct. 1219, 127 L.Ed.2d 565 (1994). Plaintiffs argued that Rochan stands for the proposition that the “frequency, regularity, and proximity” test adopted by this Court in Balbos provides sufficient protection to a defendant’s interests in an asbestos exposure case.
In both cases from Illinois on which plaintiffs relied, the courts assumed that a defendant’s purpose in offering evidence of exposure to the products of non-parties was to argue that the party-defendant should be exonerated if the exposure to the products of one or more non-parties was more substantial than the exposure to that party-defendant’s product. Tragarz, 980 F.2d at 424-25; Kochan, 182 Ill.Dec. at 819-20, 610 N.E.2d at 688-89.
In their opposition to the motion in limine, the defendants argued that the Balbos concept of substantial factor causation made relevant evidence of a Claimant’s exposure to the products of non-parties. But the defendants did not make particularly plain what the purpose of the evidence in question might be, if it were not for the kind of comparison assumed in the cases from Illinois.3
The circuit court granted the motion in limine without hearing oral argument or giving the reasons for its decision. The Court of Special Appeals affirmed, viewing evidence of [174]*174the type in question as irrelevant to the issue of whether exposure to a particular defendant’s product was a substantial cause of a plaintiffs illness. ACandS, Inc. v. Asner, 104 Md.App. at 648, 657 A.2d at 399. The Court of Special Appeals noted that its holding
“does not lessen the plaintiffs burden of proving that the defendant’s product was a substantial cause of his injury. Moreover, the defendant can still negate liability by showing that the plaintiff was not exposed to its products, that any exposure was insufficient to cause injury, or that its products contained an insufficient amount of asbestos to cause injury.”
Id. at 648, 657 A.2d at 399. Nor did the court’s holding apply to cross-claims. Id. at 649, 657 A.2d at 399.
Whether evidence of exposure to the asbestos-containing products of non-parties is relevant is controlled by the purpose for which such evidence is being offered. Such evidence is not per se irrelevant. Consequently, it would be a rare case in which a court could impose a blanket ban on such evidence in advance of trial, inasmuch as the evidentiary setting in which the evidence would be offered ordinarily would be unknown.
Defendants first argue that evidence of exposure to non-parties’ products is necessary to avoid having the jury “make its causation determination in a vacuum, without the means to determine how ‘substantial’ a cause” defendants’ products were in producing plaintiffs’ injuries. Brief of Appellant PH at 12. In general, if a defendant’s purpose were only to show that a claimant’s exposure to the product of a non-party was greater than exposure to the defendant’s product, without any further demonstration, exclusion of the defendant’s evidence ordinarily would be proper. In the typical case the fact of other substantial exposures is not exclusively “the means to determine how ‘substantial’ a cause” a defendant’s product might have been. In the case before us the defendants’ first argument essentially is premised on a comparative fault standard, i.e., on the theory that one defendant [175]*175may be exculpated simply because another defendant more substantially contributed to a plaintiffs disease. This is the argument that we rejected in Balbos when we said:
“[N]o supplier enjoys a causation defense solely on the ground that the plaintiff would probably have suffered the same disease from inhaling fibers originating from the products of other [identified] suppliers.”
326 Md. at 209, 604 A.2d at 459 (emphasis added). In a footnote to the above text, we quoted with approval the following from W. Keeton, Prosser & Keeton on the Law of Torts § 41, at 268 (5th ed. 1984): “ ‘When the conduct of two or more actors ... is a but-for cause of the event, and application of the but-for rule to them individually would absolve all of them, the conduct of each is a cause in fact of the event.’ ” Balbos, 326 Md. at 209 n. 11, 604 A.2d at 459 n. 11.
Defendants’ second argument is that they “should be able to show the jury that other factors were so dominant in causing the injuries at issue that the defendants’] own role in causation was so trivial as to not constitute a substantial factor.” Brief of Appellant PH at 13. We interpret the thrust of the argument not to be simply that a plaintiffs exposure to the products of others was greater than to that of the defendant, but to be that the exposure to a defendant’s product was not a substantial factor.
The theoretical basis for the defendants’ argument is recognized in Restatement (Second) of Torts where a distinction is made between conduct which is a substantial factor in causing harm and conduct the effect of which is so negligible that it is not a substantial or legal cause of the plaintiffs harm. For example, § 431, which we cited favorably in Balbos, presents the following discussion in Comment b:
“In many cases the question before the court is whether the actor’s negligence was in fact the cause of the other’s harm—that is, whether it had any effect in producing it—or whether it was the result of some other cause, the testimony making it clear that it must be one or the other, and that the harm is not due to the combined effects of both. In such [176]*176a case, the question, whether the defendant’s negligence has a substantial as distinguished from a merely negligible effect in bringing about the plaintiffs harm, does not arise if the testimony clearly proves that the harm is from a cause other than the actor’s negligence. Indeed, the testimony often makes it clear that, if the defendant’s conduct had any effect, the effect was substantial. It is only where the evidence permits a reasonable finding that the defendant’s conduct had some effect that the question whether the effect was substantial rather than negligible becomes important.”
(Emphasis added). See Balbos, 326 Md. at 208-09, 604 A.2d at 459.
In this Court defendants rely principally on § 433, dealing with “Considerations Important in Determining Whether Negligent Conduct is [a] Substantial Factor in Producing Harm.” Under subsection (a) a court is to consider “the number of other factors which contribute in producing the harm and the extent of the effect which they have in producing it.” Defendants particularly emphasize Comment d to § 433 which reads:
“There are frequently a number of events each of which is not only a necessary antecedent to the other’s harm, but is also recognizable as having an appreciable effect in bringing it about. Of these the actor’s conduct is only one. Some other event which is a contributing factor in producing the harm may have such a predominant effect in bringing it about as to make the effect of the actor’s negligence insignificant and, therefore, to prevent it from being a substantial factor. So too, although no one of the contributing factors may have such a predominant effect, their combined effect may, as it were, so dilute the effects of the actor’s negligence as to prevent it from being a substantial factor.”
We have no disagreement at the theoretical level of the law of torts generally regarding the portions of the Restatement on which the defendants rely. A factual defense may be based on the negligible effect of a claimant’s exposure to the defendant’s product, or on the negligible effect of the [177]*177asbestos content of a defendant’s product, or both. In such a case the degree of exposure to a non-party’s product and the extent of the asbestos content of the non-party’s product may be relevant to demonstrating the non-substantial nature of the exposure to, or of the asbestos content of, the defendant’s product.4 But, a defendant would not ordinarily generate a jury issue on lack of substantial factor causation only by showing the dangerousness of a non-party’s product to which the claimant was exposed. Ordinarily a defendant would have to follow up the evidence of exposure to the products of non-parties with evidence tending to prove that the defendant’s product was not unreasonably dangerous or was not a substantial causal factor. Under these circumstances the proposition that the defendant’s product is not a substantial cause may be made more probable by evidence tending to prove that the claimant’s disease was caused by the products of one or more non-parties. See, e.g., Becker v. Baron Bros., 138 N.J. 145, 649 A.2d 613 (1994) (whether processed chrysotile in brake products posed a risk of causing mesothelioma in users was a sharply disputed issue of fact at trial, so that trial court erred in instructing as a matter of law that the products were defective without a warning).
In the instant matter the circuit court refused to permit the defendants to elicit from product identification witnesses proffered evidence of the exposure of Asner to the asbestos products of non-parties, and the circuit court instructed the jury to disregard any evidence relating to the exposure of the Claimants to the products of non-parties. Assuming, arguendo, that these rulings were erroneous, whether they were prejudicial under the guidelines that we set forth above depends upon the record as a whole. Inasmuch as the parties have not briefed the factual aspects of their legal contentions as both bear on the issue of prejudice on this record, we do not decide that issue, even for guidance on retrial.
[178]*178Ill
In this Part III we address PH’s argument that, in the Payne case, the court erroneously excluded three purchase requisitions from the Fairfield Shipyard directly to JohnsManville (JM). The issue will undoubtedly recur, because it is part of the ongoing dispute between PH and its adversaries in these cases over the nature of PH’s “exclusive distributorship” for JM. See ACandS, Inc. v. Godwin, 340 Md. at 356-58, 667 A.2d at 126-27; Balbos, 326 Md. at 214-15, 604 A.2d at 462. The scenario is that the plaintiffs offer evidence of exposure to JM products and that PH was the “exclusive distributor” for JM in the Baltimore area. Plaintiffs then argue that the JM products to which the Claimant was exposed had to have been supplied by PH. PH, on the other hand, argues that JM would sell directly, so that the presence of JM products at the workplace does not prove that PH was the supplier. In prior cases, the plaintiffs have relied oh certain deposition testimony of Charles Holterman. Here, the plaintiffs, through at least three other witnesses, produced evidence that PH used only JM products when supplying materials and performing labor under its insulation contracts, that there was no other distributor of JM products in the Baltimore region, and that PH was the largest JM distributor on the eastern seaboard.
To counter this evidence PH offered the three purchase requisitions which the trial court excluded as irrelevant. It reasoned that the evidence of Payne’s exposure to JM products related to pipe covering, block, and cement of the type used in shipyards, whereas the subject of the requisitions dealt with a different type of product. With respect to two of the requisitions, the court acted within its discretion, inasmuch as there was evidence that the product ordered in these two requisitions was a refractory product, and PH did not offer to prove that the product ordered was used in shipyards.
The third requisition, however, was for 100 pounds of asbes[179]*179tos cement and should have been admitted.5
IV
ACandS contends that its motion for judgment in the Wilson case should have been granted for lack of evidence permitting a finding by a preponderance that Wilson’s exposure to any ACandS’s product was frequent, proximate, and regular. The argument rests on a somewhat grudging reading of the testimony of the product identification witness, John Lohr (Lohr). Set forth below is the evidence most favorable to the party who obtained the verdict.
The place of exposure is Allegany Ballistics Laboratory (ABL) where Wilson worked from 1946 to 1975 as a supervisor in the tin shop. Lohr worked in the tin shop from 1956 to 1958 when he transferred to the pipe shop where he worked until his retirement in 1983. Lohr described working with asbestos cement as a pipefitter.6 He said that they used “Armstrong,” ie., ACandS, and “Mansfield” cement.7 Between the two sources the product usage was approximately equal, according to Lohr. In the early 1970s ACandS discontinued supplying any asbestos-containing products. Thus, un[180]*180der the evidence, the period of potential exposure to ACandS cement by Wilson was approximately fourteen years or more.
The workplace, ABL, is a 1,200 acre site in Pinto, Allegany County. The facility is owned by the United States Navy and, apparently, it is operated by Hercules Powder Company for the manufacture of explosive powder, missiles, and motors. When Lohr came to work at ABL there were 560 people employed there. That number grew to possibly 2,500 employees during Lohr’s tenure.
The buddings at ABL are heated by steam. In 1956 there were four boilers at ABL. As the facility expanded, the number grew to fourteen boilers in seven boiler rooms by 1983. These heating plants eventually served almost 1,000 buildings for heat, and, one may infer, for hot water as well. Pipes ran in the open air from the various boilers to the buildings respectively serviced by the boilers. All of the boilers and all of the pipes were insulated with asbestos insulation. As Lohr put it, “[Tjhere’s no bare pipes at ABL.”
New construction at ABL was done by outside contractors and not by employees of ABL who were engaged in various trades. It appears that the employees of the tin, pipe, and other shops primarily were engaged in maintenance and repair. Wilson, as sheetmetal supervisor, apparently moved throughout ABL supervising the work of crews of “tinners,” but Wilson did not personally perform the work tasks of the tinners.
Lohr described insulating the boilers by the use of block, covered with cement, and built up in layers. Lohr described the great amounts of dust generated by dumping the bags of cement into two gallon pails, or into garbage cans, for mixing. Tin shop workers would encase the insulated pipe in sheetmetal, to protect the insulation from ladders placed against the pipes during inspections and while valves were operated. Lohr placed Wilson in the boiler rooms “sometimes,” supervising sheetmetal workers, but Lohr was unable to say how frequently the work of the two trades coincided in a boiler room.
[181]*181Lohr did not state specifically whether cement was used in the insulation of the running lines of pipe. Nor did he say that it was not. But he did testify expressly that cement was used to insulate all of the joints on pipes. Cement was also used, apparently exclusively, to insulate fifty gallon, galvanized, hot water tanks. The cement was applied to these tanks by a trowel and held in place during drying by chicken wire. There is no direct evidence placing sheetmetal workers in the proximity of the insulation by pipefitters of outdoor pipes or of hot water tanks.
The sheetmetal workers, however, were closely connected with the pipefitters in the heating of buildings. The buildings were heated by hot air passed through ducts. The air was heated by passing it over coils that were, in turn, heated by steam. These coils apparently were placed at intervals in the duct system. Wilson’s sheetmetal workers did the duct work and the bracing for the fans or blowers, while the pipefitters installed the coils. All of the ducts were insulated with asbestos insulation. This was because, as Lohr said, “[tjhere would be just as much loss of heat on that duct work as what it was like on a radiator or a boiler, if you didn’t.” The jury, therefore, properly could infer that the insulation on duct work was the same as that on boilers, namely, block covered by cement. Lohr was not asked who applied the insulation to the ducts. As the record stands, the jury could infer that the insulation was done by pipefitters, working in proximity to sheetmetal workers, or by sheetmetal workers, or by both.
The record is clear that sheetmetal workers did work with asbestos cement in air conditioning buildings, or areas of buildings, at ABL. As Lohr described it, freestanding air conditioning units would be delivered to the tin shop where the sheetmetal workers would insulate them with Armstrong cement. Then they would take the unit to the area where it was to be installed, the jury could infer, by connection to the same duct system that transported hot air during the heating season. Once installed, the sheetmetal workers would place another cover or layer of insulation over the previously applied [182]*182insulation. Wilson was in the vicinity of the dust generated by these operations.
In light of the testimony of Dr. Gabrielson, that occupational exposure to asbestos causes mesothelioma, even during temporary employment, the evidence set forth above amply supports the verdict against ACandS in the Wilson case. Consequently, retrial of the Wilson case will include ACandS as a defendant.
V
The remaining issue is whether the defendants’ motions for judgment on the issue of punitive damages should have been granted! If so, the claims for punitive damages are foreclosed on retrial.
The instant case was tried in November and December 1993, well after our decision in Owens-Illinois, Inc. v. Zenobia, 325 Md. 420, 601 A.2d 633, which established the rule for punitive damages in product liability cases. Subsequently, this Court decided ACandS, Inc. v. Godwin, 340 Md. 334, 667 A.2d 116, in which we applied the Zenobia rule to PH and to ACandS. Godwin was the consolidation of 8,555 cases and was tried on a most comprehensive record. In Godwin we held, in the language of the Zenobia rule, that “[t]here is a want of clear and convincing evidence that ACandS marketed asbestos products in bad faith, knowing of the danger to bystanders, and in conscious or deliberate disregard of the threat to the safety of bystanders.” 340 Md. at 390, 667 A.2d at 143. We also held that “[u]nder Zenobia there was insufficient evidence to take punitive damages to the jury against PH on behalf of’ the three bystander plaintiffs who had obtained punitive damage verdicts against PH. Id. at 382, 667 A.2d at 139.
The Court of Special Appeals rendered its decision in the instant matter before this Court’s opinion in Godwin was filed. Godwin controls this case, and we hold that the evidence was legally insufficient to support the verdicts for punitive damages.
[183]*183The plaintiffs submit to us that the record here differs from that in Godwin, pointing to both documentary evidence and the testimony of witnesses. As to the former, plaintiffs’ briefs chronologically catalogue documentary evidence against each defendant, as did the Court of Special Appeals. ACandS, Inc. v. Asner, 104 Md.App. at 625-31, 632-35, 657 A.2d at 387-91, 391-93. Much of the documentary evidence relied upon by plaintiffs was discussed in our opinion in Godwin. Other documentary evidence was part of the record in Godwin but was not expressly referred to in the opinion because it did not differ substantially from that discussed in the opinion. Other documents relied upon by the plaintiffs in the case now before us, that were not included in the record in Godwin, relate to workers’ compensation claims. Some of these claims were in evidence in Godwin. The new matter is simply cumulative evidence of actual knowledge on the part of the defendants that insulators in their employ had become ill with an asbestos-induced disease. There is no substantial difference between this new evidence and that considered in Godwin. The new documentary evidence relates to users, but the class with which we are concerned here is bystanders, the class to which each of the Claimants belonged.8
The plaintiffs also produced testimony that neither defendant was warning insulators of the risks of exposure to asbestos dust. One such witness was William E. Whitley, an insulator who, through the asbestos workers local union, worked primarily for PH between 1940 and 1982 at thousands of job sites. He served as a foreman and as a supervisor. He said that he had never been advised by officials of PH to wear [184]*184a respirator. Nor was he told by PH officials of the compensation claims being filed by other insulators against PH. Charles R. Fort, Sr., another union insulator, worked “on and off’ between 1954 and 1973 for ACandS, at times holding the job title of foreman. He testified that he was never shown various safety manuals and other publications dealing with protective measures to ehminate excessive exposure to asbestos dust.9 Francis Ruppreeht,10 a former branch manager for ACandS in Baltimore prior to 1970, testified by deposition. The witness said that, in a conversation with a representative of ACandS’s workers’ compensation insurance carrier sometime between 1965 and 1969, the witness expressed concern over an asbestosis claim by an ACandS employee but was assured that the insurer would take care of it by a cash settlement.
The foregoing testimony, once again, deals with users. To the extent that inferences relating to actual knowledge of risk to bystanders can be drawn from such evidence, it does not substantially assist the plaintiffs in reaching the clear and convincing standard with respect to the second element of the Zenobia test, namely, a conscious or deliberate disregard of the threat to the safety of bystanders.
The plaintiffs also emphasize the testimony of Theodore Mannell, a vice president and former member of the board of directors of PH. On cross-examination the witness was referred to provisions of the Maryland Workers’ Compensation Act dating back to 1939, dealing with occupational diseases, and providing that any process or application involving an exposure to, or direct contact with, asbestos dust was an extra-hazardous occupation. The testimony then proceeded as follows:
[185]*185“Q. Okay. Now sir, that would entail the activities of [PH], wouldn’t it?
“A. And anybody that worked in a shipyard.
“Q. Okay. And that was something you would think that [PH] would have been aware of in 1939, correct?
“A. Yes.”
Although the witness may have been interpreting “that” and “something” to be the provisions of the compensation law, plaintiffs argue that PH thereby admitted that it should have been aware in 1939 of the risk to bystanders of exposure to asbestos dust.
In somewhat the same vein plaintiffs argue, based on the testimony of Alan Stokely, a former manager for ACandS, that that company had notice of the risks to bystanders. Stokely acknowledged breathing asbestos dust while visiting job sites when asbestos “was thought of as being a product that had dangers with it and especially if you were a heavy smoker or even, I guess, even a light smoker.”
All of these arguments reveal a flawed perception of the Zenobia standard, as applied in Godwin, and as again recently illustrated in Owens-Corning Fiberglas Corp. v. Garrett, 343 Md. 500, 682 A.2d 1143 (1996). The two-part test of actual knowledge of the defect and deliberate disregard of the consequences, Godwin, 340 Md. at 359, 667 A.2d at 128, is applied as of the time the product left the defendants’ possession or control. Zenobia, 325 Md. at 462, 601 A.2d at 653-54. The state of the art was summed up in Godwin where, speaking of the claim against ACandS, we said:
“The significance of the foregoing to the punitive damages issue before us is that, as late as the spring of 1968, leading medical authorities, the Government, and both management and labor in the industry were focusing on the exposure of asbestos workers and on the development and implementation of strategies so that those who handled asbestos could do so safely.”
[186]*186Godwin, 340 Md. at 390, 667 A.2d at 143. Given the differences of opinion in the state of the art at the time over whether asbestos could be handled safely, there is insufficient evidence that the defendants supplied products in conscious or deliberate disregard of the threat to the safety of bystanders.
The ultimate submission by plaintiffs asks this Court to reject any distinction between users and bystanders in the law of punitive damages in asbestos exposure liability cases. Plaintiffs assert that inhalation of the asbestos fibers is the cause of the harm, that inhalation of the fibers is not restricted to insulators, that other trades were known to work closely with insulators in many workplaces, and that the risks to bystanders should have been known when the earliest of the asbestos-induced lung disease claims were presented under workers’ compensation laws. The user-bystander distinction, however, is valid because, in the law of products liability, the actual malice requirement of Zenobia would be unrealistic if that degree of malice had to be directed to a particular or identified individual, so that the law necessarily deals with classes. The verdicts in the 8,555 cases consolidated in God-win were structured on the distinction, and this Court’s rulings on punitive damages accepted the distinction. The instant plaintiffs, whose cases were not part of the consolidation in Godwin, have not advanced any persuasive reasons for their being treated differently.
At bottom, plaintiffs ask us to rule that a jury could find that the zone of danger in asbestos exposure cases is coterminous with the scope of causation in fact as it is known today. As plaintiffs see it, a jury may infer that, from the known diseases suffered by users, an expert in the field should have known that bystanders were unreasonably at risk as well. But the issue under consideration is punitive damages, not liability for compensatory damages. At a minimum the analysis advocated by plaintiffs ignores Zenobia’s requirements for a state of corporate mind at the time of distribution that consciously and deliberately disregards the threat to bystanders and for that state of corporate mind to be proven by clear and convincing evidence.
[187]*187
JUDGMENT OF THE COURT OF SPECIAL APPEALS REVERSED AND CASE REMANDED TO THAT COURT FOR THE ENTRY OF A JUDGMENT VACATING THE JUDGMENTS OF THE CIRCUIT COURT FOR BALTIMORE CITY AND REMANDING THESE CASES TO THAT COURT FOR A NEW TRIAL ON LIABILITY FOR COMPENSATORY DAMAGES ONLY. COSTS IN THIS COURT AND IN THE COURT OF SPECIAL APPEALS TO BE PAID BY THE RESPONDENTS.