UNITED STATES LIABILITY INSURANCE COMPANY, Plaintiff, Appellant, v. Livingstone R. SELMAN, et al., Defendants, Appellees, 70 F.3d 684

SELYA, Circuit Judge. In this appeal, plaintiff-appellant United States Liability Insurance Company (USL-IC) strives to extricate itself from coverage obligations owed to its insured, Livingstone R. Selman, vis-a-vis personal injury claims brought by Robin Razza on behalf of her minor daughter. The district court ruled that USLIC had a duty to indemnify Selman with respect to those injuries that occurred while the subject policies were in force. See USLIC v. Selman, 882 F.Supp. 1163 (D.Mass.1995). USLIC appeals. We affirm. I. BACKGROUND The chronology of events is not in dispute.

Selman owned an apartment house situated at 2 North Avenue, Roxbury, Massachusetts. In 1982, he rented apartment # 3A to Robin Razza. On May 6, 1983, Robin gave birth to Carol Ann Razza. In the fall of 1984, a physician discovered that Carol Ann had contracted lead poisoning. On February 5, 1985, an inspector from the Massachusetts Child Lead Poisoning Prevention Program (the Agency) found that both the Razzas’ apartment and the building’s common areas contained lead paint. The Agency informed Selman of its findings.

Shortly thereafter, a fire damaged apartment #3A, and Selman, responding to his tenant’s expressed desire to relocate, moved the Razzas to apartment # 1A. He also requested that the Agency inspect the apartment. The inspection occurred on March 7, 1985, and disclosed the presence of lead paint. The Agency notified Selman and he made arrangements to purge the entire building (including apartment # 1A).

Inspection reports reveal that by March 29 lead removal in apartment # 1A was “95% complete.” Beyond that date, the pace of lead removal in the Razzas’ apartment is unclear: all that we can tell from the record is that, by September of the following year (when the Razzas departed), Selman had rid the entire building of the residue of lead paint. At all times material hereto, Selman purchased insurance coverage annually. For four consecutive one-year periods commencing May 4, 1981, Selman insured the apartment house with Mutual Fire & Marine Insurance Company. In May of 1985, his allegiance shifted.

Coincident with the expiration of the latest Mutual Fire policy, Selman bought a one-year policy from USLIC, effective May 4, 1985. The next year, USLIC issued a renewal policy effective May 4,1986. Each policy covered claims for bodily injuries arising out of Selman’s ownership, maintenance, and use of the property.

The policies define “bodily injury” as “bodily injury, sickness or disease sustained by any person which occurs during the policy period,” and define an “occurrence” as “an accident, including continuous or repeated exposure to conditions, which results in bodily injury or property damage neither expected nor intended from the standpoint of the insured.” Long after the second of USLIC’s two one-year policies lapsed, Robin Razza asserted a claim against Selman for Carol Ann’s lead paint poisoning.

Bent on exonerating itself from all responsibility under its policies in regard to this claim, USLIC brought a declaratory judgment action against Selman and the Razzas in the United States District Court for the District of Massachusetts. See 28 U.S.C. §§ 2201-2202 (1988); Fed. R.Civ.P. 57. It premised jurisdiction on diversity of citizenship and the existence of a controversy in the requisite amount. See 28 U.S.C. § 1332(a). In due course, the parties tried the case to the court on stipulations of fact, documentary submissions, and the live testimony of the Razzas’ expert witness, Dr. John Graef.

The district judge determined that USLIC had no duty to indemnify Selman in respect to claims for injuries resulting from the ingestion of lead paint prior to May 4, 1985 (the inception date of its first policy), and the defendants do not challenge this determination on appeal. The judge also concluded, however, that USLIC had a duty to indemnify Selman with respect to claims arising out of Carol Ann’s ingestion of lead paint while USLIC’s policies were in force, that is, from May 4, 1985 until May 3, 1987. After the district court entered a declaratory judgment to this effect, USLIC appealed. II.

STANDARD OF REVIEW We face a preliminary dispute as to the applicable standard of review. Citing Pallet v. Gallagher, 725 F.2d 131, 134 (1st Cir.1984), the appellant insists that, inasmuch as its claims require construction of the terms of an insurance policy, appellate review is plenary. This generalization oversimplifies the matter, and, in the end, is wide of the mark. To be sure, it is for the court to determine whether the terms of an integrated agreement are unambiguous and, if so, to construe them according to their plain meaning. See Allen v.

Adage, Inc., 967 F.2d 695, 698 (1st Cir.1992); RCI Northeast Servs. Div. v. Boston Edison Co., 822 F.2d 199, 202 (1st Cir.1987); Robert Indus., Inc. v. Spence, 362 Mass. 751, 291 N.E.2d 407, 409-10 (1973). In this sense, questions about the meaning of contractual provisions are questions of law, and we review the district court’s answers to them de novo. See ITT Corp. v. LTX Corp., 926 F.2d 1258, 1261 (1st Cir.1991).

However, when the district court’s answers rest not on plain meaning but on differential findings by the trier of fact, derived from extrinsic evidence as to the parties’ intent with regard to an uncertain contract provision, appellate review proceeds under the “clearly erroneous” standard. See RCI Northeast, 822 F.2d at 202. The same standard pertains whenever the trial court decides factual matters that are essential to ascertaining the parties’ rights in a particular situation (though not dependent on the meaning of contractual terms per se). See, e.g., Reliance Steel Products Co. v. National Fire Ins.

Co., 880 F.2d 575, 576-77 (1st Cir.1989). In these types of cases, the issues are ordinarily fact-dominated rather than law-dominated, and, to that extent, the district court’s resolution of them is entitled to deference. See In re Howard, 996 F.2d 1320, 1328 (1st Cir.1993) (“Many cases involve what courts term ‘mixed’ questions — questions which, if they are to be properly resolved, necessitate combining factfinding with an elucidation of the applicable law.

The standard of review applicable to mixed questions usually depends upon where they fall along the degree-of-deference continuum: the more fact-dominated the question, the more likely it is that the trier’s resolution of it will be accepted unless shown to be clearly erroneous.”); Roland M. v. Concord Sch. Comm., 910 F.2d 983, 990-91 (1st Cir.1990) (similar), cert. denied, 499 U.S. 912, 111 S.Ct. 1122, 113 L.Ed.2d 230 (1991). These principles resonate here. The appellant attempts to escape from its contractual obligations on three alternative grounds.

First, it denies that coverage was ever triggered, taking the position that Carol Ann sustained no discernible injuries while its insurance policies were in force. Second, the appellant says that, because Carol Ann’s injuries were bound up with her earlier in-gestión of lead paint (first diagnosed in 1984), they fell outside the scope of its policies (which were written in 1985 and 1986, respectively). Both of these defenses have sizeable factual components, hinging, as they do, on whether the evidence shows that discrete injuries occurred during the relevant coverage periods.

Third, the appellant says that, because Selman knew about the looming liability on the inception date of the first policy, the known loss doctrine precludes him from insuring against the Razzas’ claims. The potency of this defense likewise depends on the facts: what Selman knew and when he knew it. At bottom, then, USLIC’s appeal challenges the district court’s factfinding; Fed.R.Civ.P. 52(

a) applies in full flower; and appellate review is circumscribed by the jurisprudence of clear error. This is of appreciable importance because clear error review ordinarily heralds a rocky road for an appellant. Under this standard, “appellate courts cannot presume to decide factual issues anew.” Cumpiano v. Banco Santander P.R., 902 F.2d 148, 152 (1st Cir.1990). Rather, “Rule 52(

a) commands, and our precedents ordain, that deference be paid to the trier’s assessment of the evidence.” Id. (citing representative cases). Moreover, the clearly erroneous rule loses none of its vigor “when the [lower] court’s findings do not rest on credibility determinations, but are based instead on physical or documentary evidence or inferences from other facts.” Anderson v. City of Bessemer City, 470 U.S. 564, 574, 105 S.Ct. 1504, 1511, 84 L.Ed.2d 518 (1985); accord In re fully, 818 F.2d 106, 108-09 (1st Cir.1987).

In the last analysis, an appellate tribunal “ought not to upset findings of fact or conclusions drawn therefrom unless, on the whole of the record, [the judges] form a strong, unyielding belief that a mistake has been made.” Cumpiano, 902 F.2d at 152. As long as the district court’s rendition of the record is plausible, our inquiry is at an end. III. ANALYSIS We divide our analysis into four segments, adding to the three grounds of appeal just mentioned a matter that speaks to the interrelationship of the liability ceilings contained in USLIC’s two insurance policies. A. Was Coverage Triggered?

Massachusetts law supplies the basis for decision in this diversity case. See Erie R. Co. v. Tompkins, 304 U.S. 64, 78, 58 S.Ct. 817, 822, 82 L.Ed. 1188 (1938). Under Massachusetts law, the insured bears the initial burden of proving that an injury occurred within the coverage ambit of the insurance policy. See, e.g., Trustees of Tufts Univ. v. Commercial Union Ins. Co., 415 Mass. 844, 616 N.E.2d 68, 74 (1993).

Once the insured establishes basic risk coverage, the devoir of persuasion shifts to the insurer to prove a defense to coverage, say, the applicability of a policy exclusion or the insured’s failure to comply with conditions precedent. See Gusson v. Boston Mut. Life Ins. Co., 326 Mass. 571, 95 N.E.2d 670, 672 (1950). The court below understood these rules and applied them appropriately. After reviewing the documentary evidence and considering Dr.

Graef s erudite testimony on the nature of lead poisoning and its manifestations in Carol Ann Razza’s case, the court found that “at least a portion” of Carol Ann’s claimed damages arose as a result of exposure to lead paint at the apartment building during the currency of the appellant’s policies. USLIC, 882 F.Supp. at 1164. If sustainable, this finding evinces that the coverage trigger had been pulled: Selman had met his entry-level burden by making a pri-ma facie showing that some part of the injury claimed falls within the coverage ambit of the subject policies.

Although the appellant attacks this finding hammer and tongs, we believe it is adequately supported by the record. The nisi prius roll includes a

summary of Carol Ann’s blood toxicity levels (which, after lead paint poisoning was first diagnosed, remained abnormally high throughout her stay at 2 North Avenue). In explaining the significance of the data, Dr. Graef testified that the sharp increases which occurred from time to time (sometimes called “spikes”) were directly traceable to the child’s sporadic ingestion of lead paint chips. The data showed — and Dr. Graef confirmed — that several such episodes occurred during the interval when the appellant’s policies were in force. Judge Tauro queried Dr.

Graef as to whether he regarded the spikes as “a manifestation of lead that [Carol Ann] had in her system” before May 4, 1985. The witness responded negatively, indicating that such levels were “spontaneously reportable.” Moreover, in the doctor’s opinion the roent-genographie evidence demonstrated that Carol Ann consumed additional chips of lead paint during the currency of the appellant’s policies. The district court’s finding that these new incidents caused further injury, see id. at 1165, is also supportable. Dr.

Graef spelled out in considerable detail the effects of ingesting lead on neurological development in early childhood, and testified that Carol Ann had suffered brain damage, including “significant gaps” in her auditory and verbal performance, as the direct result of ingesting lead while USLIC was on the risk. When Judge Tauro pressed Dr. Graef about whether a tie existed between the spikes in Carol Ann’s toxicity levels and her resulting injuries, the doctor responded in the affirmative.

He testified, among other things, that the predictable consequence of each major ingestion of lead paint “probably is that some damage is done to the brain,” and that increases in toxicity levels measurable by standard tests “refleet[ ] injury.” Given this dialogue and certain other insights — for example, the appellant neither impeached Dr. Graefs testimony nor adduced any contradictory evidence — we cannot impute clear error to the judge’s finding that Carol Ann Razza suffered new and further injuries during the relevant coverage periods.

Accordingly, coverage was triggered and the district court correctly shifted the burden to the appellant to demonstrate that some contractual exclusion or other policy defense foreclosed indemnification. The appellant claims to have carried that burden twice over. The district court disagreed. It is to those disputed defenses that we now turn. B. The Post-Manifestation Doctrine. The appellant raises no contractual provision as a defense to coverage here. Instead, it contends that what it euphemistically terms the “post-manifestation doctrine” has the same inhibitory effect.

Under the guise of this euphemism, USLIC hypothesizes that when a disease process of a certain type manifests itself before an insurance policy incepts, all future injury of the same genre should be deemed to relate back to the original condition even if the victim incurs subsequent injury from continued exposure to the causative agent during the policy period.

As applied in this case, the hypothesis holds that if a person contracts lead poisoning prior to the inception of the tortfeasor’s insurance policy but continues to be exposed to lead paint and thereby suffers further injury while the policy is in force, any claim that she may assert against the tortfeasor will not be covered because lead poisoning constitutes a single injury “occurring” before the policy incepted. As doctrines go, this one has very little in the way of a pedigree.

The appellant cites no reported case discussing anything that resembles such a doctrine, and our independent research has come up equally dry. In any event, we need not tarry over the hypothesis. As we have already indicated, see supra

Part 111(A), the district court had before it compelling evidence that Carol Ann Razza ingested several “big meals” of lead paint chips while the appellant’s policies were in force, and Dr. Graef testified that each such ingestion caused (or potentially could cause) discrete injury. On this basis, the district court warrantably found a “clear nexus” between Carol Ann’s “big meals” and the spikes in her toxicity levels. USLIC, 882 F.Supp. at 1165. Each exposure can, therefore, reasonably be seen as a separate, injury-producing occurrence. No more is exigi-ble. C.

The Known Loss Doctrine, The appellant next asseverates that the known loss doctrine renders the risk of further injury to Carol Ann uninsurable because Selman knew prior to the inception date of the initial policy that his apartment building contained lead paint and that Carol Ann was suffering from lead poisoning. The argument takes the following form.

The purpose of insurance is to protect against misfortune by permitting an actor to whom the law assigns the risk of a particular kind of loss to shift the burden of it to an institution better able to assume and manage the particular risk through diversification across risk categories. See Group Life & Health Ins. Co. v. Royal Drug Co., 440 U.S. 205, 211, 99 S.Ct. 1067, 1073, 59 L.Ed.2d 261 (1979); see also 1 Ronald A. Anderson & Mark S. Rhodes, Couch on Insurance (Second) §§ 1:3, 2:7 (rev. 2d ed. 1984). Thus, the presence of risk runs to the very essence of an insurance contract.

Where there is no risk of loss — as where a loss has already occurred before a policy takes effect — insurance ceases to serve its socially utile purpose of risk-spreading. Hence, the law, embodied in the known loss doctrine, precludes coverage when the insured knows in advance of the policy’s effective date that a specific loss has already happened or is substantially certain to happen. There are two iterations of the known loss doctrine. The doctrine exists both as a function of the standard general liability insurance contract and at common law.

We discuss the first iteration briefly, mainly for the sake of completeness. Since 1966, the insurance industry has defined an “occurrence” as that word is used in the standard general liability policy to include only accidents that result in bodily injury or property damage that is “neither expected nor intended from the standpoint of the insured.” See Barry R. Ostrager & Thomas R. Newman, Handbook on Insurance Coverage Disputes § 8.03[a] (7th ed. 1994); 11 Couch, supra, § 44:289.

Under this policy provision (which graces the policies in question here), it has been held that if an insured “knew ... that there was a substantial probability that certain consequences” would result from his acts or omissions, there is no “occurrence” within the meaning of a general liability policy, and, hence, no coverage. City of Carter Lake v. Aetna Cas. & Sur. Co., 604 F.2d 1052, 1058-59 (8th Cir.1979). In this case, the appellant did not brief a contract-based coverage defense on appeal, and at oral argument appellant’s counsel expressly disclaimed any reliance on such a defense.

Accordingly, we do not pursue this iteration of the known loss doctrine. The common law version of the known loss doctrine is part of the warp and woof of Massachusetts insurance law. The Massachusetts Supreme Judicial Court (SJC) recently inspected its composition in SCA Servs., Inc. v. Transportation Ins. Co., 419 Mass. 528, 646 N.E.2d 394, 397-98 (1995). There, the insured operated a chemical waste site in Illinois.

Local residents brought a nuisance action, alleging that its activities on the site were contaminating the local water supply, causing subsidence, filling the air with dust, and permitting the escape of noxious gasses. See Village of Wilsonville v. SCA Servs., Inc., 86 Ill.2d 1, 55 Ill.Dec. 499, 503-05, 426 N.E.2d 824, 828-30 (1981). The trial court declared the site to be a public nuisance and closed the plant. The Illinois Supreme Court affirmed. See id., 55 Ill.Dec. at 502, 426 N.E.2d at 827. Subsequently, SCA purchased an insurance policy.

Several of the same residents then brought a class action seeking damages for personal injuries suffered as the result of exposure to the conditions limned in the initial nuisance action. SCA sought a declaration that its insurer had a duty to defend and indemnify with respect to the class action.

The SJC determined that, because the prior adjudication in Illinois put SCA on actual notice that the class members had suffered injuries as the result of the same conduct and conditions that led to the shutdown of the site, it had “full knowledge” of its probable liability for their damages prior to purchasing the insurance policy. SCA, 646 N.E.2d at 398. Thus, the known loss doctrine barred coverage inasmuch as the concept of insurable risk becomes a fiction “where an insured knows there is a substantial probability that it will suffer or has already suffered a loss.” Id. at 397.

Before we can measure the case at bar against the specifications of the common law doctrine as elucidated in SCA we must address two threshold questions. The first concerns the standard — objective or subjective — by which the insured’s state of mind is to be gauged. Though Massachusetts law is not explicit on the point, there is spoor for the cognoscenti.

SCA strongly suggests the use of a subjective standard to determine whether a given loss was “known.” See id. (stating that “insurance risk is eliminated ... where an insured knows, when it purchases a policy, that there is a substantial probability that it will suffer or has already suffered a loss”). The quoted language is almost identical to that used (and more fully explicated) in Quincy Mut Fire Ins. Co. v. Abernathy, 393 Mass. 81, 469 N.E.2d 797 (1984). There, dealing with the contract-based iteration of the known loss doctrine, the SJC explicitly adopted a subjective test. See id., 469 N.E.2d at 800.

Moreover, SCA and all the cases relied on in SCA deal with insureds that had actual knowledge of a probable loss prior to securing coverage. See, e.g., Bartholomew v. Appalachian Ins. Co., 655 F.2d 27, 28 (1st Cir.1981) (insured had actual knowledge of probable loss based on its own intentional misuse of a machine that had on prior occasions caused injury); Gloucester v. Maryland Cas. Co., 668 F.Supp. 394, 403 (D.N.J.1987) (insured had actual knowledge of probable loss due to environmental contamination based on the closure of its landfill by state authorities); Outboard Marine Corp. v. Liberty Mut. Ins.

Co., 154 Ill.2d 90, 180 Ill.Dec. 691, 696-98, 607 N.E.2d 1204, 1209-11 (1992) (insured had actual knowledge of probable loss due to environmental contamination based on receipt of an EPA administrative order citing it as the source of the contamination). Guided by these clearly visible signposts, we hold that the applicability vel non of the known loss doctrine, in its common law form, depends on the insured’s actual knowledge of the looming loss. The test, therefore, is subjective, not objective. The remaining threshold issue relates to the devoir of persuasion.

The SJC apparently placed the burden of proof on this issue on the insurance company in a suit invoking the contract-based interaction of the known loss doctrine, see, e.g., City of Newton v. Krasnigor, 404 Mass. 682, 536 N.E.2d 1078, 1081-82 (1989), and we see no reason why the court would take a different tack in allocating the burden of proof on the counterpart issue in the common law setting.

Moreover, Massachusetts courts generally place the burden of proof on the party seeking to invalidate or avoid the application of a contract on analogous grounds, such as when an insurer raises the defense of fraud in the procurement of insurance. See, e.g., Roger Williams Grocery Co. v. Sykes, 357 Mass. 485, 258 N.E.2d 553, 555 (1970).

Finally, the SJC appears to have treated the known loss doctrine as an affirmative defense in SCA mimicking a majority of other courts, see, e.g., Gloucester, 668 F.Supp. at 402-03, and the usual rule, honored by Massachusetts as by most jurisdictions, is to plaee the burden of proving affirmative defenses on the party asserting them, see 19 Couch, supra, § 79:368 (discussing various affirmative defenses and assigning burden of proof to insurer).

For these reasons, we hold that, under Massachusetts law, the common law version of the known loss doctrine only applies when the insured actually knows on or before the effective date of the policy either that a loss has occurred or that one is substantially certain to occur. Relatedly, we hold that the common law version of the known loss doctrine is an affirmative defense to a suit on a Massachusetts policy. Accordingly, the insurer bears the burden of proving the insured’s actual knowledge. The district court seems to have anticipated these rulings.

It treated the known loss doctrine as an affirmative defense. After reviewing the evidence, it found the defense not proven. See USLIC, 882 F.Supp. at 1164. The court concluded that a “significant portion” of the injuries asserted arose after May 4,1985, and therefore could not be classified as “known” on that date. Extrapolating from this finding, the court held that, to the extent Carol Ann’s injuries stemmed from ingestions of lead paint occurring after May 4, 1985, but before the expiration of appellant’s second (and last) policy, Selman had not sought to insure against a known loss. See id.

While there was ample room for the court to come down the other way, we think that its crucial finding withstands scrutiny. To be sure, the matter is not open and shut. Selman knew by the spring of 1985 that his building contained lead paint. He also knew that Carol Ann Razza was suffering from lead poisoning. But these two facts, naked and unadorned, do not necessarily prove that Selman insured against a known loss. Three critical elements are lacking.

First, there is nothing in the record to show definitively that the lead paint in Selman’s building constituted the source of Carol Ann’s lead poisoning (and, more to the point, that Selman knew of the connection). Without such a showing, the known loss doctrine does not apply. Second, nothing in the record establishes that Selman actually knew that Carol Ann would suffer further injury from continued exposure to lead paint, and the trial court found in essence that he lacked any such appreciation of the disease process. See id.

Third, by late March of 1985 — six weeks before the first of the USL-IC policies became effective — the Razzas were living in an apartment in which lead removal was at least 95% complete. Selman could easily have assumed that Carol Ann was no longer exposed to any significant dose of lead paint, and would therefore suffer no further injury. These are not merely theoretical possibilities.

The deposition testimony contained in the record strongly suggests that Selman had not drawn any connection in his mind between the ongoing removal of lead paint at 2 North Avenue and the future medical risks that the condition of the premises portended to Carol Ann Razza. The court had the right to credit that testimony, see Anthony v. Sundlun, 952 F.2d 603, 606 (1st Cir.1991) (explaining that in a bench trial, credibility choices are for the trier); FDIC v.

La Rambla Shopping Ctr., Inc., 791 F.2d 215, 222 (1st Cir.1986) (similar), especially since many familiar diseases, once contracted — measles, mumps, the HIV virus, to name a few — do not result in further injury based on repeated exposure to the causative agent. There is nothing in the record to show that Selman knew that, unlike these diseases, lead poisoning was a cumulative disease. The district court’s finding is strengthened by the utter lack of any evidence that Selman attempted to conceal or misrepresent the presence of lead paint in his apartment house when he applied for insurance.

To the extent that the appellant’s application form did not request such information, the appellant was the author of its own misfortune. See Vappi & Co. v. Aetna Cas. & Sur. Co., 348 Mass. 427, 204 N.E.2d 273, 276 (1965). It does not seem unfair to hold an insurance company, knowledgeable about the prevalence of lead paint in older buildings and hardened by the rough and tumble of the business world, to the consequences of which King Solomon long ago warned. See Proverbs 11:15 (“He that is surety for a stranger shall smart for it.”).

The short of it is that the appellant had the burden to prove that its insured knew of a probable loss, and the district court’s finding that he did not, viewed in light of the record evidence, is not clearly erroneous. The appellant attempts to steer the appeal into a different channel by way of two expedients. First, it asks us to treat this ease and SCA as a matched pair of ponies. But SCA is a horse of a much different hue. The Agency’s informal notification that Selman’s apartment building contained lead paint is at a considerable remove from the adjudication of a nuisance.

The agency action here at issue lacks both the finality and the preclu-sive effect of a court judgment. Moreover, the nature and causes of the injuries alleged in the class action against SCA were identical to those alleged in the prior nuisance suit. As the SJC observed, the insured actually knew on the basis of the earlier litigation that the class action plaintiffs claimed to have been injured — and it also knew that those claims had already been adjudicated (unfavorably to it). The scenario here is not the same.

The Agency in this case only informed Selman that his apartment building contained lead paint; it did not conclude that any particular injuries, much less Carol Ann’s injuries, had been caused by the lead in Selman’s building. In a nutshell, accepting the appellant’s view that, as a matter of law, the known loss doctrine encompasses this situation would take us several steps beyond the holding in SCA We are unwilling to take those steps. The appellant, presumably to suit its own convenience, selected a federal forum in preference to an available state forum.

It has no right to grouse if a federal court, sitting in diversity jurisdiction, declines to push state law past previously established frontiers. See Martel v. Stafford, 992 F.2d 1244, 1247 (1st Cir.1993); Porter v. Nutter, 913 F.2d 37, 41 (1st Cir.1990). The organic growth of state law is best left to state courts, particularly in areas that traditionally have been committed to, and regulated by, the states. Insurance is such a field.

The appellant’s second effort to skirt the district court’s factfinding involves its contention that the court applied the wrong legal standard in determining whether Selman knew of his likely liability to Carol Ann Raz-za for injuries related to future ingestions of lead paint. This gambit is conceptually sound in the sense that a “finding of fact predicated upon, or induced by, a misapprehension of law is robbed of its customary vitality.” RCI Northeast, 822 F.2d at 203. The concept is inoperative, however, when a party attempts to play the artful Dodger, cf.

Charles Dickens, Oliver Twist (1838), recasting its objections to the district court’s findings of fact as disputes about the governing law. See Reliance Steel, 880 F.2d at 577 (declaring that litigants may not “profit by dressing factual disputes in ‘legal’ costum-ery”). So it is here. The appellant derides the district court’s finding that Selman did not know Carol Ann Razza would sustain new injuries after May 4, 1985. Embedded in this finding, appellant claims, is the legal benchmark by which the district court evaluated the evidence in determining Selman’s state of knowledge.

This benchmark is wrong, appellant postulates, because the substantive law that governs Sel-man’s putative liability is based not on knowledge but on strict liability. See Bencosme v. Kokoras, 400 Mass. 40, 507 N.E.2d 748, 749 (1987). This is a red herring. Whether Massachusetts law renders Selman strictly liable for Carol Ann’s damages is irrelevant to whether Selman knew he was virtually certain to experience a loss as the inevitable result of his tenant’s continued exposure to lead paint during the policy periods.

It is the answer to this pivotal question that determines the applicability of the known loss doctrine to this case — and that question, as we have said, is predominantly a question of fact. To say more would be supererogatory. Because the district court’s findings of fact are not clearly erroneous, its rejection of the appellant’s known loss defense must be upheld. D. Applicability of Policy Limits. In this instance, the appellant issued two consecutive one-year policies to Selman.

Each policy contains a stipulation limiting the insurer’s liability to $300,000 “per occurrence,” and each policy states that “continuous or repeated exposure to conditions” is to be treated as a single “occurrence.” In its complaint for declaratory relief, the appellant prayed that, if it were found to have any obligation at all to indemnify Selman vis-a-vis the Razza claims, then in such event, the limits of liability contained in its two policies should be interpreted so as to cap the insurer’s total potential liability at $300,000. The district court did not entertain this prayer for relief.

The appellant now invites us to do so. We decline the invitation. In general, declaratory relief is discretionary. See, e.g., Ernst & Young v. Depositors Economic Protection Corp., 45 F.3d 530, 534 (1st Cir.1995); El Dia, Inc. v. Hernandez Colon, 963 F.2d 488, 493-94 (1st Cir.1992). Thus, we view the district court’s withholding of a declaration in regard to the appellant’s “policy limit” question through a deferential glass. In the process, we focus our inquiry on the whole of the circumstances confronting the district court. See El Dia, 963 F.2d at 492.

The trial judge did not spell out his reasons for declining to declare the parties’ rights in this regal'd. While courts should articulate grounds for them actions, see Pearson v. Fair, 808 F.2d 163, 165-66 (1st Cir.1986) (per curiam), the district court’s failure to do so here is not fatal, as the basis for the declination seems evident.

The insurance policies contain no definition of the operative terms (e.g., “continuous,” “repeated,” “conditions”); and the record suggests that there were many conditions to which Carol Ann Razza might have been exposed and which could have been sources of her deleterious ingestion of lead paint. Consequently, the lack of development in the record concerning the possible sources of the lead paint ingested by Carol Ann placed the lower court at so great a disadvantage that it reasonably could conclude that it was in no position to rule intelligently on the appellant’s request.

Accordingly, the court acted within the realm of its discretion in refusing the declaration. See, e.g., Askew v. Hargrave, 401 U.S. 476, 478-79, 91 S.Ct. 856, 857-58, 28 L.Ed.2d 196 (1971) (cautioning against grant of declaratory judgment on the basis of sparse and inadequate record); Public Affairs Assocs., Inc. v. Rickover, 369 U.S. 111, 112, 82 S.Ct. 580, 581, 7 L.Ed.2d 604 (1962) (per curiam) (similar); A.L. Mechling Barge Lines, Inc. v. United States, 368 U.S. 324, 330-31, 82 S.Ct. 337, 341-42, 7 L.Ed.2d 317 (1961) (similar). IV. CONCLUSION We need go no further.

This case pivots on the facts, not on the law — and factual issues that are resolved in a bench trial may not freely be relitigated on appeal. Discerning no error, we hold the appellant to its contractual duty. Affirmed. . Selman eliminated the hazard by scraping lead paint from the walls in some locations and covering over lead paint in other locations. Since the method of abstersion is not important for present purposes, we refer to both processes as “removal.” .

The record contains no hint either that Mutual Fire canceled Selman's coverage or that the change in carriers was somehow connected to the discovery of lead paint on the premises. .While pretrial discovery was ongoing, Robin Razza sued Selman to her daughter's behoof in a Massachusetts state court, seeking damages for injuries allegedly sustained as a result of Carol Ann's exposure to lead paint in the apartment building. That suit is still pending. . In reality, the cutoff date is probably September 27, 1986 (when the Razzas moved from 2 North Avenue). .

The district court’s holding, while obvious from its reasoning, is not explicitly articulated in the text of its opinion. The final judgment cured this omission. There, the court declared that: [I]n regard to the lawsuit filed against Livingstone Selman by Robin Razza, as mother and next friend of Carol Ann Razza ...: 1. The plaintiff has no duty to indemnify Livingstone Selman with respect to injuries to Carol Ann Razza resulting from ingestions of lead paint prior to May 4, 1985; 2.

The plaintiff has a duty to indemnify Livingstone Selman with respect to injuries to Carol Ann Razza resulting from ingestions of lead paint on and after May 4, 1985, and; 3. The case is closed. . There is nothing unorthodox about these views. Courts have found in other (similar) cases that each ingestion of lead paint leads to discrete injury. See, e.g., USLIC v. Farley, 626 N.Y.S.2d 238, 239-40 (App.Div.1995); General Accident Ins. Co. v. Idbar Realty Corp., 163 Misc.2d 809, 622 N.Y.S.2d 417, 419 (Sup.Ct.1994). . The appellant does direct us to an opinion of a Maryland state court, Harford Mut. Ins. Co. v.

Jacobson, 73 Md.App. 670, 536 A.2d 120 (1988), and two unpublished dispositions of trial judges (one federal and one state), as "authority” for the "doctrine.” But none of these cases involves comparable issues or facts, and none of them adverts by name to the elusive doctrine. . The SJC repeatedly emphasized the presence of actual knowledge both in the case before it and in its discussion of the precedents on which it relied. See, e.g., SCA, 646 N.E.2d at 397, 398. . Furthermore, the appellant made no compelling demonstration of a need for the declaration.

For instance, there is no showing that Carol Ann’s claim against Selman for the injuries she sustained within the coverage period could support a recovery of more than $300,000, and, thus, insofar as the trial court was concerned, the policy limit question may have appeared to be academic. The Declaratory Judgment Act notwithstanding, courts have no obligation to answer hypothetical questions. See El Dia, 963 F.2d at 494 (cautioning that courts should not issue declaratory judgments when the need is remote or speculative); Washington Pub. Power Supply Sys. v. Pacific N.W.

Power Co., 332 F.2d 87, 88 (9th Cir.1964) (similar).