Opinion

Plaintiff Barbara Howell appeals after the trial court granted State Farm Fire and Casualty Company’s motion for summary judgment in an action for damages arising from State Farm’s refusal to compensate Ms. Howell for landslide damage to her real property. We reverse.

This case presents two primary issues for our consideration. First, we must decide whether an insurer providing coverage under an “all risk” property insurance policy may contractually exclude coverage when an insured peril is the “efficient proximate cause” of a loss. We conclude that an insurer may not contractually limit its liability in this manner. Secondly, we must determine whether there is a triable issue of fact that the “efficient proximate cause” of the loss in this case was a covered peril (namely, fire). Since we conclude there is a triable issue of fact concerning this issue, we reverse the summary judgment.

Facts

Summary judgment is proper only if there is no triable issue of fact and, as a matter of law, the moving party is entitled to judgment. (Code Civ. Proc., § 437c.) “In reviewing a summary judgment, we are limited to the facts shown in the affidavits and those admitted and uncontested in the pleadings. We determine only whether the facts so shown give rise to a triable issue of fact. [In making this determination] the moving party’s papers are strictly construed, while those of the opposing party are liberally construed.” (Steele v. Totah (1986) 180 Cal.App.3d 545, 549 [225 Cal.Rptr. 635]; see also Molko v. Holy Spirit Assn. (1988) 46 Cal.3d 1092, 1107 [252 Cal.Rptr. 122, 762 P.2d 46].) Viewed in light of these rules, the record discloses the following facts:

*1449Plaintiff and appellant Barbara Howell is the owner of a 17-acre parcel of land in Lafayette, California, which is the site of her home, three rental units, and a commercial dog kennel. At the time of the events pertinent to this appeal, State Farm insured Ms. Howell’s property through five separate policies: a homeowner’s policy for the residence, three identical rental dwelling policies for the rental units, and a commercial policy for the kennel.

The Howell property is located on a slope which is subject to landslides. In the summer of 1985, a fire destroyed much of the vegetation on the slope, and, during the following winter, unusually heavy rains drenched the property. Because of the loss of vegetation, the soils on the denuded slope could not withstand the severe rainfall and a major landslide occurred. According to appellant’s expert, the landslide probably would not have happened had the ground cover been intact. Specifically, the expert’s report concluded: “Although water was the slide’s actuating mechanism, resistance to failure was severely impaired by destruction of the area’s natural biotechnical slope protection. On 6/28/85 the slope that failed was devastated by a fire. The destruction of the vegetation and rooting at the surface caused a[n] adverse modification to the hydrologic regime in the soil. . . . [^] In my opinion, if the vegetation had not been destroyed by the fire, the slope failure . . . probably would not have occurred.” This conclusion was based in part on the fact that no slides had occurred when the vegetation was in place even though there had been greater amounts of rainfall in past years.

Shortly after the landslide occurred, Ms. Howell filed a claim for slide damage under each of her five State Farm policies. In a series of five letters written in the spring of 1986, State Farm denied Ms. Howell’s claims. Each letter stated that the loss was not covered because it was caused by perils (earth movement and water damage) which had been specifically excluded under the policies.

In particular, the homeowners and rental policies each provided coverage against “accidental direct physical loss” to the “dwelling” and “other structures on the residence premises, separated from the dwelling by clear space. ...” The exclusions section of the homeowners policy provided in pertinent part: “We do not insure under any coverage for loss (including collapse of an insured building or part of a building) which would not have occurred in the absence of one or more of the following excluded events. We do not insure for such loss regardless of: a) the cause of the excluded event; or b) other causes of the loss; or c) whether other causes acted concurrently or in any sequence with the excluded event to produce the loss . . . .” The *1450policy then listed “Earth Movement” and “Water Damage” as specific perils excluded under this section.1

The exclusion provision of the rental dwelling policies is functionally identical to the homeowner’s exclusion quoted above, except for a slightly different definition of “Earth Movement.”2

The commercial policy provided coverage for “building(s) or structure(s)” against “direct physical loss unless the loss is excluded . . . .” The exclusions section provided in pertinent part: “This policy does not insure against loss caused directly or indirectly by any of the following. Such loss is excluded regardless of any other cause or event contributing concurrently or in any sequence to the loss. . . . [fl] C. Loss caused by, resulting from, contributed to or aggravated by any of the following: [fl] 1. earth movement *1451. . . [j|] 2. flood, [or] surface water . . . [ft] 4. water below the surface of the ground . . . .”3

After State Farm denied her claims, Ms. Howell filed the instant action for damages. She alleged causes of action for breach of contract, breach of the implied covenant of good faith and fair dealing, intentional and negligent misrepresentation, violation of Insurance Code sections 790.2 and 790.3, and breach of fiduciary duty.

State Farm’s answer generally denied the complaint’s allegations and advanced several affirmative defenses, including the policies’ exclusion for earth movement and water damage. As an additional affirmative defense, State Farm contended that it was “informed and believes . . . that there has been no physical loss to any property covered under the policy.”

Subsequently, State Farm filed a motion for summary judgment, contending that Ms. Howell’s policies did not cover her losses because (1) the policies excluded losses resulting from earth movement and water damage; and (2) the policies provided coverage only for physical loss to buildings and structures and the only damage had been to Ms. Howell’s land.

In opposition, Ms. Howell argued that the 1985 fire was the efficient proximate cause of her loss since it had permitted the landslide to occur. Consequently, under Insurance Code section 530 and California case law, the earth movement and water damage exclusions did not bar recovery under the policies. In response to State Farm’s claim that there was no damage to insured property, Ms. Howell contended that there was damage to insured structures and, additionally, State Farm had waived the right to assert the absence of such damage when it failed to raise it in the letters denying coverage.

The superior court granted State Farm’s motion, reasoning that Ms. Howell’s losses were not covered because of the policies’ earth movement *1452exclusion. The court expressly declined to decide whether there had been any damage to insured property.

This timely appeal followed.

Discussion

A. California Law Requires a Property Insurer to Provide Coverage When an Insured Peril Is the Efficient Proximate Cause of a Loss.

State Farm contends that the exclusions contained in its policies preclude recovery by the plaintiff because the loss would not have occurred “but for” an excluded peril (earth movement). Stated simply, the important question presented by this case is whether a property insurer may contractually exclude coverage when a covered peril is the efficient proximate cause of the loss, but an excluded peril has contributed or was necessary to the loss. We conclude that a property insurer may not limit its liability in this manner, since the statutory and judicial law of this state make the insurer liable whenever a covered peril is the “efficient proximate cause” of the loss, regardless of other contributing causes.4 Consequently, the policy exclusions at issue in this case are not enforceable to the extent they conflict with California law.

Our analysis of this issue begins with Insurance Code sections 530 and 532. These apparently contradictory sections state: “An insurer is liable for a loss of which a peril insured against was the proximate cause, although a peril not contemplated by the contract may have been a remote cause of the loss; but he is not liable for a loss of which the peril insured against was only a remote cause.” (Ins. Code, § 530.)

“If a peril is specially excepted in a contract of insurance and there is a loss which would not have occurred but for such peril, such loss is thereby excepted even though the immediate cause of the loss was a peril which was not excepted.” (Ins. Code, § 532.)

The apparent conflict between these two sections was resolved by the Supreme Court in Sabella v. Wisler (1963) 59 Cal.2d 21 [27 Cal.Rptr. 689, *1453377 P.2d 889]. There the insureds’ home was damaged when water escaping from a negligently installed sewer pipe caused the home’s foundation to settle. The insureds’ policy covered damage caused by negligence but excluded loss caused by settlement. (Id., at pp. 23-27.) The Supreme Court held that the insurer was liable for the loss under the following general rule: “ ‘[I]n determining whether a loss is within an exception in a policy, where there is a concurrence of different causes, the efficient cause—the one that sets others in motion—is the cause to which the loss is to be attributed, though the other causes may follow it, and operate more immediately in producing the disaster.’ ” (Id., at pp. 31-32, quoting from Couch on Insurance (1930) § 1466; see also Garvey v. State Farm Fire & Casualty Co. (1989) 48 Cal.3d 395, 402 [257 Cal.Rptr. 292, 770 P.2d 704].)

In Sabella, the insurer advanced an argument which is conceptually identical to that advanced by State Farm in this case. There, the insurer contended it was not liable because settlement was a necessary link in the chain of causation leading to the loss. However, instead of relying on the language of the policy exclusion to support its argument—as State Farm does in the present case—the insurer in Sabella relied on the “but for” language of Insurance Code section 532. In rejecting the insurer’s contention the Supreme Court stated: “It is . . . apparent that if section 532 were construed in the manner contended for by defendant insurer, where an excepted peril operated to any extent in the chain of causation so that the resulting harm would not have occurred ‘but for’ the excepted peril’s operation, the insurer would be exempt even though an insured peril was the proximate cause of the loss. Such a result would be directly contrary to the provision in section 530, in accordance with the general rule, for liability of the insurer where the peril insured against proximately resulted in the loss . . . .” (Id., at p. 33, italics added.) Consequently, the Supreme Court harmonized Insurance Code sections 530 and 532 by construing the “but for” clause of section 532 to mean a “proximate cause” of the loss, and the “immediate cause” to mean the cause most immediate in time to the damage. (Id., at pp. 33-34; see also Garvey, supra, 48 Cal.3d at p. 402.)

Similarly, in the present case, if we were to give full effect to the exclusion clauses contained in State Farm’s policies “the insurer would be exempt even though an insured peril was the proximate cause of the loss. Such a result would be directly contrary to the provision in section 530, in accordance with the general rule, for liability of the insurer where the peril insured against proximately resulted in the loss . . . .” (Sabella, supra, 59 Cal.2d at p. 33.) In short, the exclusion clauses are contrary to Insurance Code section 530, which provides that an insurer “is liable for a loss” proximately caused by a covered peril. Consequently, the exclusion clauses *1454are not enforceable to the extent they purport to limit the insurer’s liability beyond what is permitted by California law.

Subsequent cases applying Sabella show that our interpretation of that case is correct. In Sauer v. General Ins. Co. (1964) 225 Cal.App.2d 275 [37 Cal.Rptr. 303], leaking water pipes caused the earth under the plaintiffs’ house to sink and its foundation to settle. The plaintiffs’ insurance policy covered loss caused by plumbing leakage but excluded loss “ ‘caused by, resulting from, contributed to, or aggravated by’ ” earth movement or water, and also excluded loss by settling. (Id., at pp. 276-277.) The trial court found that earth movement had “contributed to” and “aggravated” the damage, and concluded the insurer was not liable. The Court of Appeal reversed. Relying on Sabella, the court held that the policy covered the loss because the “efficient proximate cause” of the damage was the covered plumbing leak. (Id., at pp. 278-280.) The court reached this result despite the language in the policy excluding liability where earth movement had been a factor in the loss.

Similarly, in Gillis v. Sun Ins. Office, Ltd. (1965) 238 Cal.App.2d 408 [47 Cal.Rptr. 868, 25 A.L.R.3d 564], a windstorm lifted a gangway into the air and dashed it violently into a docking facility, causing the dock to sink into the water, where it broke up under the pounding of the waves. (Id., at pp. 410-411, 416-417.) The insurance policy covered losses caused by the wind but excluded damage “ ‘caused by, resulting from, contributed to or aggravated by’” water. (Id., at pp. 415-416.) In what we believe is an overly technical opinion, the court concluded the exclusionary language in the policy could not protect the insurance company, at least where the action of the peril insured against preceded and operated through an excluded peril to cause the damage. (Id., at pp. 419-421, 423-424.)

The insurer in Gillis argued that the damage to the dock was “contributed to or aggravated by” water and hence was not covered. The company further contended that the “contributed to or aggravated by” exclusionary language precluded application of Sabella's proximate cause doctrine and barred recovery if an excluded peril played any part, no matter how small, in producing the loss. (238 Cal.App.2d at p. 419.)

The appellate court rejected this argument. Although the court recognized that, as a factual matter, the water had “contributed to or aggravated” the damage to the dock, it nevertheless concluded Sabella imposed liability since the fall of the gangway “was the start of a chain of events which permitted the waves to cause damage to the float.” (238 Cal.App.2d at p. 419.) According to the court, it was proper to apply the Sabella-Sauer proximate cause analysis because the wind “creatfed] a condition that per*1455mitted natural forces, which alone may have caused no damage, to effect the damage for which recovery is sought.” (Id., at p. 421.) Consequently, the court applied the Sabella-Sauer analysis despite the language of the policy which, on its face, appeared to preclude that analysis.

More recently, the Sabella analysis was applied in Premier Ins. Co. v. Welch (1983) 140 Cal.App.3d 720 [189 Cal.Rptr. 657]. In that case, the plaintiffs’ hillside home slid off its foundation during a heavy rainy season because the subdrain, negligently damaged when the house was built, could not accommodate the runoff, which saturated the fill under the house and caused the slide. The parties stipulated that the slide would not have occurred if the drain had not been damaged. (Id., at pp. 722-723.)

The plaintiffs’ insurance policy covered negligence but excluded loss “caused by, resulting from, contributed to or aggravated by” surface or subsurface water or “water which backs up through sewers or drains . . . .” (140 Cal.App.3d at p. 723.) The trial court concluded that the heavy rainfall was the efficient proximate cause of the landslide, and that the loss was excluded. (Ibid.) The Court of Appeal reversed and held the insurer liable. Relying on Sabella, the court applied “the settled principle that where, as here, the loss occurs through a concurrence of covered and uncovered risks, the insurer’s liability under the policy depends on whether the efficient cause of the loss (that is the cause that sets the others in motion) is a covered peril. If it is, the insurer will be held liable even though other exempted causes also contribute to the loss.” (Id., at pp. 724-725.) Thus, in Premier, as in Sauer and Gil lis, the court held that the causation analysis was controlled by Insurance Code section 530 and Sabella, not the specific exclusionary language contained in the policy.

Finally, the Supreme Court recently reaffirmed the Sabella proximate cause analysis in Garvey v. State Farm Fire & Casualty Co., supra, 48 Cal.3d 395. In Garvey, the plaintiffs’ home was damaged when an addition, built in the early 1960’s, began to pull away from the main structure. (Id., at p. 400.) The conflicting evidence suggested that the damage was caused by negligent construction, earth movement, or a concurrence of both. (Id., at p. 412.) The plaintiffs’ homeowner’s policy provided coverage for negligence but excluded losses “caused by, resulting from, contributed to or aggravated by any earth movement . . . .” (Id., at p. 399.) The Supreme Court reversed a directed verdict in favor of the plaintiffs, because the trial court had applied the “concurrent causation” approach articulated in State Farm Mut. Auto. Ins Co. v. Partridge (1973) 10 Cal.3d 94 [109 Cal.Rptr. 811, 514 P.2d 123], (48 Cal.3d at pp. 398-399, 401, 412.) The Supreme Court held that the Partridge concurrent causation analysis should not be applied to so-called “first party” property insurance cases. Instead the Supreme Court *1456strongly reaffirmed the application of the Sabella analysis to cases where an insured is seeking coverage against loss or damage which he himself has sustained. (Id., at pp. 399, fn. 2, 404.) The court stated: “Sabella ... set forth a workable rule of coverage that provides a fair result within the reasonable expectations of both the insured and the insurer whenever there exists a causal or dependent relationship between covered and excluded perils.” (Id., at p. 404, italics supplied.) Consequently, the Supreme Court remanded the case to the trial court so that the jury could determine coverage under & Sabella efficient proximate cause analysis. (Id., at pp. 412-413.) Once again, as in Sauer, Gillis and Premier, the Supreme Court did not look to the specific exclusionary language of the policy, but instead held that the case was governed by the general principles articulated in Sabella.

In our view, Insurance Code section 530, as interpreted by Sabella and its progeny, requires a property insurer to provide coverage whenever an insured peril is the “efficient proximate cause” of the loss.5 (See Villella v. PEMCO (1986) 106 Wn.2d 806, 816-819 [725 P.2d 957, 962-964], where the Washington Supreme Court interpreted and applied Sabella and its progeny.) Consequently, the exclusionary provisions contained in the contracts at issue are not enforceable to the extent they purport to limit the insurers’ liability beyond what is permitted by section 530 and its interpreting cases.6

*1457State Farm relies on a lower federal court case to support its contention that it may contractually exclude coverage whenever an uninsured peril is a “but for” cause of a loss. In that case—State Farm Fire and Cas. Co. v. Martin (C.D. Cal. 1987) 668 F.Supp. 1379—a federal district court judge concluded that Sabella and Insurance Code section 530 do not prevent an insurer from contractually excluding coverage where an insured peril is merely a concurrent cause of the loss.7

In Martin, the insureds argued that their home was damaged by earth movement (an excluded peril) and sulfate attack of the foundation (an insured peril). (668 F.Supp. at pp. 1380, 1383.) The homeowners policy excluded coverage “for any loss that would not have occurred in the absence of. . . earth movement . . . ‘regardless of . . . any other causes, or whether other causes acted concurrently or in any sequence with the excluded event to produce the loss.’” (Id., at p. 1382.) The federal district court concluded that this provision was sufficient to exclude coverage even if the covered peril had contributed to the loss. Rejecting the insureds’ argument that Sabella and Insurance Code section 530 preempted the exclusionary language in the contract, the court stated: “The court in Sabella was addressing the construction and applicability of section 532 of the statute. Here, the insurance policy itself expressly and explicitly includes in its language a provision which states that for certain causes, the ‘but for’ argument raised by the insurer in Sabella will apply to limit coverage. While the court in Sabella would not extend the statute so as to imply such restrictions on coverage of concurrent causes, there is nothing in the law denying the insurer the right to include such language as a term of the contract itself.” (Ibid., second italics added.)

We note first that the Martin decision appears to apply only to “concurrent causes,” and not to “efficient proximate causes” as defined in Sabella. That is, Martin holds that an insurer may contractually limit liability where *1458a covered peril is merely a “concurrent cause” of the loss; Martin does not address whether the insurer may exclude coverage when an insured peril is the “efficient proximate cause” of the loss, as that term is defined in Sabella. (See State Farm Fire and Cas. Co. v. Martin, supra, 872 F.2d at p. 321.)

Moreover, to the extent Martin purports to permit an insurer to deny coverage where an insured peril is the “efficient proximate cause” of the loss, it contradicts Sabella since “[s]uch a result would be directly contrary to the provision in section 530, in accordance with the general rule, for liability of the insurer where the peril insured against proximately resulted in the loss . . .” (Sabella, supra, 59 Cal.2d at p. 33.) In addition, Martin completely ignores Sauer, Gillis, and Premier, where the courts applied the Sabella efficient proximate cause analysis despite the presence of specific exclusionary language not present in the Sabella contract.

In short, Martin does not specifically hold that an insurer may contractually exclude coverage even though an insured peril is the efficient proximate cause of the loss. Furthermore, to the extent Martin can be read to permit an insurer to limit liability in this manner it is contrary to California law and we decline to follow it. (People v. Bradley (1969) 1 Cal.3d 80, 86 [81 Cal.Rptr. 457, 460 P.2d 129].)

Finally, we note that the result we reach should hardly come as a surprise to State Farm. In a 1985 article, one of State Farm’s assistant counsel acknowledged in a law journal article that “[a]n open question remains whether sections 530 and 532 can be successfully altered by contract language. . . . [T]he possibility cannot be ignored that the courts will declare sections 530 and 532 so entrenched in the law of property insurance that they now represent the public policy of the state and cannot be contractually circumscribed.” (Bragg, Concurrent Causation and the Art of Policy Drafting: New Perils for Property Insurers (1984) 20 Forum 385, 397-398.)

We need not, as the author suggests, find that Insurance Code section 530 is “so entrenched” that it represents the public policy of this state. The simple truth—which State Farm insists on ignoring—is that this provision is a part of the statutory law of this state and is applicable to all insurers who issue “all risk” policies. For this simple reason, the exclusions contained in the policies at issue are not enforceable to the extent they violate Insurance Code section 530. Because we reach this conclusion, we need not address appellant’s alternative argument that the exclusionary language in the policies is so unclear that it is unenforceable.

*1459B. There Is a Triable Issue of Fact That the Fire Was the Efficient Proximate Cause of the Loss.

State Farm next contends that even if we apply the Sabella efficient proximate cause analysis, the uncovered peril (earth movement) was the “efficient proximate cause” as a matter of law, and hence summary judgment was properly granted. We disagree.8

Sabella defined “efficient proximate cause” as “ ‘the one that sets others in motion’ ” and as “the predominating or moving efficient cause.” (Sabella, supra, 59 Cal.2d at pp. 31-32; Garvey, supra, 48 Cal.3d at p. 403.) Although this definition lacks precision, what is clear after Garvey is that the issue of efficient proximate causation is a question of fact for the jury. (Garvey, supra, 48 Cal.3d at pp. 412-413, fn. 11.)

Here, appellant’s geotechnical expert stated in his report that “[ajlthough water was the slide’s actuating mechanism, resistance to failure was severely impaired by destruction of the area’s natural biotechnical slope protection. On 6/28/85 the slope that failed was devastated by a fire. The destruction of the vegetation and rooting at the surface caused a[n] adverse modification to the hydrologic regime in the soil, [fl] In my opinion, if the vegetation had not been destroyed by fire, the slope failure between the houses on Via Media probably would not have occurred. Many other failures of ancient landslide deposits took place in years immediately preceding the fire, for instance during March 1983 following two consecutive years of abnormally heavy rainfall. The slope that burned in 1985 did not fail in prior years of greater rainfall. ” (Italics added.)

In our view, this expert opinion is sufficient to raise a triable issue of fact that the fire was the efficient proximate cause of the landslide.9 Based on this *1460evidence, a reasonable juror could find that the burning of the slope was the “predominating cause” or the one that set the others in motion.

Conceptually, the facts here are similar to those in Gillis v. Sun Ins. Office, Ltd.., supra, 238 Cal.App.2d 408, which was cited with approval in Garvey. (48 Cal.3d at p. 404.) As we discussed in our analysis above, in Gillis a windstorm lifted a gangway into the air and dashed it violently into a docking facility, causing the dock to sink into the water, where it broke up under the pounding of the waves. (238 Cal.App.2d at pp. 410-411, 416-417.) The insurance policy covered losses caused by the wind but excluded damage caused by water. (Id., at pp. 415-416.) The appellate court nevertheless concluded Sabella imposed liability since the fall of the gangway “was the start of a chain of events which permitted the waves to cause damage to the float. . . .” (Id., at p. 419.) According to the court, it was proper to apply the Sabella-Sauer proximate cause analysis because the wind “creat[ed] a condition that permitted natural forces, which alone may have caused no damage, to effect the damage for which recovery is sought.” (Id., at p. 421.) Similarly, it can be argued here that the fire “creat[ed] a condition that permitted natural forces, which alone may have caused no damage, to effect the damage for which recovery is sought.” (Ibid.)

Since there was a triable issue of fact concerning the “efficient proximate cause” of the slide, summary judgment was improper.

C. There Is a Triable Issue of Fact That the Landslide Damaged Insured Structures.

State Farm finally contends that the summary judgment may be affirmed on the alternative ground that Ms. Howell failed to produce evidence that the landslide had damaged insured buildings or other structures.10 We disagree.

The policies issued by State Farm insured against “direct physical loss” to the “dwellings,” “buildings,” and “other structures” on the premises.

*1461State Farm’s own investigator testified in a deposition that he had examined the property in February 1986, and determined that a recent landslide had caused “minor damage” to a kennel building. In addition, the investigator took and authenticated pictures showing that the landslide had flowed over or through a retaining wall, and had blocked or damaged private roads on the property. Significantly, the investigator was unaware of the existence or location of any boundaries between the various portions of Ms. Howell’s property covered by the five different policies.

Thus, at a minimum, there is a triable issue of fact that the landslide caused physical damage to the kennel building, the retaining wall, and the private roads on the property.

State Farm has not disputed Ms. Howell’s contention that the retaining wall and private road are “other structures” within the meaning of the policies. Instead, they contend that Ms. Howell is seeking compensation only for the loss in value of her land resulting from her inability to subdivide the property. This is untrue. Ms. Howell’s opposition to the motion for summary judgment makes it very clear that she is seeking recovery for damage to insured buildings and structures. State Farm’s argument to the contrary is based on a distortion of Ms. Howell’s testimony and ignorance of the rules governing summary judgment, which require us to view the affidavits and other evidence in the light most favorable to the nonmoving party. (Steele v. Totah, supra, 180 Cal.App.3d at p. 549; see also Molko v. Holy Spirit Assn., supra, 46 Cal.3d at p. 1107.)

In sum, we conclude there is a triable issue of fact that the landslide caused direct physical damage to buildings and “other structures” on the insured premises. Consequently, summary judgment may not be supported on this ground. Because we reach this conclusion, we need not address Ms. Howell’s alternative arguments that there is a triable issue of fact that the landslide threatens the stability of insured buildings and structures, and that State Farm waived any defense based on absence of damage.

Disposition

The judgment is reversed, and the case is remanded to the trial court for proceedings consistent with the views expressed in this opinion. Costs on appeal are awarded to appellant. (Cal. Rules of Court, rule 26(a).)

Merrill, J., concurred.

I concur with the majority’s conclusion that the trial court erred in granting summary judgment in favor of respon*1462dent. I agree that triable issues of fact exist whether the efficient proximate cause of appellant’s loss was a covered peril and whether any insured property was damaged. I also agree that, in this case, the insurer may not escape liability based on a policy exclusion which purports to be effective when an excluded event contributes in any way to a loss, even though the efficient proximate cause of the loss is covered. However, because I do not agree that Insurance Code section 530 (hereafter section 530) compels that result in every case regardless of policy language, I concur separately.

This case squarely presents an issue which was left open in the leading case in this area, Garvey v. State Farm Fire & Casualty Co. (1989) 48 Cal.3d 395 [257 Cal.Rptr. 292, 770 P.2d 704]: whether an insurer may, by employing specific and unambiguous language, vary the ordinary rules of causation that apply to property casualty insurance policies. In Garvey, our Supreme Court held the courts should apply a rule of “efficient proximate cause.” (Id., at p. 412.) Under that rule, if a particular covered peril is the efficient proximate cause of a loss, then the loss should be covered; if the efficient proximate cause is excluded, then there should be no coverage. (Id., at pp. 408, 412.) However, Garvey was expressly limited to pre-1983 policy language, and did not decide what effect, if any, insurer attempts to modify that language might have. (Id., at p. 407, fn. 6.) Here, we are asked to decide whether the rule of efficient proximate cause reaffirmed in Garvey survives an insurer’s attempt to substitute a different rule of causation clearly and unambiguously. The majority hold that section 530, as interpreted in Sabella v. Wisler (1963) 59 Cal.2d 21 [27 Cal.Rptr. 689, 377 P.2d 889] and subsequent cases, declares a rule which may not be varied by contract because it expresses the public policy of the state. (Civ. Code, § 3268 [parties to contract may not waive provisions of code in contravention of public policy].)1 I do not agree that section 530 is a legislative expression of mandatory public policy, either as enacted or as interpreted. However, I find that, under well-established rules, an insurer may not simultaneously insure against loss proximately caused by a covered peril, and exclude that same coverage if a different peril contributes to the loss in any way, however remote. I therefore concur in the judgment.

Facts

As the summary of facts in the majority opinion is complete, it is sufficient for my purpose to say that after a fire defoliated the insured *1463property, heavy rains caused landslides which resulted in damage to insured property. The policies in question insured against “accidental direct physical loss,” excepting those specifically excluded. Among the losses specifically excluded was any loss “which would not have occurred in the absence of . . . [fl] . . . landslide . . . .”2 The trial court gave summary judgment based on its conclusion that the loss was not covered.3 Appellant argues that because an issue of fact exists whether the efficient proximate cause of the loss was fire, a nonexcluded peril, summary judgment was improper. Respondent argues that in light of the clear language excluding losses that would not have occurred absent landslide or other specified events, the ordinary rule of proximate cause has no application, and coverage is excluded whether or not the fire was the efficient proximate cause of the loss.

Discussion

I. Introduction

In the following discussion, I first explain why I cannot agree that section 530 is a legislative mandate which a fortiori nullifies any attempt to draft a contrary rule. In a separate section, I discuss why I conclude that judicially declared policy requires a reversal in this case.

II. Section 530 Is Not Mandatory

Section 530 provides that “[a]n insurer is liable for a loss of which a peril insured against was the proximate cause, although a peril not contemplated by the contract may have been a remote cause of the loss; but [it] is not liable for a loss of which the peril insured against was only a remote cause.” Present section 530 was enacted with the rest of the Insurance Code in 1935. (See Stats. 1935, ch. 145, § 530, p. 510.) As passed, its history is unremarkable. It was a reenactment of former Civil Code section 2626. (Stats. 1935, ch. 145, § 2, p. 496 [sections of Insurance Code substantially the same as former provisions, intended as “restatements and continuations thereof, and not as new enactments”]; Stats. 1935, Appen., p. 2731; and see Civ. Code (1872 ed.) former §2626, p. 430.)

*1464Former Civil Code section 2626 was in turn enacted in 1872 as a part of the general codification of California law and remained unchanged between its first enactment and its transposition into the Insurance Code in 1935. (Civ. Code, § 23; see Civ. Code (1872 ed.) supra, former § 2626, p. 430; cf. present Ins. Code, § 530.) Between 1872 and 1935, there is no trace of any legislative consideration of its intent or effect. The entire legislative history of section 530 exists in the notes and commentary that appear in connection with the enactment of the 1872 California Civil Code, and its counterpart and predecessor, the Civil Code of the State of New York. Those commentaries reveal that section 530 is the statutory expression of the common law rule of causation in insurance law, that where perils combine to cause a loss, coverage is determined by looking to which of the perils was the proximate cause of the loss, and not to perils which were remote causes.4 (See Raymond & Burch’s Ann. Civ. Code (1st ed. 1872) foil, former § 2626, pp. 160-161 [code comrs.’ ed., noting that former Civ. Code, § 2626 was intended to codify the general and contemporary rule of causation in insurance cases].)

The derivation of former Civil Code section 2626 confirms the opinion of the code commissioners. The statute was taken directly from the Civil Code of the State of New York, and restated section 1431 of that code.5 (See, e.g., Revised Laws of the State of Cal., Civ. Code (1871 ed.) p. 457 [noting origin of then-proposed Cal. Civ. Code, § 2626 as N.Y. Civ. Code, § 1431]; Civ. Code of the State of N.Y. (1865 ed. [“Field Draft”]) supra, fn. 3 foil. § 1431, pp. 421-422; and see Raymond & Burch’s Ann. Civ. Code (1872 ed.) supra, note foil, former § 2626, pp. 160-161.) The notes of the drafters of the New York Civil Code and the California code commissioners reiterate that the New York statute was a legislative expression of the general rule of causation in insurance, that where two or more perils concur to cause a loss, coverage either stands or falls depending on which of the perils was the primary cause of the loss.

Since present section 530 is the direct descendant of the 1865 New York Civil Code section 1431, the most that can be said of the intentions of the *1465Legislature that adopted that section is that its motive was that common to the 19th-century codification movement as a whole: to restate the principles of common law in a consistent, accessible form, or, in the words of the New York code commissioners, to “[arrange] these rules in a manner which will be approved by the scientific student, while it will help the lawyer and the citizen to an easier if not a better knowledge of the law.” (Code Comrs.’ Final Rep., Civ. Code of the State of N.Y. (1865 ed. [“Field Draft”]) supra, p. vii.)

The question remains, however, whether the common law rule of causation in insurance coverage disputes, restated in the early codes and in present section 530, was or is an embodiment of public policy which overrides specific contract language. A review of contemporary and early cases reveals that at least as it existed at the time the statutes were drafted, it was not. In Mathews v. Howard Ins. Co. (1854) 11 N.Y. 9, one of the principal cases relied on by the drafters of the New York Civil Code, a concurring opinion noted the rule that although “the maxim causa próxima non remota spectatur is the general rule for determining the liability of the underwriter, the parties may, nevertheless, by express stipulation, agree upon another rule of responsibility. [Citation.]” (Id., at p. 25, fn. 3 (cone. opn. of Parker, J.), citing Savage v. The Corn Exchange Fire and Inland Navigation Insurance Co. (1858) 4 Bosw. 1, 19; and see 1 Parsons on Marine Insurance, Excepted Risks and Losses, supra, at pp. 622-623 [noting in discussion of proximate cause that policy exclusions may always be made “at the pleasure of the parties” and must be construed so as to carry into effect the intention of the parties]; St. John v. American Mutual Ins. Co. (1854) 11 N.Y. 516, 519-520 [cited by drafters of predecessors to present Ins. Code, § 532, and holding that an excluded remote cause defeated coverage where loss would not have occurred but for a remote peril]; and see Williamsburgh City Fire Ins. Co. v. Willard (1908) 164 Fed. 404, 408-409 [finding waiver provisions of Civ. Code, § 3268 applicable to causation rule of former Civ. Code, § 2628, now Ins. Code, § 532].) At least as originally drafted, section 530 was an expression of a rule of general law, not a mandatory provision to be read into private contract.

Further, no case interpreting section 530 has ever held it to be more than it was originally intended to be. Although the majority rely on Sabella v. Wisler, supra, 59 Cal.2d 21, neither that case nor any other finds a mandatory provision in section 530. Sabella itself, though silent on the point, depends for its holding on a series of cases which rely on interpretive principles rather than legislative direction. (See Sabella, supra, at pp. 32-33, discussing Brooks v. Metropolitan Life Ins. Co. (1945) 27 Cal.2d 305, 309-310 [163 P.2d 689] [better rule of interpretation is that coverage intended where insured peril is proximate cause of loss]; Hanna v. Interstate B. M. *1466Acc. Assn. (1919) 41 Cal.App. 308, 310-311 [182 P. 771] [distinguishing cases involving different policy language and applying general rule of causation]; Edgerton & Sons, Inc.v. Minneapolis Fire & Marine Ins. Co. (1955) 142 Conn. 669, 673-674 [116 A.2d 514, 516-517] [interpreting exclusion in favor of insured and applying general rule]; Princess Garment Co. v. Fireman’s Fund Ins. Co. (6th Cir. 1940) 115 F.2d 380, 382-383 (1940 [noting rules of interpretation of insurance contracts and stating that “on such contracts [of insurance] it is assumed that it was the intention of the insurer to indemnify the insured against losses where the cause insured against was the proximate or immediate means or agency causing the loss”].) These cases demonstrate that the rule in Sabella is based on the related ideas that, first, a reasonable insured expects that a loss will be covered if it is proximately caused by a covered peril, even though other remote and excluded causes may concur in producing the loss, and second, to construe the policy to defeat that expectation would be contrary to the purpose of the insurance itself, i.e., provision against economic loss from certain classes of perils. Although I endorse that holding and believe it should be followed here, I do not agree that it is a declaration of a mandatory public policy. Any such declaration must be left to the Legislature. (Cf. Ins. Code, §§ 2071 [mandatory fire insurance form of policy], 11580.1 et seq. [requiring certain provisions to be included in automobile insurance policies].)

Because I do not find that section 530 expresses any mandatory public policy, I do not agree with the reasoning adopted by part A of the majority opinion. However, for the reasons set out in the following section, I concur in the judgment.

III. Policy Interpretation

In the present case, respondent insured appellant’s property against all direct accidental losses save those specifically excluded. Under the rule of Sabella v. Wisler, supra, 59 Cal.2d 21, as restated and reaffirmed in Garvey v. State Farm Fire & Casualty Co., supra, 48 Cal.3d at pages 402-403, it would appear that if an accidental loss to the insured property occurred, and if the efficient proximate cause of that loss was not excluded, there would be coverage. (Cf. Brodkin v. State Farm Fire & Casualty Co. (1989) 217 Cal.App.3d 210, 217-218 [265 Cal.Rptr. 710] [where all asserted efficient proximate causes of loss were excluded, summary judgment for insurer proper].) Here, appellant argues that a triable issue exists whether the efficient proximate cause of the loss was the fire which defoliated her *1467property or the ensuing landslide.6 Fire is not among the perils excluded from coverage by the policies in question. However, respondent asserts that because the claimed property damage would not have occurred absent the landslide, the loss is excluded even if we assume that the efficient proximate cause of the loss was covered. After careful review, I have concluded that respondent’s position is unsupportable under long-established principles.

It is true, as respondent strongly urges, that subject to legislative regulation, an insurer has the right to select the risks which it will insure and to exclude those that it will not. (See, e.g., Continental Cas. Co. v. Phoenix Constr. Co. (1956) 46 Cal.2d 423, 432 [296 P.2d 801, 57 A.L.R.2d 914].) If the meaning of the exclusion is unambiguous and does not mislead, “the plain language of the limitation must be respected. [Citations.]” (Ibid.) This rule is followed in a variety of circumstances. (See, e.g., National Ins. Underwriters v. Carter (1976) 17 Cal.3d 380, 384-386 [131 Cal.Rptr. 42, 551 P.2d 362] [where aircraft insurance policy provided in exclusions and on declarations page that policy was effective only if persons specified by declarations were piloting aircraft, no coverage where airplane crashed while piloted by permissive user not so named]; Argonaut Ins. Co. v. Transport Indem. Co. (1972) 6 Cal.3d 496, 508 [99 Cal.Rptr. 617, 492 P.2d 673] [secondary insurer not liable where policy endorsement provided that policy was not applicable if loss was covered by any other policy, primary or secondary]; State Farm Mut. Auto. Ins. Co. v. MacKenzie (1978) 85 Cal.App.3d 727, 732 [149 Cal.Rptr. 747] [no coverage for permissive use of nonowned auto where neither insurance policy nor statute provided for or required such coverage, noting right of insurer to limit coverage where not “ ‘. . . prohibited by public policy or statute. [Citation.]’ [Citations.]”])

Here, the exclusion relied on by the insurer is clear; it will not pay a loss which would not have occurred but for a landslide. Thus, at first glance it seems that we should affirm the judgment of the trial court, since it is obvious that the loss in this case would not have occurred but for a landslide. However, as I discuss below, policy exclusions which are asserted to make otherwise applicable coverage turn on the presence or absence of remote causes have been treated differently than other kinds of exclusions.

In a line of cases beginning at least as long ago as the first judicial interpretation of what is now section 530, California courts have consistently found that where a loss is the efficient proximate result of a covered peril, an insurer may not defeat coverage by discovering a remote and excluded cause somewhere in the chain of causation, even where the policy contains *1468clear language that would support the insurer’s interpretation. That finding is based on two related principles applicable to all insurance contracts: first, that ’’the policy or its endorsements cannot be so interpreted as to become meaningless, or to withhold coverage which the [layperson] would normally expect from it . . . ,” and second, that “ . . [t]he courts will not sanction a construction of the insurer’s language that will defeat the very purpose or object of the insurance. [Citations.] ....’” (City of Santa Monica v. Royal Indent. Co. (1958) 157 Cal.App.2d 50, 54 [320 P.2d 136].)

I begin with the first California case which construed the predecessor of present section 530, former Civil Code section 2626, Pacific etc. Co. v. Williamsburg (1910) 158 Cal. 367 [111 P. 4]. In Pacific, the insured property was destroyed by fire as a consequence of the 1906 San Francisco earthquake. (Id., at p. 369.) It was undisputed that the fire started in another building as a direct consequence of the earthquake, and then spread from structure to structure, finally destroying the insured property. (Ibid.) The insurer denied coverage based on an exclusion which recited that it would not be liable for “ ‘ “. . . loss caused directly or indirectly by invasion, insurrection, riot, civil war or commotion, or military or usurped power, or by order of any civil authority; or for loss or damage occasioned by or through any . . . earthquake . . . ; or by theft; or by neglect of the insured . . . (Ibid.) The insurer argued that under former Civil Code section 2628 (now Ins. Code, § 532), it had no liability on the policy, because the fire would not have occurred but for the earthquake. (Pacific, supra, at pp. 372-373.) The court7 disagreed, and stated that former Civil Code sections 2626 and 2628 (now Ins. Code, §§ 530, 532) must be read together, and that “ ‘. . . [n]o rule of construction would justify us in holding that the section [now Civ. Code, § 532] broadens the scope of the original exception [i.e., policy exclusion]. The exception was made in the contract of insurance, and must be held to control, as it was the deliberate words of the parties used by them in making the contract. . . .’” (Pacific, supra, at p. 373.) The court went on to reject the insurer’s argument that the existence of an excluded peril in the chain of causation defeated coverage. The court concluded that the insurer’s theory would lead to a holding that if “ ‘. . . a very slight shock of earthquake had upset a lamp and thus set fire to a building, and this building should communicate the fire to an adjoining building, and thence from building to building until the whole city had burned, not a dollar of insurance could be recovered if the policies each contained the clause under discussion. We do not think that either the plaintiff [insured] or the defendant [insurer] ever contemplated making any such contract. It would certainly require the most clear and unambiguous language to justify *1469any court in placing a construction upon a policy of insurance that would have such result. . . (Id, at p. 372.)

Two principles are at work in Pacific. First, the court emphasizes that its decision is based on the language of the insurance contract, not on any statutory or judicial policy. Second, and more important here, the court employs the general rule that the intentions of the parties govern interpretation of the contract to void arguably clear exclusionary language. Even though the fire in question was indirectly caused by earthquake, the court chose to ignore the language of the contract in favor of the presumed intention of the parties. This method of avoiding the effect of apparently clear language has been duplicated in virtually every California case to consider the question.

In the leading case, Brooks v. Metropolitan Life Ins. Co., supra, 27 Cal.2d 305, the spouse of the insured husband filed suit on an accidental death policy after he died in a fire. (Id., at p. 306.) The evidence showed that the insured was suffering from an incurable cancer and had been confined to his room. (Id., at pp. 307-308.) He was discovered dead after a fire in the room, and there was expert testimony that he had died of his burns. (Id., at p. 309.) The policy insured against death or injury “ ‘. . . caused directly and independently of all other causes by violent and accidental means,’ ” and excluded any loss which was . . caused wholly or partly, directly or indirectly, by disease or mental infirmity or medical or surgical treatment therefor.’” (Id., at p. 306.) The defending insurer argued among other things that because the deceased insured would not have died in the fire had he not been an invalid, the loss was partly caused by sickness and was not payable. (Id., at p. 309.) Rejecting case law to the contrary, the court held that even where policy language in an accidental death policy precluded coverage if the death was caused “ ‘partly’ ” or “ ‘indirectly, by disease or mental infirmity,’ ” a “preexisting disease or infirmity will not relieve the insurer from liability if the accident is the proximate cause of death; and that recovery may be had even though a diseased or infirm condition appears to actually contribute to cause the death if the accident sets in progress the chain of events leading directly to death, or if it is the prime or moving cause. [Citations.]” (Id., at pp. 306, 309-310.)

Brooks's rationale is probably best summed up in Kelly v. Prudential Ins. Co. (1939) 334 Pa. 143 [6 A.2d 55], one of the two cases on which Brooks relies. In Kelly, an accidental death case, the insurer argued it was improper to instruct the jury that a preexisting illness which contributed to death would not bar liability so long as an accident was the proximate cause of death. (Kelly, supra, at pp. 149-150 [6 A.2d at p. 58].) Explaining its decision to the contrary, the court stated that “ ‘ “. . . it is the duty of courts to *1470give such construction to a policy, if the language used fairly admits, as will make it of some substantial value and carry out the intention expressed therein that liability is incurred where death occurs from accidental injury. . . (Id., at p. 151 [6 A.2d at p. 58].) The policy language in Brooks, specifying that coverage would be defeated if the death was caused “ ‘wholly or partly’ ” by disease, could hardly be clearer, and it is similarly clear that the Brooks court refused to exclude coverage even while admitting that the death had been “ ‘partly’ ” caused by the insured’s preexisting disease. It is significant in this context to note that the authorities which were specifically rejected in Brooks supported a contrary view; that where the policy excluded death caused partly by disease, there would be no coverage unless the accident which killed the insured would have been fatal to a healthy person, or put another way, if the death would not have occurred but for the preexisting disease, there would be no coverage. (Brooks v. Metropolitan Life Ins. Co., supra, 27 Cal.2d at p. 309.) In addition, Brooks relies on discussion of the point in 6 Couch, Cyclopedia of Insurance Law (1930) section 1249, page 4569, in which the author notes that “there is considerable authority to the effect that preexisting disease does not render the exemption clause applicable, even though it concurred with or assisted the injury in causing death. For instance, the existence of a pre-existing disease, or predisposition to disease, does not preclude recovery where the accident is the proximate or efficient cause of death . . .” (fn. omitted), even though the condition or disease may have contributed to the death.

Brooks rejects clear policy language in favor of the general rule that losses proximately caused by covered perils are covered. It stands for the proposition that California courts will resolve in favor of the insured the inherent conflict between insuring language which purports to cover certain perils and excluding language which would defeat that coverage even though a covered peril directly, efficiently, and proximately caused the loss.

Other cases have reached the same result by identical means. (See, e.g., City of Santa Monica v. Royal Indem. Co., supra, 157 Cal.App.2d at p. 54 [application of exclusion would have been contrary to purpose of insurance]; Hanna v. Interstate B. M. Acc. Assn., supra, 41 Cal.App. at pp. 310-311 [accidental death policy]; Berry v. United Com. Travelers (1915) 172 Iowa 429, 432-433 [154 N.W. 598] [finding exclusion ambiguous; cited in Hanna]; Travelers’ Ins. Co. v. Murray (1891) 16 Colo. 296, 305-306 [26 P. 774] [finding exclusion ambiguous; cited in Hanna; and noting that insurance contracts must be construed to “give the insured the benefit of his [or her] contract and consideration for the premium paid . . .”]; Atlanta Accident Asso. v. Alexander (1898) 104 Ga. 709, 712-713 [30 S.E. 939] [finding exclusion ambiguous; cited in Hanna; and noting that contrary construction *1471of the policy “would make the policy very misleading to the insured . . .”]; and see Annot., Insurance—Coverage—Disease After Accident (1937) 108 A.L.R. 6-37 [collecting cases].)

Brooks is especially significant because it became the principal authority for the holding in Sabella v. Wisler, supra, 59 Cal.2d 21, which our Supreme Court has recently stated to be a model of correct reasoning in this area. (See Garvey v. State Farm Fire & Casualty Co., supra, 48 Cal.3d at p. 404.) In Sabella, the insured’s house was severely damaged when, as a result of the builder’s failure to compact the ground properly, a sewer line broke, saturated the ground, and caused the house to settle. (Sabella, supra, 59 Cal 2d at p. 26.) In the insured’s suit against the insurer, the insurer argued that it was not liable on the policy because the policy excluded any loss “ . . by . . . settling, cracking, shrinkage, or expansion of pavements, foundations, walls, floors, or ceilings ....’” (Ibid., italics omitted.) The insurer also claimed that Insurance Code section 532 (hereafter section 532) freed it from liability, because the damage to the house would not have occurred but for the settlement. (Id., at p. 33.) The court found that the policy clearly excluded losses caused by “ ‘settling,’ ” and that the exclusion could not have been misunderstood by the insureds. (Id., at pp. 30-31.) However, it concluded that because the proximate or “moving efficient” cause was the breaking of the sewer pipe as the result of negligence, coverage therefore existed as a matter of law. (Id., at pp. 31-32.) The court relied on the familiar principle that “ ‘. . . where there is a concurrence of different causes, the efficient cause—the one that sets others in motion—is the cause to which the loss is to be attributed, though the other causes may follow it, and operate more immediately in producing the disaster.’ ” (Ibid., quoting 6 Couch, Cyclopedia of Insurance Law, supra, at § 1466, pp. 5303-5304.)

In its explanation of the operation of this principle, the Sabella court relied primarily on Brooks, and quoted directly from that portion of Brooks in which the court held that an exclusion based on the concurrence of a remote cause would not prevent coverage where the loss was proximately caused by a covered event. (Sabella, supra, 59 Cal.2d at p. 32, quoting Brooks v. Metropolitan Life Ins. Co., supra, 27 Cal.2d at pp. 309-310.) Although the court did not say in so many words that it would refuse to interpret policy language that would have defeated coverage based on remote causes, its reliance on Brooks, in which that principle was the basis for the holding, is an indication of the court’s direction. Further, the other authority cited by the court supports the conclusion that the court was adopting a rule of construction which was inconsistent with any policy interpretation which would vary the general rule of causation. (See Hanna v. Interstate B. M. Acc. Assn., supra, 41 Cal.App. at pp. 310-311 [holding that where accident was the proximate cause of death, coverage existed, and *1472citing cases that support the general proposition that policies will not be construed so as to defeat the purpose of the insured in obtaining the insurance]; Edgerton & Sons, Inc. v. Minneapolis Fire & Marine Ins. Co., supra, 142 Conn, at pp. 673-674 [116 A.2d at pp. 516-517] [finding coverage over dissent where exclusion clear and unambiguous].)8

After Sabella, California courts continued to apply the principle expressed in Brooks and Sabella to cases raising the causation question.9 (See, e.g., Premier Ins. Co. v. Welch (1983) 140 Cal.App.3d 720, 725 [189 Cal.Rptr. 657]; Gillis v. Sun Ins. Office, Ltd. (1965) 238 Cal.App.2d 408, 415-420 [47 Cal.Rptr. 868, 25 A.L.R.3d 564]; Sauer v. General Ins. Co. (1964) 225 Cal.App.2d 275, 278-279 [37 Cal.Rptr. 303].)

Unlike other modern cases, in which the question was apparently not raised, the problem under consideration here was directly confronted in Gillis v. Sun Ins. Office, Ltd., supra, 238 Cal.App.2d 408, cited with approval in Garvey v. State Farm Fire & Casualty Co., supra, 48 Cal.3d at page 404.10 In Gillis, the insured operated a floating pontoon dock at Sausalito, *1473California. (Gillis, supra, at p. 410.) The dock was connected to another dock by a movable gangway. (Ibid.) The insurance policy provided coverage against “ . . direct loss by windstorm [naming other perils], except as hereinafter provided . . (id., at p. 415), but excluded any . . loss caused by, resulting from, contributed to or aggravated by any of the following—(a) Flood, surface water, waves, tidal water or tidal wave, overflow of streams or other bodies of water, or spray from any of the foregoing, all whether driven by wind or not ....”’ (Id., at p. 416.) During a large storm, high winds caused the gangway to rise and fall violently on the floating dock, causing the dock to break up and sink. (Id., at pp. 410-411.) It was conceded that wave action had contributed in some way to the destruction of the dock. (Id., at p. 416.) The insurer argued that because wave action had contributed to the loss in some proportion, the loss was excluded. (Id., at pp. 418-419.) In addition, the insurer specifically argued that the exclusionary language “void[ed] the application of the usual rule of proximate cause, and clearly and unequivocally expressed] the intent that regardless of the initiating or proximate cause of the damage, no recovery can be had if an excluded cause has any part, no matter how small, in causing the damage.” (Id., at p. 419.) The court rejected that position. It noted that although the wind did not by itself destroy the float, it “create[d] a condition that permitted natural forces, which alone may have caused no damage, to effect the damage for which recovery is sought . . . ,” and held that the case was within the efficient proximate cause rule of Sabella. (Gillis, supra, at p. 421.) The court, after some discussion of hypothetical situations, then addressed the insurer’s principal argument. It is useful to quote the passage in full.

“The third category presents the problem of whether the insured is barred whenever the excluded cause contributes to the loss [as the insurer *1474had argued]. Where the windstorm causes the initial damage and leaves the insured property in such a vulnerable state that a subsequent force of nature causes loss, the insured has generally been permitted to recover. For example, in [citation], the court held that notwithstanding the erroneous finding of the jury that the building was not damaged by snow, an expressly excluded cause of loss, the judgment on the verdict for the insured should be affirmed because windstorm damage which preceded the snow had weakened the roof and was the efficient cause which set the contributing cause in motion. . . . [Citations.] Where the property is picked up by the wind and deposited where it is subjected to further damage by high water, the insurer cannot defeat recovery by resort to a clause which excludes the latter peril. [Citation.] In the case last cited the opinion recites: ‘Common experience and understanding suggest that when personal property of this nature is blown into a body of water that some water damage will result before the property can be recovered. It is fair to assume that under these facts this incidental water damage was within the terms of the policies and contracts of the parties at the time they were made.’ [Citation.] [([] So here common experience and understanding[11] suggest that when personal property of the nature of a floating dock is damaged so that its buoyancy is affected, some water damage will result.” (Gillis v. Sun Ins. Office, Ltd., supra, 238 Cal.App.2d at p. 423.)

Thus, even where the Gillis court was confronted with a policy exclusion which it recognized was designed to vary the traditional rule of causation, it refused to do so. Its refusal was based on the premise that the parties to the insurance contract had agreed and expected that losses caused by wind would be covered, and that their expectation at the time of contracting should not be defeated by the existence of a remote cause acting in concurrence with the covered cause. In this context, the words of another court on the issue of causation bear repeating. “It is known that the wind blows; that in certain latitudes and at a certain temperature ice forms, and later melts when the temperature rises; that water will saturate and soften the soil; that flowing streams carry silt, the amounts whereof vary as the square of the velocity, and when the velocity decreases the silt is deposited on the bed of the stream. None of these things was due to the act of any person, nor can any of them be prevented by any reasonable human means. It is merely nature acting as it always has, and always will act. It seems to us that when a given act is such as to put in force a normal law of nature, which in conjunction with the original act done, produces a harmful result, such result is necessarily a proximate cause [s/c] of the act done.” (United States v. Chicago, B. & Q. R. Co. (8th Cir. 1936) 82 F.2d 131, 137 [106 A.L.R. 942].) This reasoning underlies the rule in Gillis that where the insurer has *1475issued a policy protecting the insured against the proximate results of a given peril, and a loss results directly from a force of nature set in motion by that peril, there should be coverage, because the loss was that which was contemplated by the parties when they made the contract. Coverage in such a case should not be defeated because the insurer has inserted exclusionary language designed to avoid liability even though an event has occurred which the insured intended to guard against.

Thus, where an insurer attempts to interpose a contributing remote cause as a bar to coverage, Brooks, Sabella, and Gillis look to the expectation of the parties that the insurer must pay when a covered peril is the efficient proximate cause of a loss. They use the general rule that gives effect to those expectations over conflicting contract terms even where, as in the present case, the language relied on by the insurer is clear and unambiguous. At bottom, the rule applied in these cases is a recognition that an exclusion based on remote causation, if given routine effect, could render a policy valueless almost at random. For example, it is conceivable under the policy language at issue that the insurer could claim that its earth movement exclusion would bar recovery if a negligently operated bulldozer leveled the insured property, for clearly, such damage could not occur in the absence of some earth movement.12 It was to avoid such absurd results that the rule of efficient proximate cause was devised.

To the extent that Garvey v. State Farm Fire & Casualty Co., supra, 48 Cal.3d 395, bears on the specific issue at bench, it supports the conclusion that once it is established that the efficient proximate cause of a loss is covered, even a clear exclusion based on remote causation will not be given effect. First, Garvey expressly validates the holdings in Brooks, Sabella, Gillis, and similar cases. (Garvey, supra, at p. 404.) Second, like those cases, the keystone of Garvey's analysis is “the reasonable expectations of the insurer and the insured . . . .” (Id., at p. 408.) Though Garvey's focus was on the reasonable expectation of the insurer that it would only be held liable where a covered peril was the efficient proximate cause of the loss, it follows from that premise that it is equally reasonable for the insured to expect that, under the same circumstances, coverage would exist even though an excluded event might have contributed to the loss. (See ibid.) Thus, I conclude that Garvey is consistent with the analysis I adopt here.

*1476 Conclusion

The principle most fundamental to the interpretation of insurance policies is that which focuses on the expectations of the parties to the contract, since all insurance is at bottom an attempt to provide against unknown future events. No reasonable person would pay for insurance against some future peril if it were possible for the insurer to avoid liability by discovering an excluded peril somewhere in the chain of causation. As one court pointed out, a policy interpretation which would allow that result would be “in direct contradiction of the terms and purpose of the insurance . . . . [ft] ... To say that such was the intent of the defendant company in executing the policy would be to charge it with fraud.” (Moore v. Insurance Co. (1912) 158 N.C. 305, 306-3070 [73 S.E. 1002-1003] [rejecting insurer’s attempt to defeat accidental injury coverage where coverage excluded if disability resulted “ ‘wholly or in part’ ” from paralysis, and the insured was paralyzed in a train wreck].) I would therefore hold that where an insurer chooses to insure against the direct and proximate results of a certain peril, it may not rely on the concurrence of an excluded cause to deny coverage. I reach this conclusion based on the inherent inconsistency between policy language which clearly is intended to protect against certain perils, and language which would defeat coverage even though those same perils have directly resulted in a loss. In such a case, an exclusion which relies on a remote cause must give way to the dominant purpose of the insurance, in line with the reasonable expectations of the insured that he or she will be protected against loss, and those of the insurer, that upon the occurrence of the anticipated event, it will be liable on its policy.

In the present case, the insurer has chosen to provide against all risks of physical loss save those excluded. In this review on appeal from summary judgment, I join the finding of the majority that a triable issue of fact exists whether the efficient proximate cause of this loss was fire, a nonexcluded risk, or landslide, an excluded risk.13 It would be unfair to the insured, and against the reasonable expectation of the insurer, to find as a matter of law that even though a covered peril was the efficient proximate cause of the loss, no coverage exists because an excluded peril may have contributed in some way, however slight, to the loss.

For the above reasons, I concur in the judgment.

Respondent’s petition for review by the Supreme Court was denied June 21, 1990.