(dissenting).
With all deference I must dissent from the opinion of the court, because I am convinced that it extends the already wide field of judicial rewriting of private contracts.
Faced with the grammatical rule that a qualifying phrase refers to its nearest antecedent, unless the intention of the parties appears otherwise, the court does not deny that the language of Exclusion E, standing alone, has a definite unambiguous meaning, namely that if a loan or loans totalling more than $3500 should be made to any one borrower, any loss resulting therefrom is excluded from the policy. On the contrary, the court seeks to avoid the effect of this rule by finding a meaning inconsistent with this, expressed in other parts of the policy; or to find some sort of ambiguity in the whole policy, though none is found in Exclusion E itself. To find what Exclusion E means, the court shuns like the plague considering what it says; a studied avoidance of the obvious, and a detailed analysis, instead, of the most uncertain and apocryphal indications of the parties’ intention is the means whereby the court concludes that the exclusion means something other than what it says. At the point where the court says, “Considering the policy as a whole, it seems to us that the more reasonable construction is that the loss resulting from loans to any one borrower was limited to $3500.00, not that loans exceeding such sum were excluded in their entirety from the coverage of the policy”, it seems to me that the court has not really considered the policy as a whole, but the most far-fetched portions of it, and then drawn erroneous inferences from them.
While recognizing the literary ability which enables the court, with conviction which I do not doubt, to draw substantial conclusions from matters which seem of no substance to me, I am constrained to object that in ascertaining the intention of the parties some considerable weight ought to be given to the language of Exclusion E, the precise clause that is before us for application, and that no matters even approaching its importance have been considered in the court’s opinion. The court, so it seems to me, needs a reminder that what they are supposed to be seeking is the intention objectively *211manifested in the entire policy; but the court appears to be so occupied in the search for minute clues, that it overlooks the real nub of the matter, namely the grammatical meaning of Exclusion E.
Now, the matters which the court regards as significant indications that the parties intended the construction adopted by the court, may fairly be summarized as follows:
(1) If the construction urged by appellant is correct, the policy does not contain a monetary limit to claims for attorney’s fees and collection costs indemnified in any particular loss; however, one would expect to find such a limitation in the policy.
(2) If the construction urged by appellant is correct, the premiums paid for the excluded loans would be forfeited to the insurer.
(3) The policy provides for priority of the insured’s claim for an mindemnifled portion of a loss over the insurer’s subrogation rights; the policy contains express provisions for partial indemnification where the retail market value of the loan security is less than the amount of the loss, and where some loan payments are due more than 36 months after the making of the loan ;1 from this it is a justified inference that the parties intended partial indemnification also for losses resulting from loans to one person aggregating more than $3500, in spite of the contrary language of Exclusion E.
Now, it is my belief that these matters do not express any intention contrary to the literal meaning of Exclusion E; and certainly the unambiguous language of that exclusion must be given effect unless a contrary intention is expressed elsewhere in the policy. And there is no ambiguity — so far as is material here— in the policy, once we give due consideration to Exclusion E in our deliberations. However, the court seems to assume gratuitously that Exclusion E is equally susceptible to two interpretations, and that the three minutiae stated above tip the scales in favor of showing that the intention of the parties was contrary to the intention expressed in Exclusion E. So reasoning, the court proceeds to say that the manifested intention of the parties was contrary to what they said there, justifying the court in ignoring the grammatical rule in interpreting Exclusion E. But in concluding that it might properly ignore the grammatical signification, the court first assumed, so it seems to me, that the clause was equally susceptible to two meanings, an assumption correct only if it is proper to ignore the grammatical rule. The court’s reasoning therefore seems to me circular.
Let us examine one by one the matters held by the court to indicate an intention contrary to the literal meaning of Exclusion E:
(1) The court says one would expect to find in the policy, a monetary limit to the attorney’s fees and collection costs for which insured was indemnified in any one loss. Now, by what rule of construction recognized in Georgia, the place of this contract, can we import into the agreement as an aid to the finding of an ambiguity what the court might expect to find in such a contract?
But even if we had the right to test the intent of the parties by what we think they should have embodied in their contract, this is not a persuasive point, for there is clearly a limit of the amount of recovery.
When one examines Paragraph 2 of the “Agreements, Limitations and Conditions” with care, the argument loses all of its cogency. For by that Paragraph’s terms, such fees and costs were indemnified only to the extent that they were incurred by insured before acquiring reasonable grounds of belief of the existence of a claim by a third person to a superior *212title in the property; that is, insured is indemnified for such expenses only up to the time it had reason to believe the insurance policy covered the loss. Thus the self-interest of the insured in keeping its unindemnified expenses to a minimum could be relied upon to protect the insurer from the payment of disproportionate or unreasonable collection costs, the indemnity for such costs ceasing as soon as the insured had reasonable ground to believe the loss is covered by the insurance. Since the literal interpretation of Exclusion E, as the court concedes, would result in a most effective limitation of the insured’s claims for defaulted loans, and the indemnified collection costs-could be counted on to be proportionate and predictable in the actuarial sense, there seems to me no sound reason to infer that the parties intended a specific monetary limit to the collection costs indemnified.
(2) The occasional forfeiture of the premium which would follow from interpreting Exclusion E-as it is written, is no more a proper ground to infer that Exclusion E means something other than what it says, than it is to infer that the parties intended to -provide for a refund of the premium if paid by mistake.2 But after all, forfeiture of premiums in the sense used here, is a common enough phenomenon in insurance contracts. If the insured fails to-, file a timely proof of loss as required in an unambiguous clause,- is the resulting forfeiture a basis to infer that-the parties.meant something different? I trust that the court would answer no to my rhetorical question, but after reading its opinion I cannot be sure. Furthermore, a forfeiture may result just as certainly from the court’s construction as from the literal interpretation. Suppose the insured makes three $3500.00 loans to X, paying the insurer the agreed premium on each loan. X repays nothing, sells the security, and a,bsconds. The insured recovers nothing on the policy according to my view, $3500 according to the court’s. But is the forfeiture any less real if the court’s construction is adopted? In my opinion, this argument advanced by the court is purely ad hominem.
(3) The third argument is not even plausible as I understand it. This argument is of the form:
“Some losses are partially unindemnified. Situations A, B and C give rise to partially unindemnified losses. Therefore, situation D gives rise to a partially unindemnified loss.”
This argument is about as valid as the following, which is based on the same type of inference:
“Some men are Greeks. Homer, Socrates, and. Demosthenes are Greeks. Therefore, Charlemagne is a Greek.”
I am firmly of the opinion that not only is Exclusion E in itself unambiguous, but also that there is no significant indication of any intention contrary to its .plain grammatical meaning, to be found elsewhere in the policy. The grammatical rule that modifying phrases relate to the nearest antecedent unless a contrary intention is manifested, does not permit the finding of two meanings of Exclusion E, whereof the more favorable to the insured must be adopted. Resting firmly in the conviction that there is no ambiguity, I think it is appropriate nevertheless to point out some common-sense observations about the purpose of Exclusion E, 'neither to bolster my conclusions by unnecessary make-weights, nor to weaken them by suggesting there may be doubt as to the meaning, but simply to show the superficiality of what to me seem trivial' matters which the court says manifest the true intention of the parties, which the court makes the basis of ■its decision.
*213This policy of insurance is clearly a device to shift a certain class of risks to the insurer; to be specific, the risk that a consumer may fraudulently resell consumer goods which he is purchasing, on ordinary installment terms, before he has fully paid for them. Now it is evident that this is a kind of risk which can be estimated by actuaries and become the subject of insurance. It is also evident that it is a risk differing substantially from the risk that a retail merchant might so resell goods, or the risk that a person buying on extremely long terms might do so, or the risk that a consumer who has bought far more than he will ever consume might do so. The common sense of the actuarial situation is so far as possible to eliminate those risks from coverage in their entirety, not simply in part. That is what Exclusion F does, in excluding “any loss resulting from any loan due and payable more than thirty-six (36) calendar months after the making thereof.” Not only would it be unpermissible grammar to read this as “any loss (resulting from any loan due and payable) more, than thirty-six months after the making thereof,” but it would be a flagrant disregard of the. actuarial purpose of the exclusion; the length of the loan is the significant risk factor, and not the date of the loss. Here, the factor which is significant to those who must calculate the risk, is clearly the total amount of the loans to a single person. We all agree to this, I suppose. Now, if the risk of loss on those loans aggregating more than $3500 to one person is qualitatively different and relatively greater than the normal risks insured against, would it be sensible to exclude part of that risk, or all of it? Would a sensible life insurance company include heart patients in its ordinary rate classifications up to a maximum of $3500 ? Or would it exclude such persons entirely from its normal rate schedules ?
In interpreting the policy as a whole to find the parties’ intention, then, not only does the court ignore the language of Exclusion E, but it also glosses over the common sense purpose of the insurance. To disregard the grammatical meaning of that exclusion, however, there must be an inconsistent expression of intention elsewhere in the document itself. Certainly the court’s opinion points out no inconsistent expression of intention; at face value the matters pointed out are no more than obscure hints, and on close scrutiny mean nothing. On this point, Williston says:
“The freedom of interpretation permissible is, however, necessarily limited by the principle that unexpressed intention is of no legal effect. The reason for interpolating, omitting, or disregarding specific words is that in the remainder of the writing an intention is expressed which makes it evident that particular words were erroneously used.” 3 Williston, Contracts § 619 (Rev. Ed.); see also O. W. Holmes, “The Theory of Legal Interpretation,” 12 Harv.L.Rev. 417.
The court’s opinion conjectures that the insured “might naturally assume” that the language of Exclusion E meant something different from what the rules of grammar say it means. That kind of supposition of what might have been the insured’s secret state of mind is not the sort of thing which renders an integrated contract ambiguous. The parol evidence rule forbids that. See 9 Wigmore, Evidence § 2458 et seq. (3d Ed.); Restatement, Contracts § 230, Comment b:
“They have assented to the writing as the expression of the things to which they agree, therefore the terms of the writing are conclusive and a contract may have a meaning different from that which either party supposed it to have.”
See also Reese v. Fidelity Mutual Life Ass’n., 111 Ga. 482, 485, 36 S.E. 637; Maddox v. Life & Casualty Ins. Co., 79 Ga.App. 164, 173-174, 53 S.E.2d 235; Provident Life & Accident Ins. Co. v. Anderson, 4 Cir., 166 F.2d 492, 494, certiorari denied 334 U.S. 846, 68 S.Ct. 1514, 92 L.Ed. 1769; Gulf Refining Co. v. Home Indemnity Co., 8 Cir., 78 F.2d *214842, 843-844. And see our recent ease of Willingham v. Life & Casualty Ins. Co., 5 Cir., 216 F.2d 226.
I think the judgment below should be reversed.
. Another conceivable situation not mentioned by the court where a right to partial indemnification might arise, bringing Paragraph 8 of the “Agreements, Limitatíons and Conditions” into operation, is the contingency that the insurer might become insolvent, so that it would be unable to pay a claim in its entirety.
. Aside from the fact that the premium said by'the'court to be forfeited amounted to only $.40 each loan,1 it is to be noted that'the policy provided for adjustment of the premium' at the end of the year in accordance with actual loss experience. Thus the forfeiture does not appear to be - of a very mischievous sort, for'it would operate to benefit policyholders primarily, not the insurer. " ■ .'