dissenting.
The court today concludes that a disputed issue of material fact exists as to what the. parties intended when the Matanuska-Susit-na Borough presented and Joseph Philbin signed the release in question. Because I believe that a reasonable person in Philbin’s position, viewing the evidence in the light most favorable to Philbin, must have understood the release language to mean that Phil-bin “release[d] and discharge^] the MaV-Su Borough from any and all further claim, debt, charge, demand or liability whatsoever under or arising from” the contract he had entered, I would affirm the decision of the superior court.
Our cases make clear, and the court today reiterates, that the legal standard is “what a reasonable person would have understood the release language to have meant.”1 I agree that this is the test. But none of the six reasons posited by the court to support its conclusion that there are material facts in dispute on this issue withstands scrutiny. For that reason, I dissent.
The court first relies on two paragraphs in Philbin’s affidavit to support the proposition that “Philbin ... justifiably believed that the borough would purchase the remaining mate*1271rial.”2 First, we have made clear that after-the-fact subjective assertions of intent in contract situations are entitled to no evidentiary weight:
Differences of opinion among the parties as to their subjective intent, expressed during the litigation, do not establish an issue of fact regarding the parties’ reasonable expectations at the time they entered into the contract, since such self-serving statements are not considered to be probative.3
Moreover, we have held that where the language of an agreement in a commercial context is clear on its face, and the releasor realizes he or she is signing a release, the conclusion that the release is enforceable may be compelled as a matter of law.4 In Ahwinona v. State5 we specifically rejected an argument based on the releasor’s claim that he had not understood the release to preclude a later lawsuit: “Absent any showing of coercion or fraud, Ahwinona’s mistaken understanding of the release is not sufficient to set it aside.”6 And in Mitchell v. Mitchell we said that the releasor’s mistake was “legally irrelevant because the [settlement] clearly and unambiguously dismisses the entire lawsuit.”7 Finally, whether the borough and Philbin might reestablish a commercial relationship the following spring is not the issue. Even if Philbin believed a continuing relationship was possible, he could not have reasonably believed that such a relationship, if it did come about, would have been compelled by the contract which the borough clearly terminated several days earlier.
The court’s second reason to reverse is Kaueic’s deposition testimony that a winter shutdown was necessary as of November 2 or 3, and that he discussed this with Philbin, permitting an inference that Philbin might be allowed to complete the project in the spring.8 There is no question that the borough and Philbin discussed that possibility before November 6. But there is likewise no question that the borough abandoned that approach when it unambiguously terminated the contract for nonperformance on November 6. The mere discussion of options cannot suffice to create a dispute of material fact when those discussions are followed by termination of the contract and an unambiguous release.9
The court’s third reason- — “there is also some evidence that the borough itself did not consider the release, when Philbin signed it, to foreclose a breach of contract claim”10— fails both legally and factually. Legally, we look to the language of the release and the parties’ reasonable expectations concerning that language. The release’s language is broad and all-encompassing: the contractor “releases and discharges the Matanuska-Susitna Borough, its officers, agents and employees of and from any and all further claim, debt, charge, demand, liability or other obligation whatsoever under or arising from said contract, whether known or unknown and whether or not ascertainable at the time of the execution of this instrument except specified claims .... ” Factually, it should be of little consequence that a borough employee referred to the release as a “lien release.” *1272Indeed, the only reasonable interpretation of the document is that it is a general release with a possible exception for lien and related claims.
The court’s fourth rationale is that “Phil-bin’s theory that the release covered only that part of the contract relating to Beverly Lakes Road is also supported by the language of the payment request submitted by the public works department to the borough receiving department. That request sought ‘partial’ payment under the contract and indicated that the contract performance was ‘incqmplete.’ ”11 But this is irrelevant for several reasons. First, the request was an internal document, not communicated to Philbin. Second, the request was submitted on November 3, when the Borough had not yet announced its decision to terminate the contract; that announcement came three days later. As the Engineering Manager for the borough testified: “When this purchase order request for payment was received by borough purchasing personnel, they changed the ‘P’ [partial] to ‘F’ [full] on both lines, as it was determined, by that date that this would definitely be a closed contract.” (Emphasis added.) Finally, the release is clear on its face. How can an internal document, unseen by Philbin and subsequently properly modified to reflect the decision to terminate the contract, be used to justify a conclusion that Philbin’s intent in signing the release is in dispute?
The court’s fifth justification relies on the observation that “the Borough could not have insisted that Philbin execute a complete release as a condition to payment for the initial gravel deliveries because that would have been coercive and in bad faith.”12 I have no quarrel with this statement. But the borough did not do that. The release specifically allowed the releasor to except whatever claims he wished to except. Counsel for the borough properly admitted at oral argument that this would have allowed Philbin to except a breach of contract claim. The majority somehow turns this into the unwarranted conclusion that “[the borough] did not intend the release to encompass a breach of contract claim.”13 Philbin did not except any breach of contract claim, and the language of the release is otherwise extremely broad. Under these circumstances, the only reasonable interpretation of the release is that “any and all” claims, demands, or other obligations arising from the contract were released, unless specifically excepted.
The court last points to the borough’s alleged failure to mention the release in subsequent negotiations when Philbin tried to resurrect the terminated contract the following spring. Using a party’s negotiating technique in these circumstances as evidence of the party’s intent in signing a document months earlier seems questionable. For one thing, the borough’s reliance on the release in negotiations with Philbin would probably have been inconsistent with reaching a negotiated settlement of the dispute. For another, positions taken in settlement negotiations have never been relevant later when negotiations have proved unsuccessful.14 I would give no weight to the borough’s alleged failure to mention the release several months later in settlement negotiations when assessing the meaning of the release.
The court fails to set out15 the compelling evidence that Philbin knew, when he signed the release, that the contract had been terminated and that he was releasing any and all claims arising under the contract (except *1273those that he specifically excepted). While he now argues that the contract remained open and he therefore had no reason to understand that the release he signed actually was a release of all claims “arising under the contract,” there is no doubt that he knew the contract was terminated. His own words prove it:
[F]our or five days [after the meeting with Mr. Kaucic] ... I got another letter ... stating that they’ve decided to terminate the contract, I was irate. I’ve never had that happen. You don’t tell somebody you’re going to make a winter shutdown, especially before the duration of the contract, and then let that time pass and then hand them a termination notice ... I arranged a meeting with the Borough and I told them that ivithin a two-week period come spring once the ground was thawed, I could absolutely guarantee that I could crush the material and well before the road restrictions lifted-I would like to go ahead and crush the material. And they said no, they just didn’t believe I could do it. My record didn’t look good because of this project.
(Emphasis added.) This testimony was given in a related case pitting Philbin against an equipment supplier on the contract in this case. In the same case, Philbin testified that the proposed winter shut-down option had been discarded by the borough, a decision with which he heartily disagreed. Specifically, he testified that he was “shocked” when he received the termination letter instead of a shut down order:
In the context of the conversation [on November 2 or 3, Chuck Kaucic] informed me that the road surfaces were too hard to blade and prepare to accept gravel and then if we put gravel down on it, it wouldn’t bond, it would come off. And at that point I was glad to hear that because I wasn’t getting any production out of the plant at all. The material was frozen, lumped up. It wouldn’t process right. And it came as a shock after he’d — he’d orally told me that [ — ] to get this letter saying they terminated it. That it no longer was considering picking back up in the spring to let me finish it.
(Emphasis added.)
Given Philbin’s own testimony, no reasonable fact-finder could conclude that, as of the time he signed the release, he reasonably believed that the road contract was still in effect, that he would resume work on it in the spring, and that the release only concerned the work already performed. A reasonable person in Philbin’s position must have understood that he was releasing the borough from the claims he now seeks to pursue. Moreover, the law is clear that, absent coercion or fraud, even a mistaken understanding of the contents of a release is not sufficient to set it aside. Under these circumstances, and given the breadth of the release language, I believe that the superior court was correct when it found no material issues of fact in dispute. I would affirm that decision.
. Op. at 1266.
. Id.
. Peterson v. Wirum, 625 P.2d 866, 870 (Alaska 1981).
. Ahwinona v. State, 922 P.2d 884, 886 (Alaska 1996).
. Id.
. Id. at 887.
. 655 P.2d 748, 752-53 (Alaska 1982).
. Op. at 1267.
. In this regard, the court's statement about the discussion is puzzling: "Although Kaucic characterized this as a 'hypothetical' possibility, his testimony that freezing weather prevented continuation removes that factual contingency.” Op. at 1267. That freezing weather as of early November forced a shutdown did not preclude the borough from terminating the contract, because Philbin was significantly behind the contract’s production schedule well before that date. The borough was free to offer a continuation of the contract, but it was not required to do so.
.Op. at 1267-1268.
. Id. at 1268.
. Op. at 1268.
. Id. at 1269.
. See Alaska Rule of Evidence 408.
. The court satisfies itself with the observation that this evidence "support[s] the borough’s interpretation ... [b]ut ... is not conclusive given the averments in Philbin’s affidavit” and the inferences which must be drawn in favor of the non-movant. Op. at 1268. This approach ignores our well-established rule that subjective statements made during litigation are entitled to no weight "since such self-serving statements are not considered to be probative.” Peterson v. Wirum, 625 P.2d at 870.