dissenting:
I respectfully dissent. I believe that the majority opinion has gone astray in two ways.
First, it fails to recognize that the relevant language in the parties’ arbitration agreement is ambiguous. The arbitrator and the district judge both concluded that the language is ambiguous, and I agree with them. The arbitrator also concluded that the language is better interpreted differently than the majority has read it here, and I tend to agree with the arbitrator’s specific interpretation, as well. The two judges making up the majority insist, nonetheless, that the language can have but one reasonable interpretation, an interpretation different from that reached by the arbitrator. In light of the contrary views of the neutral arbitrator, the district court, and me, the majority’s conclusion that the language is unambiguous flies in the face of a strong headwind, and the reasoning offered by the majority to support its conclusion is unpersuasive.
Second, as a result of its refusal to recognize the ambiguity of the contractual language, the majority opinion usurps the arbitrator’s authority to interpret an ambiguous contractual term, in conflict with our decision in China National Metal Products Import/Export Co. v. Apex Digital, Inc., 379 F.3d 796 (9th Cir.2004).
1. The ambiguity of the relevant language
The problem posed by this case is how, after an arbitration between the parties has been initiated, to deal with a claim back by the respondent against the original claimant. In federal court, we call this type of claim a “counterclaim.” See Fed. R.Civ.P. 13. The parties advocate different solutions to the problem based upon different interpretations of the arbitration term in the contract: “In case of failure to settle the mentioned disputes by means of negotiations they should be settled by means of arbitration at the defendant’s side [site].”
RAE Systems maintains that the word “defendant” refers to the initial respondent, the party against which the first claim is brought, such that any subsequent counterclaim can be brought in the same arbitration proceeding, even though that *845proceeding is not located at the site of the target of the counterclaim. Polimaster, on the other hand, contends that both the initial respondent and the counter-respondent are “defendant[s]” within the meaning of the contractual term, requiring the original respondent to initiate and pursue a separate arbitral proceeding at the site of the counter-respondent in order to pursue a counterclaim.
The majority opinion accepts Polimaster’s multiple-defendant, multiple-arbitration interpretation as the right one. It further decides that this interpretation is so clearly correct that the arbitration clause is not ambiguous. The majority does not satisfactorily explain, however, why RAE’s alternative interpretation is not also a reasonable reading of the arbitration clause. “[A] contract is ambiguous if reasonable people could find its terms susceptible to more than one interpretation.” Doe 1 v. AOL LLC, 552 F.3d 1077, 1081 (9th Cir.2009) (internal quotation marks omitted). The arbitration clause is susceptible to more than one reasonable interpretation, as the arbitrator and the district court concluded. The language is ambiguous and should be recognized as such.
The majority argues that its reading is the only reasonable one based on the notion that any party who defends against “any dispute where another party seeks damages or some other form of relief against him” is necessarily a “defendant.” Maj. op. at 837. But none of the sources the majority cites, at 838-39, establish as a matter of terminology that the term “defendant” must be used only as the majority supposes.1 At least two of the cited sources actually support the opposite conclusion: that one might use the term “defendant,” especially with the definite article and without a preceding adjective, to distinguish the original defending party from the party initially on the offensive. For example, Moore’s Federal Practice, upon which the majority relies, refers to “the defendant” in precisely this way. See Daniel Coquilette et al., eds., Moore’s Federal Practice § 13.90[2] (3d ed. 2010) (especially the text following n. 9 and preceding n. 17) (using “the plaintiff’ and “the defendant” to designate substantive sides in litigation, in contrast to the term “defending party,” which refers to either side whenever it defends against an affirmative claim).2 The plaintiff, under this terminology, might later become a “counterclaim defendant” or a “crossclaim defendant” as well, but such additional designations, despite including the word “defendant,” would not transform that party into “the defendant” in the contemplated sense — a shorthand for a particular side in the litigation.
To the extent that any answer can be gleaned from the language used in the agreement, I think the language cuts slightly against the majority opinion’s in*846terpretation. Look carefully at the sentence in question: “In case of failure to settle the mentioned disputes by means of negotiations they should be settled by means of arbitration at the defendant’s side.” The word “disputes” is plural, but the words “defendant” and “side” (or “site”) are singular. The parties anticipated that there could be multiple disagreements, yet “the defendant’s site” refers to only one location. The arbitration clause does not say “defendants’ sites.” Of course, I need not be as sure of this interpretation as the majority must be of its own, because I do not contend that only this reading is reasonable. To establish that the language is ambiguous, it is enough to demonstrate that a reasonable interpretation other than the majority’s exists, and the language of the arbitration clause is more than adequate for that purpose.
The reasoning provided .by the arbitrator to support his conclusion that the existing arbitration should encompass counterclaims as well was logical. The arbitrator rejected the interpretation embraced by the majority — that arbitration of a counterclaim must be conducted in a separate proceeding at the counter-respondent’s site — because that would represent an inefficient way to resolve disputes. Parties contractually adopting arbitration as the method for resolving disputes commonly do so to achieve efficiency. It is logical to reason that such parties do not intend inefficient results. See Restatement (Second) of Contracts § 202(1) (1981) (“Words and other conduct are interpreted in the light of all the circumstances.”); id. cmt. b (“The circumstances for this purpose in-elude the entire situation, as it appeared to the parties.”).
In addition, the arbitrator recognized that it is not a novel or obscure practice to resolve all claims, including counterclaims, in a single proceeding that has already commenced. The arbitrator here specifically referenced the treatment of counterclaims in the Federal Rules of Civil Procedure, the California Rules of Civil Procedure, and the rules of the arbitration forum agreed upon by the parties, JAMS. The prosecution of counterclaims in the same proceeding is broadly recognized in international arbitration. Prominent international arbitration organizations address counterclaims explicitly in their rules. See, e.g., International Chamber of Commerce Rules of Arbitration art. 5; London Court of International Arbitration Rules art. 2.1(b); German Institute of Arbitration Rules § 10; United Nations Commission on International Trade Law Arbitration Rules art. 19.
Considering the context in which the parties made their agreement does not improperly assume any conclusion or wrongly impute any particular motivation to the parties. It merely recognizes one good reason the parties may have intended to agree to something different from the interpretation of the arbitration clause that the majority espouses: because the majority’s interpretation ignores both the common desire for efficiency and the widespread procedural practice of litigating counterclaims in the same proceeding.3 Given this context, it is not so clear to me, let alone unambiguously clear from the words of the arbitration clause, that the parties agreed to require piecemeal litiga*847tion. The majority opinion does not persuasively explain why we should conclude that they did.
The majority opinion’s arguments are, in reality, circular — the arguments for its preferred reading of the arbitration clause assume the correctness of that reading. For example, the majority opinion asserts, at 16562, that “the parties’ clause was adequate to express their intent ... to provide for separate arbitrations at the defendant’s site.” I agree that if the parties intended separate arbitrations at the sites of each defendant or counterclaim defendant, then that is how the agreement should be interpreted and applied. But the inference that the parties intended the interpretation favored by the majority rests on nothing other than the majority’s own interpretation of the contractual language. Nothing else is cited by the majority opinion to support its assertion that the parties intended that result, nor is any persuasive explanation given to counter the reasoning of the arbitrator that reached a different conclusion.
Similarly, the majority opinion asserts, at 16562, that this dissent rests on an “assumption” that counterclaims will be joined into an existing proceeding, and, at 16562, that the relevant language is clear because there was “no reason to require the parties to include contractual language specifically defeating or negating the joinder of claims.” But the parties obviously recognized the possibility of conflicting claims. The possibility of combining those claims into a single proceeding was by no means unknown. As noted above, that is the result suggested by both rules of courts and rules of international arbitration organizations. It is no less an “assumption” to conclude that, in the absence of a contractual agreement, multiple claims should be litigated piecemeal in separate arbitrations.
In the end, the reasoning offered by the majority to demonstrate that the agreement unambiguously provided for piecemeal arbitrations rests on nothing more than the majority’s own assumption that its interpretation of the arbitration clause is correct. In the face of contrary conclusions by the arbitrator, the district court, and this dissent, that is much too thin a reed to support the majority’s conclusion that the relevant language is unambiguous.
2. The China National decision
Our decision in China National Metal Products Import/Export Co. v. Apex Digital, Inc., 379 F.3d 796 (9th Cir.2004), requires that we respect the arbitrator’s interpretation of an ambiguous contractual provision. Indeed, it also provides further support for the conclusion that the arbitration clause here is ambiguous.
The arbitration clause in China National provided that all disputes arising from or in connection with the contract would be submitted to a specified forum, the China International Economic and Trade Arbitration Commission (“CIETAC”), for arbitration in Beijing, Shenzhen, or Shanghai, “at the Claimant’s option.” China National, 379 F.3d at 800. Just as the parties here argue over who qualifies as “defendant,” the parties in China National debated who qualified as “claimant” under their arbitration clause. At stake was the right to determine where arbitration would take place.
Apex first commenced arbitration against China National in Shanghai. Days later, China National brought its own claims against Apex in a separate arbitration in Beijing. Id. at 798-99. Apex argued that only it, as the party that first initiated arbitration, was a “claimant,” and that its selection of Shanghai as the arbitral forum required China National to bring its claims, which arose largely out of *848the same set of facts, as counterclaims in the ongoing Shanghai arbitration. Id. at 801. China National countered that “[i]t too was a rightful claimant with respect to its claims against Apex” and that it therefore retained the right, under the arbitration clause, “to pick a forum for its own claims.” Id. After considering its own rules, CIETAC decided in favor of China National’s position and let the claims proceed separately before separate panels. The Beijing panel entered an award in favor of China National and against Apex. China National brought an action in federal district court to confirm the Beijing panel’s award, the court confirmed the award, and Apex appealed, arguing that only one proceeding, the Shanghai arbitration, should have taken place. We rejected the challenge and affirmed the confirmation order. Id. at 797-98.
We held that “[b]oth positions are arguable, and in the face of an assertion that there can be two claimants, the text of the arbitration clause alone is indeterminate and does not resolve the matter.” Id. at 801. The parties’ disagreement in this case over who is a “defendant” — a respondent to an initial claim only, or both an initial respondent and a party who responds to counterclaims' — matches the China National debate over the meaning of the contractual term “claimant.” See id. (describing how the parties argued “claimant” should be interpreted). And the arbitration clause here is equally “indeterminate” with respect to the dispositive interpretive question of the precise meaning of “defendant.”
Because China National dealt with an analogous contractual ambiguity, it controls our decision here and requires us to affirm. China National established that an arbitrator does not impermissibly “trump specific terms of the parties’ [agreement] by turning to its own rules” when, as here, an “arbitral clause [does] not resolve the parties’ dispute itself.”4 Id. In our case the arbitrator applied JAMS rules and the Federal and California Rules of Civil Procedure because the parties’ agreement left a dispute about counterclaims unresolved. The application of extrinsic procedural rules did not contradict the parties’ agreement, but merely supplemented it. The arbitrator thus did not violate Article V, section 1(d) of the New York Convention, and the judgment of the district court confirming the arbitration award should be affirmed.
The majority opinion tries to distinguish China National on several grounds, but none are persuasive. It points out that “the parties in China National adopted CIETAC rules in their arbitration agreement,” so the gap-filling application of those rules by CIETAC “did not trump the specific terms of the parties’ agreement.” Maj. op. at 842; see China National, 379 F.3d at 801. But it was the arbitration panel, not our court, that interpreted CIETAC rules and made the decision as to where the arbitration could proceed. The majority opinion in this case overrides the arbitrator’s decision. While the parties in this case did not specify a choice of forum or a choice of procedural law to address issues their arbitration clause did not resolve, they subsequently did expressly agree on JAMS as a forum.
Moreover, while the election of default procedural rules in China National may have strengthened the argument that applying those rules did not violate the par*849ties’ agreement, a choice of law was not necessary to China National’s result. “CIETAC did not trump specific terms of the parties’ purchase orders by turning to its own rules because the arbitral clause did not resolve the parties’ dispute itself,” 379 F.3d at 801 (emphasis added), not because the CIETAC rules used to fill a gap in the arbitration clause were incorporated by reference into the contract. Rules entirely extrinsic to an agreement, such as the JAMS rules and Federal and California Rules of Civil Procedure applied in this case, do not automatically conflict with that agreement. China National makes clear that it is the presence of a gap in an arbitration clause, not the specification of default rules for filling such a gap, that makes an arbitrator’s reference to extrinsic rules appropriate. A gap — that is, a dispute between the parties that “the text of the arbitration clause alone” does not resolve — exists as much in this case as it did in China National. Reference to outside procedural rules was appropriate here, even though the rules employed were not specified in advance by the parties.
The majority opinion also attempts to differentiate this case from China National by highlighting the fact that the arbitration clause at issue here, once properly interpreted, mandates a particular forum (“the defendant’s side”), while the clause in China National gave one party a choice among three Chinese cities. Maj. op. at 842-43. That distinction is beside the point. The parties debated the meaning of “claimant” in China National because its definition determined which party got to choose among Beijing, Shenzhen, and Shanghai. The winner of the interpretive debate received the right to select a city instead of, as in this ease, a predetermined forum that the victorious party would prefer. But that detail does not disturb the parallel contractual analysis that underlies the two cases. Both cases are fundamentally about how contractual terms, “claimant” and “defendant,” respectively, should be interpreted when the contracts themselves do not resolve the parties’ disputes.
Because I conclude that the judgment on the merits should be affirmed, I would reach the issue of whether the district court exceeded the scope of our earlier limited remand to correct its failure to enter a judgment. I would affirm the district court’s belatedly entered judgment. Because the court’s oversight was its failure to enter any judgment at all, it was empowered to correct its mistake by entering a judgment that included any relief, including pre- and post-judgment interest, that could have been included in the judgment in the first instance. A court is allowed, under Federal Rule of Civil Procedure 60(a), to make corrections to effectuate what it “originally intended to do.” Robi v. Five Platters, Inc., 918 F.2d 1439, 1445 (9th Cir.1990).
I would affirm.
. See, e.g., Rainbow Mgmt. Group, Ltd. v. Atlantis Submarines Haw., L.P., 158 F.R.D. 656, 659 (D.Haw.1994) (holding that "[c]o-parties become opposing parties [emphasis added] within the meaning of Fed.R.Civ.P. 13(a) after one such party pleads an initial cross-claim against the other,” with no discussion of whether an "opposing party” must be deemed a "defendant”); In re Concept Clubs, Inc., 154 B.R. 581, 586 n. 4 (distinguishing between a counterclaim and a setoff or recoupment, with no mention of whether a counterclaim makes the party that opposes it a “defendant”); Earle M. Jorgenson Co. v. T.I. U.S., Ltd., 133 F.R.D. 472, 475 (E.D.Pa.1991) (dealing, like In re Concept Clubs, with the "opposing party” issue, without endorsing any theory of how the term "defendant” can or must be used).
. See also FDIC v. F.S.S.S., 829 F.Supp. 317, 322 n. 11 (D.Alaska 1993) (also cited by the majority) (describing a counterclaim brought by "Defendants!]” against "the plaintiff”).
. We need not (indeed, should not, in determining whether a contract is ambiguous) seek to establish whether the parties subjectively took efficiency concerns or the concept of counterclaims into account. The relevant point is that the influence of these considerations as background to the agreement makes an alternate reading of the arbitration clause reasonable.
. None of the parties’ contracts defined "defendant’s site" or "defendant’s side.” The controversial term in China National, "claimant,” was similarly undefined: the arbitration clause "[did] not define 'Claimant' but [left] it open as a variable term (i.e., either party could be a claimant).” China National, 379 F.3d at 801.