concurring:
The parties will no doubt be surprised by the majority opinion, as it decides a question they neither briefed nor argued, and that wasn’t even raised below. This is a dangerous thing for a court to do, and wholly unnecessary. The case can and should be decided on the issues presented by the parties. Because I cannot join in the court’s analysis, I concur only in the result.
I
A. The majority opens with the following question: “Which law, state or federal, governs the creation of an implied, nonex-elusive copyright license?” Maj. Op. at 824. It is an interesting question that deserves exploring in the right case. It is also a difficult question that bears some analysis. Instead, the majority cites Effects Associates, Inc. v. Cohen, 908 F.2d 555 (9th Cir.1990), and then proceeds to apply state law. Given the importance of the question, one might have hoped for some explanation of how the majority got there.
As best I can tell, the majority assumes that anything that is called a contract-including an implied contract-must be governed by state law. But not every implied contract is, in fact, a contract. Certainly, some implied contracts are governed by state law. Those contracts really are contracts; they are actual agreements between parties, albeit imperfectly articulated. The cases on which the majority relies all involve this type of contract. Maj. Op. at 825-26.
But there is another type of implied contract, one that is “created otherwise than by assent and without any words or conduct that are interpreted as promissory.” 8 Arthur Linton Corbin, Corbin on Contracts § 561, at 276-77 (1960). Such an implied contract is not a contract at all; it is a legal obligation the law imposes between certain parties where there is no actual agreement between them. Id. If the implied contract that gives rise to the nonexclusive license discussed in Effects Associates is this kind of contract, then it has nothing at all to do with contract law. Rather, it is an incident of the copyright and is therefore governed by federal law.
*833In Effects Associates, we imposed a nonexclusive license as an incident of the transfer of the copyrighted work but did not expressly say that it was a creation of federal law. I believe that this was the clear implication of our ruling, but I would leave resolution of the question to a case where the issue is properly raised below and the parties have briefed and argued it before us. In this case, I would decide the question as a matter of federal law, as the parties and district court assumed it is.
B. Purporting to apply state law, the majority infers “from the terms of the contract” that Foad and GenCom agreed to a nonexclusive license. Maj. Op. at 828. Yet there is nothing in the written contract between Foad and GenCom that remotely touches upon the plans’ copyright. The contract says nothing about whether Gen-Com had the right to modify or distribute the plans. It says nothing about the circumstances in which GenCom can use the plans to build the project. The parties seem to have negotiated, written and signed the contract without ever discussing (or perhaps being aware of) these copyright issues. The majority acknowledges as much when it concedes that neither contract was sufficient to provide for a written transfer of copyright ownership because “[tjhere is nothing in either contract between GenCom and Foad which purports to transfer any of Foad’s exclusive rights to GenCom.” Maj. Op. at 825.
Which leads to the question: Exactly which terms of the contract give rise to the license? The majority never says, relying instead on what the contract does not say. The majority notes that Foad failed to “spell[ ] out explicitly in the agreement” a requirement that “GenCom ... seek Foad’s permission before using the plans to build the project.” Maj. Op. at 829. In other words, Foad’s failure to expressly reserve its rights raises an implication that Foad licensed those rights. This makes intuitive sense but, as we are applying state law, I would have expected to see state caselaw or a state law treatise cited on this point. The majority offers no state law authority to support its conclusion that a court can infer an agreement from the absence of a contrary agreement. Id.
Even more troubling is the majority’s reliance on the penumbras emanating from an indemnification clause. The majority says that the written contract’s “indemnification clause plus the absence of any prohibition against modification by others indicates that the contract granted GenCom an implied license to hire others to create derivative works using the revised plot plan for the purpose of completing the project.” Maj. Op. at 830. The majority cites no state law authority for this point, either, and I doubt it can be found. Indemnification is not authorization. Usually, it’s the other way around: A party seeks indemnification against another party’s unauthorized actions. For example, rental car companies routinely require customers to indemnify them against harm caused by underage drivers, but this does not mean they authorize their customers to give the car keys to children. The majority’s conclusion that an indemnification clause creates a presumption that the indemnified-for act is authorized is almost certainly inconsistent with state law.
C. Part III of the opinion devotes considerable effort to figuring out whether the implied copyright license previously found to exist is nonetheless defeated by parol evidence. This makes about as much sense as calculating how high is up.
Parol evidence consists of words or conduct outside the contract that tends to vary or explain the contract’s written terms. The majority correctly recognizes that California has a particularly broad version of the parol evidence rule, which *834permits a court to examine such evidence even when the contract appears to be clear on its face; other states have a narrower version of the rule, which prohibits reference to parol evidence unless the contract terms are ambiguous. In either event, however, the parol evidence is used to shed light on words actually used in the contract.
The implied copyright license here is not a term in the contract; rather, as the name suggests, it’s a term that is implied from the relationship of the parties. In such circumstances, it makes no sense at all to talk about parol evidence, or to consider whether California applies a broad or narrow parol evidence rule. There are no words of the contact that the parol evidence here can be used to interpret or clarify.
Evidence extrinsic to the contract is, of course, not irrelevant. Because the implied license is derived from the relationship of the parties-which may well extend beyond the contract-it is entirely appropriate to look at any words or conduct that bear on whether a copyright license should be implied. But that is not a question of parol evidence; rather, it goes to whether such a license exists in the first place.
D. The majority’s preemption analysis is incomplete. The Copyright Act has a specific preemption provision, 17 U.S.C. § 301, but the majority does not cite it. Normally, of course, section 301 is the start of any preemption analysis in a copyright case, see, e.g., Kodadek v. MTV Networks, Inc., 152 F.3d 1209, 1212 (9th Cir.1998); Anderson v. Nidorf 26 F.3d 100, 102 (9th Cir.1994), so the majority’s assertion that “[i]n enacting the Copyright Act, Congress did not preempt the field,” is at least incomplete. Maj. Op. at 827. Other courts have devoted considerable discussion to section 301 in resolving the tension between state contract law and federal copyright laws. See Wrench LLC v. Taco Bell Corp., 256 F.3d 446, 453-59 (6th Cir.2001) (holding that the Copyright Act does not preempt a state breach of implied-in-fact contract claim); Selby v. New Line Cinema Corp., 96 F.Supp.2d 1053, 1057-62 (C.D.Cal.2000) (holding that the Copyright Act preempts a state breach of contract claim); Endemol Entm’t B.V. v. Twentieth Television Inc., 48 U.S.P.Q.2d 1524, 1526-28, 1998 WL 785300 (C.D.Cal.1998) (same).
This analysis is also beside the point. The majority purports to hold that application of the parol evidence rule is not preempted by the Copyright Act but, for reasons pointed out above, the parol evidence rule has no relevance to our case, so we have no occasion to speak to the issue. See Part I.C supra. However, the parol evidence rule may well be quite relevant in a case where the contract contains an express copyright license and the parol evidence is being offered to undermine or vary the terms of that license. I am not at all sure whether, in such a case, applying California’s broad version of the parol evidence rule is consistent with the federal copyright laws, which are designed to ensure certainty and predictability in the transfer of copyright. It is a question I would leave for a case where it is actually presented, rather than answering it in a case like ours, where no certainty is possible in any event because the written contract does not deal with the issue. I can only hope that future courts will not give undue deference to the majority’s answer to a question that was neither posed by the parties nor presented by the facts of the case before us.
II
For my part, I would answer the question actually presented by the parties, which presupposes that the implied license in Effects Associates is a matter of federal *835law. Because I find the majority’s analysis inapposite, I offer my own:
A. While an exclusive copyright license must be in writing, a nonexclusive license “may be granted orally, or may even be implied from conduct.” Effects Assocs., 908 F.2d at 558 (citing 3 Melville B. Nimmer & David Nimmer, Nimmer on Copyright § 10.03[A], at 10-36 (1989)). In Effects Associates, we found that a special effects studio granted a nonexclusive implied license when it handed over film footage to the producer who had commissioned it. See Effects Assocs., 908 F.2d at 558-59. The studio had created and delivered special shots knowing they were to be incorporated into a film, and were of little value for anything else. See id. We held that the transaction had to be understood as contemplating a license to use the footage for the purpose for which it was created; otherwise, the producer’s payment of a large sum of money would have been senseless. See id. at 559. Two other circuits have applied the teachings of Effects Associates to situations very similar to the one now before us, each involving the completion of building plans by someone other than the original architect. See Johnson v. Jones, 149 F.3d 494 (6th Cir.1998); I.A.E., Inc. v. Shaver, 74 F.3d 768 (7th Cir.1996).
In Shaver, an architect had created schematic design drawings for an air cargo building. See 74 F.3d at 770. The owners of the project hired another firm to finish the architectural work, and Shaver threatened to sue for copyright infringement. See id. at 771. The owners sought a declaratory judgment that they were not infringing, and won summary judgment. As the district court in that case put it, “Shaver granted the Airport a nonexclusive license to use his drawings as the basis for preparing working drawings and completing the project. Shaver’s argument that the Airport can use the drawings only as pieces of paper (wallhangings? placemats?) is untenable.” Id. at 772. The Seventh Circuit affirmed.
The architect in Johnson was hired to draw up plans for a client’s “dream house.” 149 F.3d at 497. After these plans had been submitted to the city for approval, the parties reached an impasse, and the client fired the architect. See id. at 498-99. When successor architects used the drawings to complete the project, Johnson successfully sued for copyright infringement, defeating defendants’ claim that they had an implied license to use the drawings. See id. at 499-502.
In each of these cases, the architect, having agreed to provide the basic plans for a project, later asserted that the client had no right to make the derivative works necessary to complete that project without the architect’s continued involvement. See Johnson, 149 F.3d at 500; Shaver, 74 F.3d at 778. If accepted, such a claim would allow the architect to hold the entire project hostage, forcing the owner either to continue to employ the architect or to fore-go the value of all work completed so far and start from scratch. Even assuming the client’s willingness to incur this cost, going back to the drawing board may not be an option where necessary government approvals have already been obtained and the approving authority is unwilling to reconsider the issue, as happened here.
While Shaver and Johnson reach different results, they are not necessarily inconsistent. As I read the cases, the result in each turned on its facts. In Shaver; the court found no evidence that the implied nonexclusive license normally expected in such a situation had been withheld. See 74 F.3d at 776-77. The certificates of registration on Shaver’s drawings stated that they were to be used for the airport facility. See id. at 776. The contract Shaver *836drew up demonstrated that he was hired to create preliminary drawings, but it reflected no expectation of further participation in the project. See id. Shaver had delivered his designs without any warning that their use would constitute copyright infringement, and even expressed hope that the ideas exhibited in his work would assist in successful realization of the project. See id. at 777.
In Johnson, the parties had never come to agreement on a final contract; that’s why the client fired the architect. See 149 F.3d at 499. Johnson, however, had twice presented the client with a contract which expressly provided that the copyrighted drawings “shall not be used by the owner or others on other projects, for additions to this Project or for completion of this Project by others, unless the architect is adjudged to be in default under this agreement, except by agreement in writing with appropriate compensation to the Architect.” Id. at 498. While the client never signed this contract, she asked Johnson to continue working after he had given it to her, and she led him to believe that she eventually would sign it. See id. at 498. The client thus had fair notice that the architect intended to retain control over his designs and any derivative works necessary to complete the project. Under these circumstances, the district court had sufficient evidence from which to find that Johnson had never granted an implied license for the client to use the drawings without retaining him on the project.1
B. Looking at the relationship between Foad and GenCom, I conclude that this case is cut from the cloth of Shaver, not Johnson. Like Shaver, Foad drafted letter contracts that define a specific scope of work and mention no expectation of a continued role in the project. On each occasion, Foad agreed to generate a particular set of documents in exchange for a lump sum payment. In August 1995, Foad contracted to prepare “a preliminary site grading plan,” for which it was paid $11,500. The February 1996 contract is more elaborate, and states that the work is to consist of “the preparation of final engineering drawings (Grading, Improvements, Final Map, Dimension Plot Plan, Site Plan) together with necessary field survey for the proposed ... project.” In exchange for all this, GenCom paid Foad a lump sum of $175,000. The overall project obviously was not finished, but for all I can tell from this contract, the relationship between GenCom and Foad was.
Nowhere does the contract contain a clause like those Johnson proposed, forbidding any additions to the project or completion of it by others. Foad asserts that the contract does say this, and points to the following paragraphs:
12. All original drawings, plans, documents, other papers and copies thereof prepared in connection with this agreement will remain the property of FOAD CONSULTING GROUP, INC. and may be used *837without the consent of the client and/owner(s). [T]he aforementioned papers will be kept on file by FOAD CONSULTING GROUP, INC. and copies will be provided to the client and/owner(s) at client and/owner(s)’s request, and at client and/owner(s)’s expense.
13. All documents, including drawings and specifications prepared by FOAD CONSULTING GROUP, INC., are not intended or represented to be suitable for reuse by client and/owner(s) or others without adaptation by FOAD CONSULTING GROUP, INC. [Client and/owner(s) shall indemnify and hold harmless FOAD CONSULTING GROUP, INC. from any and all claims, damages, losses and expenses including attorney’s fees arising out of or resulting from any use other than for the project which is the subject of this contract. Any such additions or adaptations will be billed as extra work at the current rate schedule.
Neither of these provisions speak to copyright. Paragraph 12 is clearly about the ownership of certain original papers, and is silent about what GenCom may or may not do with copies. Paragraph 13 is a disclaimer and indemnification clause. As I have discussed above, this has no bearing on the copyright issues in this case. See pp. 833-34 supra.
Foad also points to the plan itself, which bears the following legend:
All ideas, designs, arrangements and plans indicated or represented by this drawing are owned by, and the property of Foad Consulting Group, Inc. and were created, evolved, and developed for use on, and in connection with the specified project. None of such ideas, designs, arrangements or plans shall be used without written permission of Foad Consulting Group, Inc.
The first problem with this, of course, is that copyright law does not authorize the ownership of ideas. See Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539, 547, 105 S.Ct. 2218, 85 L.Ed.2d 588 (1985). In any case, this legend is reminiscent of Shaver’s registration certificates, which said that the copyrighted designs were to be used for the airport facility in question. See Shaver, 74 F.3d at 776. It is difficult to see how an announcement that the plans were “created, evolved, and developed for use on, and in connection with the specified project” can be read to prohibit their use on and in connection with that very project. This sentence would serve no purpose if the one following it were intended to require written permission even for the very uses Foad intended.
Finally, Foad submitted a declaration from an architectural expert who asserted that, under custom and practice in the industry, a plan “may not be used to produce a similar plan, without the permission of the original designer.”2 This may be true, but begs the question whether “the totality of the parties’ conduct indicates an intent to grant such permission.” 3 Melville B. Nimmer & David Nimmer, Nimmer on Copyright § 10.03[A][7], at 10-42 (2000). Foad created its revised plan at GenCom’s request, knowing that this document was being used to obtain city approvals which would then be binding on the project. Foad drafted a contract which makes no mention of its copyright. Foad *838has presented nothing that casts doubt on the conclusion that the relationship between the parties gave rise to an implicit nonexclusive license in GenCom’s favor.
Finding there was an implied license does not end the inquiry, however. We must also ask what the scope of this license was, and whether GenCom exceeded it. See Oddo v. Ries, 743 F.2d 630, 634 (9th Cir.1984) (implied license to use articles in a manuscript did not include license to use them in a book). Here, Foad’s own legend announces that the plans were created “for use on, and in connection with the specified project.” The project was to build a shopping center on the site depicted in Foad’s plan, and the derivative works created by defendants were used solely in furtherance of that purpose.
It is true that the details of this project have evolved since Foad’s contribution to it. Any project, however, will undergo modifications along the way-whether because of site conditions that are discovered during construction, because the city planners make unanticipated demands, or because the economics of the situation change. To regard each such modification as resulting in a new project for purposes of the license would be to render the implied license useless, once again giving the plan designer, rather than the owner, total control over the development.
In making those derivative works necessary to complete the shopping center, defendants did not exceed the scope of their license to use Foad’s plan in connection with the project for which it was created.
. I disagree, however, with another reason the Sixth Circuit gave to support its holding on this point: "It is one thing for an architect to allow the owner to hire someone else to build a house from the architect's plans. It is quite another for the architect to allow a competitor to come in and claim the architect’s work as his own.” Johnson, 149 F.3d at 501. Foad too has complained about the fact that bragging rights for the completed project will accrue to the defendants, even though the plan was based largely on Foad's work. While misrepresentation as to the source of a design may be actionable under the Lanham Act, see id. at 502-03, or under state law, the author’s interest in proper attribution is foreign to the Copyright Act, which "seeks to vindicate the economic, rather than the personal, rights of authors.” Gilliam v. Am. Broad. Cos., 538 F.2d 14, 24 (2d Cir.1976).
. In Shaver too, the architect had offered extrinsic evidence of industry practice and custom, which the district court rejected as being unnecessary given that the contract was not ambiguous. See 74 F.3d at 772.