Mr. Justice Clark

delivered the opinion of the Court.

After a lapse of 15 years, the Court again focuses its attention on the patentability of inventions under the standard of Art. I, § 8, cl. 8, of the Constitution and under the conditions prescribed by the laws of the United States. Since our last expression on patent validity, A. & P. Tea Co. v. Supermarket Corp., 340 U. S. 147 (1950), the Congress has for the first time expressly added a third statutory dimension to the two requirements of novelty and utility that had been the sole statutory test since the Patent Act of 1793. This is the test of obviousness, i. e., whether “the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negatived by the manner in which the invention was made.” § 103 of the Patent Act of 1952, 35 U. S. C. § 103 (1964 ed.).

The questions, involved in each of the companion cases before us, are what effect the 1952 Act had upon traditional statutory and judicial tests of patenta-bility and what definitive tests are now required. We have concluded that the 1952 Act was intended to codify judicial precedents embracing the principle long ago *4announced by this Court in Hotchkiss v. Greenwood, 11 How. 248 (1851), and that, while the clear language of § 103 places emphasis on an inquiry into obviousness, the general level of innovation necessary to sustain patent-ability remains the same.

The Cases.

(a). No. 11, Graham v. John Deere Co., an infringement suit by petitioners, presents a conflict between two Circuits over the validity of a single patent on a “Clamp for vibrating Shank Plows.” The invention, a combination of old mechanical elements, involves a device designed to absorb shock from plow shanks as they plow through rocky soil and thus to prevent damage to the plow. In 1955, the Fifth Circuit had held the patent valid under its rule that when a combination produces an “old result in a cheaper and otherwise more advantageous way,” it is patentable. Jeoffroy Mfg., Inc. v. Graham, 219 F. 2d 511, cert. denied, 350 U. S. 826. In 1964, the Eighth Circuit held, in the case at bar, that there was no new result in the patented combination and that the patent was, therefore, not valid. 333 F. 2d 529, reversing 216 F. Supp. 272. We granted certiorari, 379 U. S. 956. Although we have determined that neither Circuit applied the correct test, we conclude that the patent is invalid under § 103 and, therefore, we affirm the judgment of the Eighth Circuit.

(b). No. 37, Calmar, Inc. v. Cook Chemical Co., and No. 43, Colgate-Palmolive, Co., v. Cook Chemical Co., both from the Eighth Circuit, were separate declaratory judgment actions, but were filed contemporaneously. Petitioner in Calmar is the manufacturer of a finger-operated sprayer with a “hold-down” cap of the type commonly seen on grocers’ shelves inserted in bottles of insecticides and other liquids prior to shipment. Petitioner in Colgate-Palmolive is a purchaser of the sprayers *5and uses them in the distribution of its products. Each action sought a declaration of invalidity and noninfringement of a patent on similar sprayers issued to. Cook Chemical as assignee of Baxter I. Scoggin, Jr., the inventor. By cross-action, Cook Chemical claimed infringement. The actions were consolidated for trial and the patent was sustained by the District Court. 220 F. Supp. 414. The Court of Appeals affirmed, 336 F.'2d 110, and we granted certiorari, 380 U. S. 949. We reverse.

Manifestly, the validity of each of these patents turns on the facts. The basic problems, however, are the same in each case and require initially a discussion of the constitutional and statutory provisions covering the patentability of the inventions.

II.

At the outset it must be remembered that the federal patent power stems from a specific constitutional provision which authorizes the Congress “To promote the Progress of . . . useful Arts, by securing for limited Times to . . . Inventors the exclusive Right to their . . . Discoveries.” Art. I, § 8, cl. 8.1 The clause is both a grant of power and a limitation. This qualified authority, unlike the power often exercised in the sixteenth and seventeenth centuries by the English Crown, is limited to the promotion of advances in the “useful arts.” It was written against the backdrop of the practices — even-, tually curtailed by the Statute of Monopolies — of the Crown in granting monopolies to court favorites in goods or businesses which had long before been enjoyed by the public. See Meinhardt, Inventions, Patents and Monopoly, pp. 30-35 (London, 1946). The Congress in the *6exercise of the patent power may not overreach the restraints imposed by the stated constitutional purpose. Nor may it enlarge the patent monopoly without regard to the innovation, advancement or social benefit gained thereby. Moreover, Congress may not authorize the issuance of patents whose effects are to remove existent knowledge from the public domain, or to restrict free access to materials already available. Innovation, advancement, and things which add to the sum of useful knowledge are inherent requisites in a patent system which by constitutional command must “promote the Progress of . . . useful Arts.” This is the standard, expressed in the Constitution and it may not be ignored. And it is in this light that patent validity “requires reference to a standard written into the Constitution.” A. & P. Tea Co. v. Supermarket Corp., supra, at 154 (concurring opinion).

Within the limits of the constitutional grant, the Congress may, of course, implement the stated purpose of the Framers by selecting the policy which in its judgment best effectuates the constitutional aim. This is but a corollary to the grant to Congress of any Article I power. Gibbons v. Ogden, 9 Wheat. 1. Within the scope established by the Constitution, Congress may set out conditions and tests for patentability. McClurg v. Kingsland, 1 How. 202, 206. It is the duty of the Commissioner of Patents and of the courts in the administration of the patent system to give effect to the constitutional standard by appropriate application, in each case, of the statutory scheme of the Congress.

Congress quickly responded to the bidding of the Constitution by enacting the Patent Act of 1790 during the second session of the First Congress. It created an agency in the Department of State headed by the Secretary of State, the Secretary of the Department of War *7and the Attorney General, any two of whom could issue a patent for a period not exceeding 14 years to any petitioner that “hath . . . invented or discovered any useful art, manufacture, ... or device, or any improvement therein not before known or used” if the board found that “the invention or discovery [was] sufficiently useful and important . . . .” 1 Stat. 110. This group, whose members administered the patent system along with their other public duties, was known by its own designation as “Commissioners for the Promotion of Useful Arts.”

Thomas Jefferson, who as Secretary of State was a member of the group, was its moving spirit and might well be called the “first administrator of our patent system.” See Federico, Operation of the Patent Act of 1790, 18 J. Pat. Off. Soc. 237, 238 (1936). He was not only an administrator of the patent system under the 1790 Act, but was also the author of the 1793 Patent Act. In addition, Jefferson was himself an inventor of great note. His unpatented improvements on plows, to mention but one line of his inventions, won acclaim and recognition on both sides of the Atlantic. Because of his active interest and influence in the early development of the patent system, Jefferson’s views on the general nature of the limited patent monopoly under the Constitution, as well as his conclusions as to conditions for patentability under the statutory scheme, are worthy of note.

Jefferson, like other Americans, had an instinctive aversion to monopolies. It was a monopoly on tea that sparked the Revolution and Jefferson certainly did not favor an equivalent form of monopoly under the new government. His abhorrence of monopoly extended initially to patents as well. From France, he wrote to Madison (July 1788) urging a Bill of Rights provision restricting monopoly, and as against the argument that *8limited monopoly might serve to incite “ingenuity,” he argued forcefully that “the benefit even of limited monopolies is too doubtful to be opposed to that of their general suppression,” Y Writings of Thomas Jefferson, at 47 (Ford ed., 1895).

His views ripened, however, and in another letter to Madison (Aug. 1789) after the drafting of the Bill of Rights, Jefferson stated that he would have been pleased by an express provision in this form:

“Art. 9. Monopolies may be allowed to persons for their own productions in literature & their own inventions in the arts, for a term not exceeding — years but for no longer term & no other purpose.” Id,., at 113.

And he later wrote:

“Certainly an inventor ought to be allowed a right to the benefit of his invention for some certain time. . . . Nobody wishes more than I do that ingenuity should receive a liberal encouragement.” Letter to Oliver Evans (May 1807), V Writings of Thomas Jefferson, at 75-76 (Washington ed.).

Jefferson’s philosophy on the nature and purpose of the patent monopoly is expressed in a letter to Isaac McPherson (Aug. 1813), a portion of which we set out in the margin.2 He rejected a natural-rights theory in *9intellectual property rights and clearly recognized the social and economic rationale of the patent system. The patent monopoly was not designed to secure to the inventor his natural right in his discoveries. Rather, it was a reward, an inducement, to bring forth new knowledge. The grant of an exclusive right to an invention was the creation of society — at odds with the inherent free nature of disclosed ideas — and was not to be freely given. Only inventions and discoveries which furthered human knowledge, and were new and useful, justified the special inducement of a limited private monopoly. Jefferson did not believe in granting patents for small details, obvious improvements, or frivolous devices. His writings evidence his insistence upon a high level of patentability.

As a member of the patent board for several years, Jefferson saw clearly the difficulty in “drawing a line between the things which are worth to the public the embarrassment of an exclusive patent, and those which are not.” The board on which he served sought to draw such a line and formulated several rules which are preserved in *10Jefferson's correspondence.3 Despite the board's efforts, Jefferson saw “with what slow progress a system of general rules could be matured.” Because of the “abundance” of cases and the fact that the investigations occupied “more time of the members of the board than they could spare from higher duties, the whole was turned over to the judiciary, to be matured into a system, under which every one might know when his actions were safe and lawful.” Letter to McPherson, supra, at 181, 182. Apparently Congress agreed with Jefferson and the board that the courts should develop additional conditions for patentability. Although the Patent Act was amended, revised or codified some 50 times between 1790 and 1950, Congress steered clear of a statutory set of requirements other than the bare novelty and utility tests reformulated in Jefferson’s draft of the 1793 Patent Act.

III.

The difficulty of formulating conditions for patent-ability was heightened by the generality of the constitutional grant and the statutes implementing it, together with the underlying policy of the patent system that “the things which are worth to the public the embarrassment *11of an exclusive patent,” as Jefferson put it, must outweigh the restrictive effect of the limited patent monopoly. The inherent problem was to develop some means of weeding out those inventions which would not be disclosed or devised but for the inducement of a patent.

This Court formulated a general condition of patent-ability in 1851 in Hotchkiss v. Greenwood, 11 How. 248. The patent involved a mere substitution of materials— porcelain or clay for wood or metal in doorknobs — and the Court condemned it, holding:4

“[U]nless more ingenuity and skill . . . were required . . . than were possessed by an ordinary mechanic acquainted with the business, there was an absence of that degree of skill and ingenuity which constitute essential elements of every invention. In other words, the improvement is the work of the skilful mechanic, not that of the inventor.” At p. 267.

Hotchkiss, by positing the condition that a patentable invention evidence more ingenuity and skill than that possessed by an ordinary mechanic acquainted with the business, merely distinguished between new and useful innovations that were capable of sustaining a patent and those that were not. The Hotchkiss test laid the cornerstone of the judicial evolution suggested by Jefferson and left to the courts by Congress. The language in the case, and in those which followed, gave birth to “invention” as a word of legal art signifying patentable inventions. Yet, as this Court has observed, “[t]he truth is the word [‘invention’] cannot be defined in such manner as to afford any substantial aid in determining whether a particular device involves an exercise of the inventive faculty *12or not.” McClain v. Ortmayer, 141 U. S. 419, 427 (1891); A. & P. Tea Co. v. Supermarket Corp., supra, at 151. Its use as a label brought about a large variety of opinions as to its meaning both in the Patent Office, in the courts, and at the bar. The Hotchkiss formulation, however, lies not in any label, but in its functional approach to questions of patentability. In practice, Hotchkiss has required a comparison between the subject matter of the patent, or patent application, and the background skill of the calling. It has been from this comparison that patentability was in each case determined.

The 1952 Patent Act.

The Act sets out the conditions of patentability in three sections. An analysis of the structure of these three sections indicates that patentability is dependent upon three explicit conditions: novelty and utility as articulated and defined in § 101 and § 1Q2, and non-obviousness, the new statutory formulation, as set out in § 103. The first two sections, which trace closely the 1874 codification, express the “new and useful” tests which have always existed in the statutory scheme and, for our purposes here, need no clarification.5 The pivotal *13section around which the present controversy centers is § 103. It provides:

Ҥ 103. Conditions for patentability; non-obvious subject matter

“A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negatived by the manner in which the invention was made.”

*14The section is cast in relatively unambiguous terms. Patentability is to depend, in addition to novelty and utility, upon the “non-obvious” nature of the “subject matter sought to be patented” to a person having ordinary skill in the pertinent art.

The first sentence of this section is strongly reminiscent of the language in Hotchkiss. Both formulations place emphasis on the pertinent art existing at the time the invention was made and both are implicitly tied to advances in that art. The major distinction is that Congress has emphasized “nonobviousness” as the operative- test of the section, rather than the less definite “invention” language of Hotchkiss that Congress thought had led to “a large variety” of expressions in decisions and writings. In the title itself the Congress used the phrase “Conditions for patentability; non-obvious subject matter” (italics added), thus focusing upon “non-obviousness” rather than “invention.” 6 The Senate and House Reports, S. Rep. No. 1979, 82d Cong., 2d Sess. (1952); H. R. Rep. No. 1923, 82d Cong., 2d Sess. (1952), reflect this emphasis in these terms:

“Section 103, for the first time in our statute, provides a condition which exists in the law and has existed for more than 100 years, but only by reason of decisions of the courts. An invention which has been made, and which is new in the sense that the same thing has not been made before, may still not be patentable if the difference between the new thing and what was known before is not considered sufficiently great to warrant a patent. That has been expressed in a large variety of ways in decisions of *15the courts and in writings. Section 103 states this requirement in the title. It refers to the difference between the subject matter sought to be patented and the prior art, meaning what was known before as described in section 102. If this difference is such that the subject matter as a whole would have been obvious at the time to a person skilled in the art, then the subject matter cannot be patented.

“That provision paraphrases language which has often been used in decisions of the courts, and the section is added to the statute for uniformity and definiteness. This section should have a stabilizing effect and minimize great departures which have appeared in some cases.” H. R. Rep., supra, at 7; S. Rep., supra, at 6.

It is undisputed that this section was, for the first time, a statutory expression of an additional requirement for patentability, originally expressed in Hotchkiss. It also seems apparent that Congress intended by the last sentence of § 103 to abolish the test it believed this Court announced in the controversial phrase “flash of creative genius,” used in Cuno Corp. v. Automatic Devices Corp., 314 U. S. 84 (1941).7

*16It is contended, however, by some of the parties and by several of the amici that the first sentence of § 103 was intended to sweep away judicial precedents and to lower the level of patentability. Others contend that the Congress intended to codify the essential purpose reflected in existing judicial precedents — the rejection of insignificant variations and innovations of a commonplace sort — and also to focus inquiries under § 103 upon nonobviousness, rather than upon “invention,” as a means of achieving more stability and predictability in determining patentability and validity.

The Reviser’s Note to this section,8 with apparent reference to Hotchkiss, recognizes that judicial requirements as to “lack of patentable novelty [have] been followed since at least as early as 1850.” The note indicates that the section was inserted because it “may have some stabilizing effect, and also to serve as a basis for the addition at a later time of some criteria which may be worked out.” To this same effect are the reports of both Houses, supra, which state that the first sentence *17of the section “paraphrases language which has often been used in decisions of the courts, and the section is added to the statute for uniformity and definiteness.”

We believe that this legislative history, as well as other sources,9 shows that the revision was not intended by Congress to change the general level of patentable invention. We conclude that the section was intended merely as a codification of judicial precedents embracing the Hotchkiss condition, with congressional directions that inquiries into the obviousness of the subject matter sought to be patented are a prerequisite to patentability.

V.

Approached in this light, the § 103 additional condition, when followed realistically, will permit a more practical test of patentability. The emphasis on non-obviousness is one of inquiry, not quality, and, as such, comports with the constitutional strictures.

While the ultimate question of patent validity is one of law, A. & P. Tea Co. v. Supermarket Corp., supra, at 155, the § 103 condition, which is but one of three conditions, each of which must be satisfied, lends itself to several basic factual inquiries. Under § 103, the scope and content of the prior art are to be determined; differences between the prior art and the claims at issue are to be ascertained; and the level of ordinary shill in the pertinent art resolved. Against this background, the obviousness or nonobviousness of the subject matter is determined. Such secondary considerations as commercial success, long felt but unsolved needs, failure of others, etc., might be utilized to give light to the circum*18stances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness or nonobviousness, these inquiries may have relevancy. See Note, Sub tests of “Nonobviousness”: A Nontechnical Approach to Patent Validity, 1.12 U. Pa. L. Rev. 1169 (1964).

This is not to say, however, that there will not be difficulties in applying the nonobviousness test. What is obvious is not a question upon which there is likely to be uniformity of thought in every given factual context. The difficulties, however, are comparable to those encountered daily by the courts in such frames of reference as negligence and scienter, and should be amenable to a case-by-case development. We believe that strict observance of the requirements laid down here will result in that uniformity and definiteness which Congress called for in the 1952 Act.

While we have focused attention on the appropriate standard to be applied by the courts, it must be remembered that the primary responsibility for sifting out unpatentable material lies in the Patent Office. To await litigation is — for all practical purposes — to debilitate the patent system. We have observed a notorious difference between the standards applied by the Patent Office and by the courts. While many reasons can be adduced to explain the discrepancy, one may well be the free rein often exercised by Examiners in their use of the concept of “invention.” In this connection we note that the Patent Office is confronted with a most difficult task. Almost 100,000 applications for patents are filed each year. Of these, about 50,000 are granted and the backlog now runs well over 200,000. 1965 Annual Report of the Commissioner of Patents 13-14. This is itself a compelling reason for the Commissioner to strictly adhere to the 1952 Act as interpreted here. This would, we believe, not only expedite disposition but *19bring about a closer concurrence between administrative and judicial precedent.10

Although we conclude here that the inquiry which the Patent Office and the courts must make as to patent-ability must be beamed with greater intensity on the requirements of § 103, it bears repeating that we find no change in the general strictness with which the overall test is to be applied. We have been urged to find in § 103 a relaxed standard, supposedly a congressional reaction to the “increased standard” applied by this Court in its decisions over the last 20 or 30 years. The standard has remained invariable in this Court. Technology, however, has advanced — and with remarkable rapidity in the last 50 years. Moreover, the ambit of applicable art in given fields of science has widened by disciplines unheard of a half century ago. It is but an evenhanded application to require that those persons granted the benefit of a patent monopoly be charged with an awareness of these changed conditions. The same is true of the less technical, but still useful arts. He who seeks to build a better mousetrap today has a long path to tread before reaching the Patent Office.

VI.

We now turn to the application of the conditions found necessary for patentability to the cases involved here:

A. The Patent in Issue in No. 11, Graham v. John Deere Co.

This patent, No. 2,627,798 (hereinafter called the-’798 patent) relates to a spring clamp which permits plow shanks to be pushed upward when they hit obstructions *20in the soil, and then springs the shanks back into normal position when the obstruction is passed over. The device, which we show diagrammatically in the accompanying sketches (Appendix, Fig. 1), is fixed to the plow frame as a unit. The mechanism around which the controversy centers is basically a hinge. The top half of it, known as the upper plate (marked 1 in the sketches), is a heavy metal piece clamped to the plow frame (2) and is stationary relative to the plow frame. The lower half of the hinge, known as the hinge plate (3), is connected to the rear of the upper plate by a hinge pin (4) and rotates downward with respect to it. The shank (5), which is bolted to the forward end of the hinge plate (at 6), runs beneath the plate and parallel to it for about nine inches, passes through a stirrup (7), and then continues backward for several feet curving down toward the ground. The chisel (8), which does the actual plowing, is attached to the rear end of the shank. As the plow frame is pulled forward, the chisel rips through the soil, thereby plowing it. In the normal position, the hinge plate and the shank are kept tight against the upper plate by a spring (9), which is atop the upper plate. A rod (10) runs through the center of the spring, extending down through holes in both plates and the shank. Its upper end is bolted to the top of the spring while its lower end is hooked against the underside of the shank.

When the chisel hits a rock or other obstruction in the soil, the obstruction forces the chisel and the rear portion of the shank to move upward. The shank is pivoted (at 11) against the rear of the hinge plate and pries open the hinge against the closing tendency of the spring. (See sketch labeled “Open Position,” Appendix, Fig. 1.) This closing tendency is caused by the fact that, as the hinge is opened, the connecting rod is pulled downward and the spring is compressed. When the obstruc*21tion is passed over, the upward force on the chisel disappears and the spring pulls the shank and hinge plate back into their original position. The lower, rear portion of the hinge plate is constructed in the form of a stirrup (7) which brackets the shank, passing around and beneath it. The shank fits loosely into the stirrup (permitting a slight up and down play). The stirrup is designed to prevent the shank from recoiling away from the hinge plate, and thus prevents excessive strain on the shank near its bolted connection. The stirrup also girds the shank, preventing it from fishtailing from side to side.

In practical use, a number of spring-hinge-shank combinations are clamped to a plow frame, forming a set of ground-working chisels capable of withstanding the shock of rocks and other obstructions in the soil without breaking the shanks.

Background of the Patent.

Chisel plows, as they are called, were developed for plowing in areas where the ground is relatively free from rocks or stones. Originally, the shanks were rigidly attached to the plow frames. When such plows were used in the rocky, glacial soils of some of the Northern States, they were found to have serious defects. As the chisels hit buried rocks, a vibratory motion was set up and tremendous forces were transmitted to the shank near its connection to the frame. The shanks would break. Graham, one of the petitioners, sought to meet that problem, and in 1950 obtained a patent, U. S. No. 2,493,811 (hereinafter '811), on a spring clamp which solved some of the difficulties. Graham and his companies manufactured and sold the ’811 clamps. In 1950, Graham modified the ’811 structure and filed for a patent. That patent, the one in issue, was granted in 1953. This suit against competing plow manufacturers resulted from charges by petitioners that several of respondents’ devices infringed the ’798 patent.

*22 The Prior Art.

Five prior patents indicating the state of the art were cited by the Patent Office in the prosecution of the 798 application. Four of these patents, 10 other United States patents and two prior-use spring-clamp arrangements not of record in the 798 file wrapper were relied upon by respondents as revealing the prior art. The District Court and the Court of Appeals found that the prior art “as a whole in one form or another contains all of the mechanical elements of the 798 Patent.” One of the prior-use clamp devices not before the Patent Examiner- — Glencoe—was found to have “all of the elements.”

We confine our discussion to the prior patent of Graham, ’811, and to the Glencoe clamp device, both among the references asserted by respondents. The Graham ’811 and 798 patent devices are similar in all elements, save two: (1) the stirrup and the bolted connection of the shank to the hinge plate do not appear in ’811; and (2) the position of the shank is reversed, being placed in patent ’811 above the hinge plate, sandwiched between it and the upper plate. The shank is held in place by the spring rod which is hooked against the bottom of the hinge plate passing through a slot in the shank. Other differences are of no consequence to our examination. In practice the ’811 patent arrangement permitted the shank to wobble or fishtail because it was not rigidly fixed to the hinge plate; moreover, as the hinge plate was below the shank, the latter caused wear on the upper plate, a member difficult to repair or replace.

Graham’s 798 patent application contained 12 claims. All were rejected as not distinguished from the Graham ’811 patent. The inverted position of the shank was specifically rejected as was the bolting of the shank to the hinge plate. The Patent Office examiner found these to be “matters of design well within the expected skill of *23the art and devoid of invention.” Graham withdrew the original claims and substituted the two new ones which are substantially those in issue here. His contention was that wear was reduced in patent ’798 between the shank and the heel or rear of the upper plate.11 He also emphasized several new features, the relevant one here being that the bolt used to connect the hinge plate and shank maintained the upper face of the shank in continuing and constant contact with the underface of the hinge plate.

Graham did not urge before the Patent Office the greater “flexing” qualities of the 798 patent arrangement which he so heavily relied on in the courts. The sole element in patent 798 which petitioners argue before us is the interchanging of the shank and hinge plate and the consequences flowing from this arrangement. The contention is that this arrangement — which petitioners claim is not disclosed in the prior art — permits the shank to flex under stress for its entire length. As we have sketched (see sketch, “Graham 798 Patent” in Appendix, Fig. 2), when the chisel hits an obstruction the resultant force (A) pushes the rear of the shank upward and the shank pivots against the rear of the hinge plate at (C). The natural tendency is for that portion of the shank between the pivot point and the bolted connection (i. e., between C and D) to bow downward and away from the hinge plate. The maximum dis*24tance (B) that the shank moves away from the plate is slight — for emphasis, greatly exaggerated in the sketches. This is so because of the strength of the shank and the short — nine inches or so — length of that portion of the shank between (C) and (D). On the contrary, in patent ’811 (see sketch, “Graham ’811 Patent” in Appendix, Fig. 2), the pivot point is the upper plate at point (c); and while the tendency for the shank to bow between points (c) and (d) is the same as in ’798, the shank is restricted because of the underlying hinge plate and cannot flex as freely. In practical effect, the shank flexes only between points (a) and (c), and not along the entire length of the shank, as in ’798. Petitioners say that this difference in flex, though small, effectively absorbs the tremendous forces of the shock of obstructions whereas prior art arrangements failed.

The Obviousness of the Differences.

We cannot agree with petitioners. We assume that the prior art does not disclose such an arrangement as petitioners claim in patent ’798. Still we do not believe that the argument on which petitioners’ contention is bottomed supports the validity of the patent. The tendency of the shank to flex is the same in all cases. If free-flexing, as petitioners now argue, is the crucial difference above the prior art, then it appears evident that the desired result would be obtainable by not boxing the shank within the confines of the hinge.12 The only other effective place available in the arrangement was to attach it below the hinge plate and run it through a *25stirrup or bracket that would not disturb its flexing qualities. Certainly a person having ordinary skill in the ■prior art, given the fact that the flex in the shank could be utilized more effectively if allowed to run the entire length of the shank, would immediately see that the thing to do was what Graham did, i. e., invert the shank and the hinge plate.

Petitioners’ argument basing validity on the free-flex theory raised for the first time on appeal is reminiscent of Lincoln Engineering Co. v. Stewart-Warner Corp., 303 U. S. 545 (1938), where the Court called such an effort “an afterthought. No such function ... is hinted at in the specifications of the patent. If this were so vital an element in the functioning of the apparatus it is strange that all mention of it was omitted.” At p. 550. No “flexing” argument was raised in the Patent Office. Indeed, the trial judge specifically found that “flexing is not a claim of the patent in suit . . .” and would not permit interrogation as to flexing in the accused devices. Moreover, the clear testimony of petitioners’ experts shows that the flexing advantages flowing from the ’798 arrangement are not, in fact, a significant feature in the patent-.13

We find no nonobvious facets in the ’798 arrangement. The wear and repair claims were sufficient to overcome *26the patent examiner’s original conclusions as to the validity of the patent. However, some of the prior art, notably Glencoe, was not before him. There the hinge plate is below the shank but, as the courts below found, all of the elements in the 798 patent are present in the Glencoe structure. Furthermore, even though the position of the shank and hinge plate appears reversed in Glencoe, the mechanical operation is identical. The shank there pivots about the underside of the stirrup, which in Glen-coe is above the shank. In other words, the stirrup in Glencoe serves exactly the same function as the heel of the hinge plate in 798. The mere shifting of the wear point to the heel of the 798- hinge plate from the stirrup of Glencoe — itself a part of the hinge plate — presents no operative mechanical distinctions, much less nonobvious differences.

B. The Patent in Issue in No. 87, Calmar, Inc. v. Cook Chemical Co., and in No. 43, Colgate-Palmolive Co. v. Cook Chemical Co.

The single patent14 involved in these cases relates to a plastic finger sprayer with a <lhold-down” lid used as a built-in dispenser for containers or bottles packaging liquid products, principally household insecticides. Only the first two of the four claims in the patent are involved here and we, therefore, limit our discussion to them. We do not set out those claims here since they are printed in 220 F. Supp., at 417-418.

In essence the device here combines a finger-operated pump sprayer, mounted in a container or bottle by means of a container cap, with a plastic overcap which screws over the top of and depresses the sprayer (see Appendix, *27Fig. 3). The pump sprayer passes through the container cap and extends down into the liquid in the container; the overcap fits over the pump sprayer and screws down on the outside of a collar mounting or retainer which is molded around the body of the sprayer. When the overcap is screwed down on this collar mounting a seal is formed by the engagement of a circular ridge or rib located above the threads on the collar mounting with a mating shoulder located inside the overcap above its threads.15 The overcap, as it is screwed down, depresses the pump plunger rendering the pump inoperable and when the seal is effected, any liquid which might seep into the overcap through or around the pump is prevented from leaking out of the overcap. The overcap serves also to protect the sprayer head and prevent damage to it during shipment or merchandising. When the overcap is in place it does not reach the cap of the container or bottle and in no way engages it since a slight space is left between those two pieces.

The device, called a shipper-sprayer in the industry, is sold as an integrated unit with the overcap in place enabling the insecticide manufacturer to install it on the container or bottle of liquid in a single operation in an automated bottling process. The ultimate consumer simply unscrews and discards the overcap, the pump plunger springs up and the sprayer is ready for use.

The Background of the Patent.

For many years manufacturers engaged in the insecticide business had faced a serious problem in developing sprayers that could be integrated with the containers or bottles in which the insecticides were marketed. Originally, insecticides were applied through the use of tin *28sprayers, not supplied by the manufacturer. In 1947, Cook Chemical, an insecticide manufacturer, began to furnish its customers with plastic pump dispensers purchased from Calmar. The dispenser was an unpatented finger-operated device mounted in a perforated cardboard holder and hung over the neck of the bottle or container. It was necessary for the ultimate consumer to remove the cap of the container and insert and attach the sprayer to the latter for use.

Hanging the sprayer on the side of the container or bottle was both expensive and troublesome. Packaging for shipment had to be a hand operation, and breakage and pilferage as well as the loss of the sprayer during shipment and retail display often occurred. Cook Chemical urged Calmar to develop an integrated sprayer that could be mounted directly in a container or bottle during the automated filling process and that would not leak during shipment or retail handling. Calmar did develop some such devices but for various reasons they were not completely successful. The situation was aggravated in 1954 by the entry of Colgate-Palmolive into the insecticide trade with its product marketed in aerosol spray cans. These containers, which used compressed gas as a propellent to dispense the liquid, did not require pump sprayers.

During the same year Calmar was acquired by the Drackett Company. Cook Chemical became apprehensive of its source of supply for pump sprayers and decided to manufacture its own through a subsidiary, Bakan Plastics, Inc. Initially, it copied its design from the unpatented Calmar sprayer, but an officer of Cook Chemical, Scoggin, was assigned to develop a more efficient device. By 1956 Scoggin had perfected the shipper-sprayer in suit and a patent was granted in 1959 to Cook Chemical as his assignee. In the interim Cook Chemical began to use Scoggin⅛ device and also mar*29keted it to the trade. The device was well received and soon became widely used.

In the meanwhile, Calmar employed two engineers, Corsette and Cooprider, to perfect a shipper-sprayer and by 1958 it began to market its SS-40, a device very much similar to Scoggin’s. When the Scoggin patent issued, Cook Chemical charged Calmar’s SS-40 with infringement and this suit followed.

The Opinions of the District Court and the Court of Appeals.

At the outset it is well to point up that the parties have always disagreed as to the scope and definition of the invention claimed in the patent in suit. Cook Chemical contends that the invention encompasses a unique combination of admittedly old elements and that patentability is found in the result produced. Its expert testified that the invention was “the first commercially successful, inexpensive integrated shipping closure pump unit which permitted automated assembly with a container of household insecticide or similar liquids to produce a practical, ready-to-use package which could be shipped without external leakage and which was so organized that the pump unit with its hold-down cap could be itself assembled and sealed and then later assembled and sealed on the container without breaking the first seal.” Cook Chemical stresses the long-felt need in the industry for such a device; the inability of others to produce it; and its commercial success — all of which, contends Cook, evidences the nonobvious nature of the device at the time it was developed. On the other hand, Calmar says that the differences between Scoggin’s shipper-sprayer and the prior art relate only to the design of the overcap and that the differences are so inconsequential that the device as a whole would have been obvious at the time of its invention to a person having ordinary skill in the art.

*30Both courts accepted Cook Chemical’s contentions. While the exact basis of the District Court’s holding is uncertain, the court did find the subject matter of the patent new, useful and nonobvious. It concluded that Scoggin “had produced a sealed and protected sprayer unit which the manufacturer need only screw onto the top of its container in much the same fashion as a simple metal cap.” 220 F. Supp., at 418. Its decision seems to be bottomed on the finding that the Scoggin sprayer solved the long-standing problem that had confronted the industry.16 The Court of Appeals also found validity in the “novel ‘marriage’ of the sprayer with the insecticide container” which took years in discovery and in “the immediate commercial success” which it enjoyed. While finding that the individual elements of the invention were “not novel per se” the court found “nothing in the prior art suggesting Scoggin’s unique combination of these old features ... as would solve the . . . problems which for years beset the insecticide industry.” It concluded that “the . . . [device] meets the exacting standard required for a combination of old elements to rise to the level of patentable invention by fulfilling the long-felt need with an economical, efficient, utilitarian apparatus which achieved novel results and immediate comiqercial success.” 336 F. 2d, at 114.

The Prior Art.

Only two of the five prior art patents cited by the Patent Office Examiner in the prosecution of Scoggin’s application are necessary to our discussion, i. e., Lohse *31U. S. Patent No. 2,119,884 (1938) and Mellon U. S. Patent No. 2,586,687 (1952). Others are cited by Cal-mar that were not before the Examiner, but of these our purposes require discussion of only the Livingstone U. S. Patent No. 2,715,480 (1953). Simplified drawings of each of these patents are reproduced in the Appendix, Figs. 4^6, for comparison and description.

The Lohse patent (Fig. 4) is a shipper-sprayer designed to perform the same function as Scoggin’s device. The differences, recognized by the District Court, are found in the overcap seal which in Lohse is formed by the skirt of the overcap engaging a washer or gasket which rests upon the upper surface of the container cap. The court emphasized that in Lohse “[t]here are no seals above the threads and below the sprayer head.” 220 F. Supp., at 419.

The Mellon patent (Fig. 5), however, discloses the idea of effecting a seal above the threads of the over-cap. Mellon’s device, likewise a shipper-sprayer, differs from Scoggin’s in that its overcap screws directly on the container, and a gasket, rather than a rib, is used to effect the seal.

Finally, Livingstone (Fig. 6) shows a seal above the threads accomplished without the use of a gasket or washer.17 Although Livingstone’s arrangement was designed to cover and protect pouring spouts, his sealing feature is strikingly similar to Scoggin’s. Livingstone uses a tongue and groove technique in which the tongue, located on the upper surface of the collar, fits into a groove on the inside of the overcap. Scoggin employed the rib and shoulder seal in the identical position and with less efficiency because the Livingstone technique *32is inherently a more stable structure, forming an interlock that withstands distortion of the overcap when subjected to rough handling. Indeed, Cook Chemical has now incorporated the Livingstone closure into its own shipper-sprayers as had Calmar in its SS-40.

The Invalidity of the Patent.

Let us first return to the fundamental disagreement between the parties. Cook Chemical, as we noted at the outset, urges that the invention must be viewed as the overall combination, or — putting it in the language of the statute — that we must consider the subject matter sought to be patented taken as a whole. With this position, taken in the abstract, there is, of course, no quibble. But the history of the prosecution of the Scoggin application in the Patent Office reveals a substantial divergence in respondent’s present position.

As originally submitted, the Scoggin application contained 15 claims which in very broad terms claimed the entire combination of spray pump and overcap. No mention of, or claim for, the sealing features was made. All 15 claims were rejected by the Examiner because (1) the applicant was vague and indefinite as to what the invention was, and (2) the claims were met by Lohse. Scoggin canceled these claims and submitted new ones. Upon a further series of rejections and new submissions, the Patent Office Examiner, after an office interview, at last relented. It is crystal clear that after the first rejection, Scoggin relied entirely upon the sealing arrangement as the exclusive patentable difference in his combination. It is likewise clear that it was on that feature that the Examiner allowed the claims. In fact, in a letter accompanying the final submission of claims, Scoggin, through his attorney, stated that “agreement was reached between the Honorable Examiner and applicant’s attorney relative to limitations which must be in the claims in *33order to define novelty over the previously applied disclosure of Lohse when considered in view of the newly cited patents of Mellon and Darley, Jr.” (Italics added.)

Moreover, those limitations were specifically spelled out as (1) the use of a rib seal and (2) an overcap whose lower edge did not contact the container cap. Mellon was distinguished, as was the Darley patent, infra, n. 18, on the basis that although it disclosed a hold-down cap with a seal located above the threads, it did not disclose a rib seal disposed in such position as to cause the lower peripheral edge of the overcap “to be maintained out of contacting relationship with [the container] cap . . . when . . . [the overcap] was screwed [on] tightly . . . .” Scoggin maintained that the “obvious modification” of Lohse in view of Mellon would be merely to place the Lohse gasket above the threads with the lower edge of the overcap remaining in tight contact with the container cap or neck of the container itself. In other words, the Scoggin invention was limited to the use of a rib — rather than a washer or gasket — and the existence of a slight space between the overcap and the container cap.

It is, of course, well settled that an invention is construed not only in the light of the claims, but also with reference to the file wrapper or prosecution history in the Patent Office. Hogg v. Emerson, 11 How. 587 (1850); Crawford v. Heysinger, 123 U. S. 589 (1887). Claims as allowed must be read and interpreted with reference to rejected ones and to the state of the prior art; and claims that have been narrowed in order to obtain the issuance of a patent by distinguishing the prior art cannot be sustained to cover that which was previously by limitation eliminated from the patent. Powers-Kennedy Co. v. Concrete Co., 282 U. S. 175, 185-186 (1930); Schriber Co. v. Cleveland Trust Co., 311 U. S. 211, 220-221 (1940).

*34Here, the patentee obtained his patent only by accepting the limitations imposed by the Examiner. The claims were carefully drafted to reflect these limitations and Cook Chemical is not now free to assert a broader view of Scoggin’s invention. The subject matter as a whole reduces, then, to the distinguishing features clearly incorporated into the claims. We now turn to those features.

As to the space between the skirt of the overcap and the container cap, the District Court found:

“Certainly without a space so described, there could be no inner seal within the cap, but such a space is not new or novel, but it is necessary to the formation of the seal within the hold-down cap.

“To me this language is descriptive of an element of the patent but not a part of the invention. It is too simple, really, to require much- discussion. In this device the hold-down cap was intended to perform two functions — to hold down the sprayer head and to form a solid tight seal between the shoulder and the collar below. In assembling the element it is necessary to provide this space in order to form the seal.” 220 F. Supp., at 420. (Italics added.)

The court correctly viewed the significance of that feature. We are at a loss to explain the Examiner’s allowance on the basis of such a distinction. Scoggin was able to convince the Examiner that Mellon’s cap contacted the bottle neck while his did not. Although the drawings included in the Mellon application show that the cap might touch the neck of the bottle when fully screwed down, there is nothing — absolutely nothing — which indicates that the cap was designed at any time to engage the bottle neck. It is palpably evident that Mellon embodies a seal formed by a gasket com*35pressed between the cap and the bottle neck. It follows that the cap in Mellon will not seal if it does not bear down on the gasket and this would be impractical, if not impossible, under the construction urged by Scoggin before the Examiner. Moreover, the space so strongly asserted by Cook Chemical appears quite plainly on the Livingstone device, a reference not cited by the Examiner.

The substitution of a rib built into a collar likewise presents no patentable difference above the prior art. It was fully disclosed and dedicated to the public in the Livingstone patent. Cook Chemical argues, however, that Livingstone is not in the pertinent prior art because it relates to liquid containers having pouring spouts rather than pump sprayers. Apart from the fact that respondent made no such objection to similar references cited by the Examiner,18 so restricted a view of the applicable prior art is not justified. The problems confronting Scoggin and the insecticide industry were not insecticide problems; they were mechanical closure problems. Closure devices in such a closely related art as pouring spouts for liquid containers are at the very least pertinent references. See, II Walker on Patents § 260 (Deller ed. 1937).

Cook Chemical insists, however, that the development of a workable shipper-sprayer eluded Calmar, who had long and unsuccessfully sought to solve the problem. And, further, that the long-felt need in the industry for a device such as Scoggin’s together with its wide commercial success supports its patentability. These legal in*36ferences or subtests do focus attention on economic and motivational rather than technical issues and are, therefore, more susceptible of judicial treatment than are the highly technical facts often present in patent litigation. See Judge Learned Hand in Reiner v. I. Leon Co., 285 F. 2d 501, 504 (1960). See also Note, Subtests of “Non-obviousness”: A Nontechnical Approach to Patent Validity, 112 U. Pa. L. Rev. 1169 (1964). Such inquiries may lend a helping hand to the judiciary which, as Mr. Justice Frankfurter observed, is Inost ill-fitted to discharge the technological duties cast upon it by patent legislation. Marconi Wireless Co. v. United States, 320 U. S. 1, 60 (1943). They may also serve to “guard against slipping into use of hindsight,” Monroe Auto Equipment Co. v. Heckethom Mfg. & Sup. Co., 332 F. 2d 406, 412 (1964), and to resist the temptation to read into the prior art the teachings of the invention in issue.

However, these factors do not, in the circumstances of this case, tip the scales of patentability. The Scoggin invention, as limited by the Patent Office and accepted by Scoggin, rests upon exceedingly small and quite nontechnical mechanical differences in a device which was old in the art. At the latest, those differences were rendered apparent in 1953 by the appearance of the Livingstone patent, and unsuccessful attempts to reach a solution to the problems confronting Scoggin made before that time became wholly irrelevant. It is also irrelevant that no one apparently chose to avail himself of knowledge stored in the Patent Office and readily available by the simple expedient of conducting a patent search — a prudent and nowadays common preliminary to well organized research. Mast, Foos & Co. v. Stover Mfg. Co., 177 U. S. 485 (1900). To us, the limited claims of the Scoggin patent are clearly evident from the prior art as it stood at the time of the invention.

*37We conclude that the claims in issue in the Scoggin patent must fall as not meeting the test of § 103, since the differences between them and the pertinent prior art would have been obvious to a person reasonably skilled in that art.

The judgment of the Court of Appeals in No. 11 is affirmed. The judgment of the Court of Appeals in Nos. 37 and 43 is reversed and the cases remanded to the District Court for disposition not inconsistent with this opinion.

It is so ordered.

Me. Justice Stewart took no part in the consideration or decision of Nos. 37 and 43.

Mb. Justice Fortas took no part in the consideration or decision of these cases.

*0

MOORE, Circuit Judge.

Bombardier Recreational Products Inc. and BRP U.S. Inc. (collectively, “BRP”) appeal from the United States District Court for the Southern District of Florida’s denial of judgment as a matter of law that the asserted claims of U.S. Patent Nos. 6,568,969 (“’969 patent”) and 6,793,-545 (“’545 patent”) would have been obvious, that Arctic Cat Inc. (“Arctic Cat”) failed to mark patented products, that the jury’s royalty award was based on improper expert testimony, and that BRP did not willfully infringe the asserted claims. BRP also appeals the district court’s decision to treble damages and its award of an ongoing royalty to Arctic Cat. We affirm the district court’s denial of judgment as a matter of law as to obviousness, the jury’s royalty rate, and willfulness. We affirm the *1357district court’s decision to treble damages and award an ongoing royalty to Arctic Cat. We vacate the court’s denial of judgment as a matter of law as to marking and remand for further consideration limited to that issue.

Background

The ’969 and ’645 patents disclose a thrust steering system for personal watercraft (“PWC”) propelled by jet stream. This type of watercraft is propelled by discharging water out of a discharge nozzle at the rear of the watercraft. E.g., ’545 patent at 1:22-24. The rider controls the thrust of water out of the discharge nozzle by pressing a lever mounted on the steering handle. Id. at 1:38-40. A sufficient amount of thrust out of the steering nozzle is required for these watercraft to steer properly because decreasing the thrust of the water out of the discharge nozzle decreases the steering capability of the watercraft. Id. at 1:34-36,1:51-55.

Because steering capabilities are affected by the amount of thrust applied, the patents explain that, to avoid obstacles at high speed, riders should apply constant pressure on the throttle lever while simultaneously turning the steering handle away from the obstacle. Id. at 1:59-61. This is counter-intuitive to inexperienced riders who often slow down to turn out of the way. Id. at 1:55-65. In these situations a rider may not be able to avoid the obstacle because steering capability has been decreased. Id. at 1:65-67. The patents seek to overcome this issue by automatically providing thrust when riders turn the steering system. Id. at 2:11-27. Claim 15 of the ’545 patent is representative:

A watercraft including:

a steering mechanism;

a steering nozzle;

a thrust mechanism; a lever adapted to allow an operator to manually control thrust of said thrust mechanism, said lever mounted on said steering mechanism and biased toward an idle position; and

a controlled thrust steering system for controlling thrust of said thrust mechanism independently of the operator;

wherein said controlled thrust steering system activates said thrust mechanism to provide a steerable thrust after said lever is positioned other than to provide a steerable thrust and after the steering mechanism is positioned for turning said watercraft.

Arctic Cat sued BRP for infringement of claims 13, 15, 17,19, 25, and 30 of the ’545 patent and claims 15-17, and 19 of the ’969 patent, accusing the off-throttle thrust reapplication system in several of BRP’s Sea-Doo PWC. BRP refers to its proprietary off-throttle thrust reapplication system as Off-Throttle Assisted Steering (“OTAS”). Before trial, BRP unsuccessfully moved for summary judgment on several issues, including that Arctic Cat’s sole licensee Honda failed to mark its products with the licensed patent numbers.

At trial, the jury found both patents not invalid, awarded a royalty consistent with Arctic Cat’s model ($102.54 per unit) to begin on October 16, 2008, and found by clear and convincing evidence that BRP willfully infringed the asserted claims. Based on the willfulness verdict, the district court trebled damages, a decision it further explained in a subsequent order.

After post-trial briefing, the district court denied BRP’s renewed motion for judgment as a matter of law on all issues. It granted Arctic Cat’s motion for an ongoing royalty, awarding $205.08 per unit. BRP appeals the district court’s denial of judgment as a matter of law on validity, marking, damages, and willfulness, as well *1358as its grant of an ongoing royalty- and decision to treble damages. We have jurisdiction under 28 U.S.C. § 1295(a)(1).

Discussion

In appeals of patent cases, we apply the law of the regional circuit “tó which district court appeals normally lie, unless the issue pertains to or is unique to patent law.” AbbVie Deutschland GmbH & Co., KG v. Janssen Biotech, Inc., 759. F.3d 1285, 1295 (Fed. Cir. 2014) (internal quotation marks omitted). We review rulings on motions for judgment as a matter of law under the law of the regional circuit. Id. The Eleventh Circuit reviews the denial of judgment as a matter- of law de novo, viewing the evidence in the light most favorable to the non-moving party. Howard v. Walgreen Co., 605 F.3d 1239, 1242 (11th Cir. 2010). “The motion should be granted only when the plaintiff presents no legally sufficient evidentiary, basis, for a reasonable jury to find for him on a material element of his cause of action.” Id. (internal quotation marks omitted).

I, Obviousness

Obviousness is a question of law based on underlying facts. WBIP, LLC v. Kohler Co., 829 F.3d 1317, 1326 (Fed. Cir. 2016), In KSR International Co. v. Teleflex Inc., 550 U.S. 398, 419, 127 S.Ct. 1727, 167 L.Ed.2d 705 (2007), the Supreme Court cautioned that the obviousness analysis should not-be reduced, to “rigid and mandatory formulas.” In Graham v. John Deere Co., the Supreme Court set the framework for the obviousness inquiry under 35 U.S.C. § 103:

Under § 103, the scope and content , of the prior art are to be determined; differences between-the prior art and the claims at issue are to be ascertained; and the level of ordinary skill in the pertinent art resolved. Against this background, the obviousness or nonobvi-ousness of the subject matter is determined. Such secondary considerations as commercial success,' long felt but unsolved, needs, failure of others, etc., "might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented.

383 U.S. 1, 17-18, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966). The Graham factors—(1) the scope and content of the prior art; (2) the differences between the claims and the prior art; (3) the level of ordinary skill in the art; and (4) objective considerations of nonobviousness—are questions of fact reviewed for substantial evidence. See Apple Inc. v. Samsung Elecs. Co., 839 F.3d 1034, 1047-48 (Fed. Cir. 2016) (en banc); In re Cyclobenzaprine Hydrochloride Extended-Release Capsule Patent Litig., 676 F.3d 1063, 1068 (Fed. Cir. 2012). “When reviewing a denial of judgment as a matter of law of obviousness, where there is a black box jury verdict, as is the case here, we presume the jury resolved underlying factual disputes in favor of the verdict winner and leave those presumed findings undisturbed if supported by substantial evidence.” WBIP, 829 F.3d at 1326. We examine the legal conclusion de novo in light of those facts. Id.

“A determination of whether a patent claim is invalid as obvious under § 103 requires consideration of all four Graham factors, and it is error to reach a conclusion of obviousness until all of those factors are considered.” Apple, 839 F.3d at 1048. This includes objective indicia of non-obviousness, which must be considered in every case where present. See, e.g., id. at 1048 & n.13; Millennium Pharm., Inc. v. Sandoz Inc., 862 F.3d 1356, 1368-69 (Fed. Cir. 2017); Merck & Cie v. Gnosis S.p.A,, 808 F.3d 829, 837 (Fed. Cir. 2015); Bristol-Myers Squibb Co. v. Teva Pharm., USA *1359 Inc., 752 F.3d 967, 977 (Fed. Cir. 2014); Transocean Offshore Deepwater Drilling, Inc. v. Maersk Drilling USA, Inc., 699 F.3d 1340, 1349 (Fed. Cir. 2012). “This requirement is in recognition of the fact that each of the Graham factors helps inform the ultimate obviousness determination.” Apple, 839 F.3d at 1048. Objective indicia of nonobviousness are considered collectively with the other Graham factors because they “serve to ‘guard against slipping into use of hindsight,’ and to resist the temptation .to read into the prior art the teachings of the invention,in issue.” Graham, 383 U.S. at 36, 86 S.Ct. 684 (citation omitted); see also KSR, 550 U.S. at 415, 127 S.Ct. 1727 (inviting court “to look at any secondary considerations that would prove instructive”); Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1538-39 (Fed. Cir. 1983) (noting that evidence of these factors must be considered with all the evidence and “not just when the decision-maker remains in doubt after reviewing the art”).

Also a fact question is whether one of ordinary skill in the art had a motivation to combine the prior art to achieve the claimed combination. Apple, 839 F.3d at 1047-48, 1051; Wyers v. Master Lock Co., 616 F.3d 1231, 1237-39 (Fed. Cir. 2010). “In KSR, the Supreme Court criticized a rigid approach to determining obviousness based on the disclosures of individual prior-art references, with little recourse to the knowledge, creativity, and common sense that an ordinarily skilled artisan would have brought to bear when considering .combinations or modifications.” Randall Mfg. v. Rea, 733 F.3d 1355, 1362 (Fed. Cir. 2018). Therefore, a motivation to combine can be found explicitly or implicitly in the prior art, references themselves, in market forces, in design incentives, or in “any need or problem known in the field of endeavor at the time of invention and addressed by the patent.” KSR, 550 U.S. at 420-21, 127 S.Ct. 1727; accord Plantronics, Inc. v. Aliph, Inc., 724 F.3d 1343, 1354 (Fed. Cir. 2013) (“Motivation to combine may be found explicitly or implicitly in market forces; design incentives; the ‘interrelated teachings of multiple patents’; ‘any need or problem known in the field of endeavor at the time of invention and addressed by the patent’; and the background knowledge, creativity, and common sense of the person of ordinary skill.” (quoting Perfect Web Techs., Inc. v. InfoUSA, Inc., 587 F.3d 1324, 1328-29 (Fed. Cir. 2009))).’ “The court should consider a range of real-world facts to determine ‘whether' there was an apparent reason to combine the known elements in the fashion claimed by the patent at issue.’ ” Intercontinental Great Brands LLC v. Kellogg N. Am. Co., 869 F.3d 1336, 1344 (Fed. Cir. 2017) (quoting KSR, 550 U.S. at 418, 127 S.Ct. 1727). Motivation to combine is a factual determination as to whether there is a known reason a skilled artisan would have been motivated to combine elements 'to arrive át a claimed combination. This is not the ultimate'legal determination of whether the claimed combination would have been obvious to the ordinary artisan—meaning that it is possible that a reason or m’otivation may exist, but nonetheless the ordinary artisan would not have found the combination obvious.

When a challenger shows that a “motivation” existed for a relevant skilled artisan to combine prior art in the way claimed in the patent at issue, such a showing commonly supports and leads readily to the further, ultimate determination that such an artisan; using ordinary creativity, would actually have found the claimed invention obvious. But the latter conclusion does not follow automatically from the former finding, and additional evidence may prevent drawing it. ... Even with a motivation *1360proved, the record may reveal reasons that, after all, the court should not- conclude that the combination would have been obvious ....

Intercontinental Great Brands, 869 F.3d at 1346-47.

Determining whether a claimed combination would have been obvious to a skilled artisan requires consideration of all the facts, no one of which is dispositive. The prior art, skill, and knowledge of the ordinarily skilled artisan may present a motivation or reason to combine. The prior art, skill, and knowledge of an ordinary artisan may also provide reasons not to combine which would likewise be a question of fact. For example, a reference may be found to teach away from a claimed combination, also a question of fact. Allergan, Inc. v. Sandoz Inc., 796 F.3d 1293, 1305 (Fed. Cir. 2015). Prior art teaches, away when “a person of ordinary skill, upon reading the reference, would be discouraged from following the path set out in the reference, or would be led in a direction divergent from the path that was taken by the applicant.” Id. (citing In re Gurley, 27 F.3d 551, 553 (Fed. Cir. 1994)). In fact, the prior art could contain one reference suggesting a combination and others critiquing or otherwise discouraging the same. Even a .single reference can include both types of statements, and we have held that it is error to fail to consider the entirety of the art. See, e.g., W.L. Gore & Assocs., Inc. v. Garlock, Inc., 721 F.2d 1540, 1550 (Fed. Cir. 1983) (holding the district court erred by “considering the references in less than their entireties, i.e., in disregarding disclosures in the references that diverge from and teach away from the invention at hand”).

“The degree of teaching away will of course depend on the particular facts; in general, a reference will teach away if it suggests that the line of development flowing from the reference’s disclosure is unlikely to be productive of the résult sought by the applicant.” Gurley, 27 F.3d at 553. As our precedent reflects, prior art need not explicitly “teach away” to be relevant to the obviousness determination. Implicit in our discussion of the “degree” of teaching away is an understanding that some references may discourage more than others. Id.; see also Meiresonne v. Google, Inc., 849 F.3d 1379, 1382 (Fed. Cir. 2017) (“A reference that ‘merely expresses a general preference for an alternative invention but does not criticize, discredit, or otherwise discourage investigation into’ the claimed invention does not teach away.” (quoting Galderma Labs., L.P. v. Tolmar, Inc., 737 F.3d 731, 738 (Fed. Cir. 2013))); DePuy Spine, Inc. v. Medtronic Sofamor Danek, Inc., 567 F.3d 1314, 1327 (Fed. Cir. 2009) (affirming district court’s finding of teaching away where the reference “expressed] concern for failure”); ATD Corp. v. Lydall, Inc., 159 F.3d 534, 546 (Fed. Cir. 1998) (reversing a judgment of invalidity in part because references “cautioned against compressing the layers in a multilayer insulator”). .Indeed, the Supreme Court has long held that “known disadvantages in old devices which would naturally discourage the search for new inventions may be taken into account in determining obviousness.” United States v. Adams, 383 U.S. 39, 52, 86 S.Ct. 708, 15 L.Ed.2d 572 (1966) (emphasis added).

Such understandings about reasons to combine or countervailing reasons not to combine could come from the knowledge, skill, and creativity of the ordinarily skilled artisan. KSR, 550 U.S. at 418, 127 S.Ct. 1727. We have held that where a party argues a skilled artisan would have been motivated to combine references, it must show the artisan “would have had a reasonable expectation of sue-*1361cess from doing so.” Cyclóbenzapñne, 676 F.3d at 1068-69. Thus, if an ordinarily skilled artisan would not believe that a particular combination would have a reasonable expectation of “anticipated success,” the combination may not be obvious. See KSR, 550 U.S. at 421, 127 S.Ct. 1727. Whether a reasonable expectation of success exists is yet another fact question. PAR Pharm., Inc. v. TWI Pharm., Inc., 773 F.3d 1186, 1196 (Fed. Cir. 2014). Thus, whether there exist reasons a skilled artisan would combine or reasons a skilled artisan would not combine are entirely factual determinations to which deference must be given. Once all relevant facts are found, the ultimate legal determination involves the weighing of the fact findings to conclude whether the claimed combination would have been obvious to an ordinary artisan.

The jury, in this case, determined that BRP failed to prove by clear and convincing evidence that the claims at issue would have been obvious to a skilled artisan. BRP moved for judgment as a matter of law on obviousness, which the district court denied. On appeal, BRP argues that the district court erred in refusing to grant it judgment as a matter of law that the asserted claims would have been obvious in light of the off-throttle thrust reapplication system in its 1997 Sea-Doo Challenger 1800 Jet Boat (“Challenger”) and an existing PWC such as the 1998 Sea-Doo GTX RFI (“GTX”). There is no serious dispute that the Challenger system and a PWC like the GTX disclose all elements of most of the asserted claims.1 BRP argues a reasonable jury could only have concluded an ordinarily skilled artisan would have been motivated to combine Challenger and a PWC, and that objective indicia of nonobviousness confirm the asserted claims would have been obvious. BRP disputes whether substantial evidence exists for particular jury fact findings and the ultimate legal determination of obviousness.

BRP argues that “KSR compels a finding of obviousness.” Appellant’s Br. 21. It argues that a conclusion of obviousness must be reached because there was “a design need or market pressure to solve a problem” and the combination is one of “a finite number of identified, predictable solutions.” Id. at 24 (quoting KSR, 550 U.S. at 421, 127 S.Ct. 1727).

In order to show a skilled artisan would have been motivated to modify a PWC with Challenger’s off-throttle thrust reapplication system, BRP principally relies on two prior art reports written by the Society of Automotive Engineers (“SAE”) that studied “personal watercraft as test vehicles in order to evaluate and test emerging off throttle steering concepts and devices.” J.A. 7530. Specifically, the SAE Interim and Draft Final Reports suggested using the Challenger system in a PWC to address the off-throttle steering problem. The Draft Final Report concluded if the Challenger system were applied to PWCs, “performance characteristics would remain unchanged when operated properly, but when off-throttle steering and panic was sensed, then some additional steering torque would automatically be restored.” J.A. 7577. BRP also cites additional references that it argues provided PWC manufacturers with overwhelming pressure to implement solutions to off-throttle steering so that riders could safely avoid obstacles. For example, the National Transportation *1362Safety Board recommended PWC manufacturers “consider ... off-throttle steerr ing” “to improve operator control and to help prevent personal injuries.” J.A. 7944. The National Association of State Boating Law Administrators similarly pressured the PWC industry to address the “disproportionate number of accidents” attributed to “ ‘off-throttle’ 'Steering loss” in PWCs. J.A. 9586. BRP also argues its own patent application, Canadian Patent Appl. 2,207,-938 (“Rheault ’938”), and patent, U.S. Patent No. 6,336,833 (“Rheault ’833”), disclose a throttle reapplication system and suggest its use in a PWC.

BRP argues a skilled artisan would have selected the Challenger system because it was one of a finite number of,identified, predictable solutions to the problem of off-throttle steering in PWCs. See KSR, 550 U.S. at 421, 127 S.Ct. 1727. The SAE Interim and Draft Final Reports identified the Challenger’s throttle reapplication as one of four solutions to the problem of off-throttle steering, along with rudders, flaps, and scoops. The National Association of State Boating Law Administrators also explained that jet boats and PWCs are similar and off-throttle directional control is a problem for both. Rheault ’833 disclosed that its steer-responsive throttle “is- applicable to single-engine, personal watercraft,” and Rheault ’938 states the Challenger jet boat’s thrust steering “is applicable to all types of watercraft vehicles, including personal watercraft vehicles.” J.A. 8942 at Abstract; J.A, 8920 at 8:15-17. For these reasons, BRP argues a reasonable jury could only have found a skilled artisan would have been motivated to modify a PWC with Challenger’s off-throttle thrust reapplication system, which would have combined known elements to improve the system in the same way and yield expected results.

We presume the jury found that an ordinarily skilled artisan would not have been motivated to combine the Challenger system with a PWC given its determination that the asserted claims are not invalid as obvious. If such a fact finding is supported by substantial evidence, we may not reverse it. In briefing and oral argument, BRP devoted much of its argument to re-litigating its case and its evidence rather than addressing the evidence that could have supported the jury’s finding of no motivation to combine. We do not reweigh the evidence, and reach our own factual determination regarding motivation. The question for us on appeal is only whether substantial evidence supports the jury’s presumed finding. See, e.g., Apple, 839 F.3d at 1052 (“Our job is not to review whether Samsung’s losing position was also supported by substantial evidence or to weigh the relative strength of Samsung’s evidence against Apple’s evidence. We are limited to determining whether there was substantial' evidence for ■ the jury’s findings, on the entirety of the record.”). We conclude that it does.

A reasonable jury could have found that a skilled artisan would not have been moti-vatéd to combine Challenger and a PWC. The SAE reports identified the combination of the Challenger system with a PWC to address the off-throttle steering problem, tested the Challenger for that pun-pose, and noted potential benefits of the combination. E.g,, J.A. 7577. But the reports did not stop there. The Draft Final Report also stated that “additional new hazards can be envisioned with such a steering system,” including collisions “when inadvertent activation of restored thrust might occur close to other boats, swimmers or fixed objects.” Id, It explained that because these hazards do not currently exist, “it is difficult to predict the frequency with which such events may occur.” • Id. Kevin Breen,- an author of the *1363SAE reports and BRP’s expert at trial, testified that automatic throttle reapplicar tion without “smart” engine controls could be dangerous. J.A. 2361-r62. The Draft Final Report likewise identified potential problems with proposed “smart” engine controls, which “would only become activated when needed and would not otherwise effect [sic] handling.” J.A. 7577. The report noted some “obvious” problems with this technology, such as the.system performing in a manner contrary to the operator’s intentions, and the need for the system to account for several variables “to be beneficial.” Id. As to the thought process behind “smart” engine controls, Mr. Breen testified that throttle reapplication “would be useful only if they were smart or on demand, as opposed to they just happened.” J.A. 2231-32. The claimed invention, in contrast, “just happen[s]” when the rider turns the steering mechanism. Id.-, see, e.g., 545 patent at claim 1.

This evidence may not rise to the level of teaching away; Nonetheless, in light of this record, the jury’s determination that there was no motivation to make this particular combination is supported by substantial evidence. Evidence suggesting reasons to combine cannot be viewed in a vacuum apart from evidence suggesting reasons not to combine. In this case, the same reference suggests a reason to combine, but also suggests reasons that a skilled artisan would be discouraged from pursuing such a combination. Under such circumstances, the jury’s fact finding regarding motivation is supported by substantial evidence. Coupled with testimony confirming the potential problems of automatic throttle reapplication and suggesting an alternative approach might reduce those same problems, J.A. 2230-32, a jury could find a skilled artisan would not have been motivated to combine the Challenger system with a PWC to .arrive at the claimed combination. ,.

BRP argues that the SAE reports demonstrate market pressure to solve a problem and a finite number of predictable solutions; in fact, BRP argues there were only four articulated solutions. Appellant’s Br. 9. In KSR, the Supreme Court explained when there is “market pressure to solve a problem and there- are a finite number of identified, predictable solutions, a person of ordinary skill has good reason to pursue the known options-within his or her technical grasp.” 550 U.S. at 421, 127 S.Ct. 1727. While the SAE reports identified the Challenger system, rudders, flaps, and scoops as potential solutions to the problem of off-throttle steering in PWCs, a reasonable jury could have determined that more than four solutions existed. At trial, Arctic Cat’s expert and named inventor Fred Bernier testified there were “various fins” and “a variety of things tried over a course of a number of years,” including modifying where the appendages attached to the PWC. J.A. 1219-21. BRP’s expert Richard Simard also testified BRP built seventeen prototypes incorporating various approaches over the course of five years. J.A.. 1951-57. An internal BRP “brainstorming” session identified thirty-two possible designs directed to off-throttle steering. J.A. 9454.. And there is evidence that other potential solutions to the off-throttle steering problem existed but were not fully disclosed for confidentiality and other concerns. See, e.g., J.A. 7532 (noting that some ideas offered in response to SAE’s inquiry “have typically either not conveyed sufficient information or have patent, propriety, or litigation concerns”).

A reasonable jury also could have found that modifying a PWC with the Challenger system would not have been a predictable solution yielding, expected . results. Mr. Bernier testified “[i]t was quite a—quite a *1364surprise, actually” when his team realized the technology worked on a prototype PWC. J.A. 1232. So did Mr. Simard, who testified “[w]e were surprised” that Proto-14, BRP’s prototype incorporating the Challenger system with a PWC, was “pretty good in forward speed.” J.A. 1960. He also admitted “what works on a jet boat may not work on a personal watercraft.” Id, And although Arctic Cat’s expert Dr. Bernard Cuzzillo testified the Challenger system reapplies some throttle when steering, he also testified he did not know whether the Challenger system was “adequate to qualify as a steerable thrust” and that it would “not necessarily” comprise a “controlled-thrust steering system.” J.A. 2876-78. This testimony along with the SAE reports’ own cautions about potential hazards of the combination provide substantial evidence upon which a jury could conclude that a skilled artisan would not have “anticipated success” with the claimed combination.

A. Objective Considerations

At trial, Arctic Cat argued the claimed invention received industry praise and satisfied a long-felt need. We presume the jury found in favor of Arctic Cat as to each of these objective considerations. We will not reverse these presumed findings if supported by substantial evidence.

1. Industry Praise

“Evidence that the industry praised a claimed invention or a product that embodies the patent claims weighs against an assertion that the same claimed invention would have been obvious.” Apple, 839 F.3d at 1053; accord Institut Pasteur & Universite Pierre Et Marie Curie v. Focarino, 738 F.3d 1337, 1347 (Fed. Cir. 2013) (“[Ijndustry praise ... provides probative and cogent evidence that one of ordinary skill in the art would not have reasonably expected [the claimed invention].”).

At trial, Arctic Cat introduced a press release it issued after Captain Michael Holmes, chief of the U.S. Coast Guard Office of Boating Safety, rode and evaluated an Arctic Cat prototype incorporating the claimed invention. J.A. 9537. After his test ride, Captain Holmes stated: “I like it. It’s one of the most impressive innovations I’ve seen all year.” J.A. 9537. He continued, “What I saw today will help us move forward in developing a realistic, achievable standard for a control and safety issue that we need to address. I’m particularly encouraged that this amount of quick-turn control can be achieved without some of the negative handling or safety ramifications that seem to accompany fins or rudders.” J.A. 9537. And Mr. Bernier testified others at the prototype demonstration “were very impressed with the system and how it worked” and said “it was the first time they had seen something that had a viable chance of resolving the off-throttle steering issues.” J.A. 1237.

BRP argues that substantial evidence does not support the jury’s presumed factual finding that the claimed invention received industry praise because “praise from a Coast Guard official in Arctic Cat’s own press release” is a “hearsay statement [that] cannot overcome persuasive evidence that the claimed technology described the same approach as BRP’s system.” Appellant’s Br. 35-36 (citing J.A. 7828; J.A. 7871; In re Cree, Inc., 818 F.3d 694, 702 (Fed. Cir. 2016)). We disagree for at least two reasons. First, Arctic Cat contends—and BRP does not contest—that BRP failed to object to this evidence as hearsay at trial, so the jury was entitled to credit the statement. Appellee’s Br. 12. Second, that Captain Holmes’ statements appear in Arctic Cat’s press release goes to evidentiary weight. Captain Holmes’ *1365statements and Mr. Bernier’s testimony constitute substantial evidence to support the jury’s presumed factual finding that the claimed invention received praise from the industry. This evidence of industry recognition of the significance and value of the claimed invention weighs in favor of nonobviousness.

2. Long-Felt Need

“Evidence of a long felt but unresolved need tends to show non-obviousness because it is reasonable to infer that the need would have not persisted had the solution been obvious.” WBIP, 829 F.3d at 1332; see also Iron Grip Barbell Co. v. USA Sports, Inc., 392 F.3d 1317, 1325 (Fed. Cir. 2004) (“Absent a showing of long-felt need or the failure of others, the mere passage of time without the claimed invention is not evidence of nonobviousness.”).

BRP does not dispute there was a long-felt need in the area of off-throttle steering and PWC rider safety. Rather, it argues Arctic Cat’s invention did not satisfy this long-felt need because the Challenger system already solved off-throttle steering. Substantial evidence supports the jury’s presumed finding that the claimed invention solved the problem of off-throttle steering. The SAE Draft Final Report noted “an effort has been ongoing to develop this [off-throttle steering] technology for more than three decades with little commercially viable success.” J.A. 7575. The report summarized test results of the Challenger system for its off-throttle steering capabilities and offered potential pros and cons of its use in a PWC; it did not summarize test results of a Challenger-PWC combination. And Mr. Breen conceded at trial that despite a number of people working to address the off-throttle steering problem, there “was not a publicly available personal watercraft with throttle reapplication” before Arctic Cat’s invention. Appellee’s Br. 41 (citing J.A. 2337). This is substantial evidence to support the jury’s fact finding about long-felt need. This long-felt need weighs in favor of the nonobviousness of the claimed invention.

B. Legal Conclusion

We consider whether the claimed invention would have been obvious de novo, in light of the jury’s underlying factual findings. WBIP, 829 F.3d at 1326. The evidence suggesting a motivation to combine is tempered by the evidence suggesting the Challenger-PWC combination could have serious problems, that “smart” engine controls might better address those problems, that the combination was not one of only four possible solutions, and that the combination did not yield expected results. We cannot under these circumstances reverse any of the jury’s presumed fact findings regarding motivation to combine or expectations of success. In light of these fact findings along with the objective indicia of nonobviousness, which also weigh in favor of nonobviousness, we see no error in the conclusion that BRP failed to prove that the-asserted claims would have been obvious by clear and convincing evidence.

II. Marking -

Pursuant to 35 U.S.C. § 287(a), a patentee who makes or sells a patented article must mark his articles or notify infringers of his patent in order to recover damages. See Dunlap v. Schofield, 152 U.S. 244, 248, 14 S.Ct. 576, 38 L.Ed. 426 (1894). Section 287(a) provides:

Patentees, and persons making, offering for sale, or selling within the United States any patented article for or under them, or importing any patented article into the United States, may give notice to the public that the same is patented, either by fixing thereon the word “pat*1366ent” or the abbreviation “pat”, together with the number of the patent .... In the event of failure so to mark, no damages shall be recovered by the patentee in any action for infringement, except on proof that the infringer was notified of the infringement and continued to infringe thereafter, in which event damages may be recovered only for infringement occurring after such notice.

The patentee bears the burden of pleading and proving he complied with § 287(a)’s marking requirement. Maxwell v. J. Baker, Inc., 86 F.3d 1098, 1111 (Fed. Cir. 1996); see also Dunlap, 162 U.S. at 248, 14 S.Ct. 576 (“[T]he duty of alleging and the burden of proving either [actual or constructive notice] is upon the [paten-tee].”). Whether a patentee’s articles have been marked “is a matter peculiarly within his own knowledge ....” Dunlap, 152 U.S. at 248, 14 S.Ct. 576. If a patentee who makes, sells, offers for sale, or imports his patented articles has not “given notice of his right” by marking his articles pursuant to the marking statute, he is not entitled to damages before the date of actual notice. I'd; see also 35 U.S.C. § 287 (noting the patentee’s “failure so to mark’’ limits his damages to those incurred after actual notice). Section 287 is thus a limitation on damages, and not an affirmative, defense. Motorola, Inc. v. United States, 729 F.2d 765, 770 (Fed. Cir. 1984). Compliance with § 287 is a question of fact. Maxwell, 86 F.3d at 1111.

A patentee’s licensees must also comply with § -287, because the statute extends to “persons making or selling any patented article- for .or under [the patentee].” Id. (quoting § 287(a)). Recognizing that it may be difficult for a.patentee to ensure his licensees’ compliance with the marking provisions, we have held that where third parties are involved, courts may consider “whether the patentee made reasonable efforts to ensure compliance with the marking -requirements,” Id. at 1111-12, This “rule of reason” inquiry is “consistent with the purpose of the constructive notice provision—to encourage patentees to mark their products in order to provide notice to the public of the existence of the patent and to prevent innocent infringement,” Id. at 1112.

We have explained that the marking statute serves three related purposes: (1) helping to avoid innocent infringement; (2) encburaging patentees to give public notice that the article is patented; and (3) aiding the public to identify whether an article is patented. Nike, Inc. v. Wal-Mart Stores, Inc., 138 F.3d 1437, 1443 (Fed. Cir. 1998) (collecting authorities). Although patent infringement is a strict liability tort, a patentee who sells or permits the sale of unmarked, patented articles misleads others into believing they are free to make and sell an article actually covered by patent. Marking helps reduce innocent infringement by notifying the public that the article is patented. See Motorola, 729 F.2d at 772.

The parties dispute whether Arctic Cat’s licensee’s-failure to mark certain products limits Arctic Cat’s damages. In February -2002, Arctic Cat entered into a fully paid-up license agreement with Honda in which Honda paid $315,000 for licenses to two earlier-issued Arctic Cat patents and any later patents “that patently cover Arctic Cat’s Controlled Thrust Steering methods, systems and developments.” J.A. 3540 ¶ EE;, J.A. 7830-31 §§ 1.01, 3.01. The agreement includes the patents-in-suit. J.A, 3540 ¶ EE. The agreement specifically states Honda “shall have no obligation or requirement to mark” its licensed products. J.A. 7833 § 6.01. Honda sold PWCs in the United States through 2009 and Arctic Cat made .no effort to ensure Honda marked those PWCs, J.A. *13673540-41 ¶¶ II, JJ. At trial, the jury found damages began on October 16, 2008, before BRP received actual notice of infringement, J.A. 94.

There is no dispute that the patentee bears the burden of pleading and proving he complied with § 287(a). Maxwell, 86 F.3d at 1111. There is no dispute that Arctic Cat did not require Honda to mark; in fact, it expressly authorized Honda to sell licensed' products without marking. And it is likewise undisputed that Honda did not mark any of its PWCs with the patent numbers at issue. Thus, if Honda sold PWC products covered by the patents at issue, Arctic Cat has failed to satisfy the marking requirements. The only dispute between the parties is whether any of the Honda PWCs was covered by the patent claims at issue. BRP explains the issue on appeal: “The only area of dispute, between the parties was whether the PWCs that Honda sold were patented articles that were required to be marked. Which party bears the burden on this issue is a question of first impression for this Court and has split district courts.” Appellant’s Br. 37.

On summary judgment, the district court in this' case held that the burden of proving compliance with marking is placed on the defendant and that “the burden of production does not shift to the plaintiff to show compliance with a marking statute.” J.A. 58-59. And again in the denial of judgment as a matter of law, the district court repeated its belief that “BRP bears the burden of proving the defen[se] of marking.” J.A. 75. This was a legal error. The burden of proving compliance with marking is and at all times remains on the patentee. As in this case where BRP identified fourteen unmarked Honda PWCs, which it argued fell within the patent claims, it was the patentee’s burden to establish compliance with the marking statute—that these products did not fall within the patent claims.

There is a split among the district courts regarding which party must initially identify the products which it believes the paten-tee failed to mark. Some courts require the alleged infringer to initially identify products it believes practice the asserted patents. See, e.g., Fortinet, Inc. v. Sophos, Inc., No. 13-CV-05831-EMC, 2015 WL 5971585, at *4-5 (N.D. Cal. Oct. 14, 2015). These courts reason that “[ajbsent guidance from the other side as to which specific products are alleged to have been sold in contravention of the marking requirement, a patentee ... is left to guess exactly what it must prove up to establish compliance with the marking statute.” Sealant Sys. Int’l, Inc. v. TEK Glob. S.B.L., No. 5:11-CV-00774-PSG, 2014 WL 1008183, at *31 (N.D. Cal. Mar. 7, 2014), rev’d in part on other grounds, 616 Fed.Appx. 987 (Fed. Cir. 2015) (placing the initial burden on the alleged infringer to put the patentee “on notice” of unmarked products and finding it failed to meet its burden because of conflicting expert testimony and failure to produce admissible evidence showing a patented product was sold); Fortinet, 2015 WL 5971585, at *5 (adopting a “burden of production on [the alleged infringer] to identify the. [unmarked products] it believes practice the inventions claimed” and granting partial summary judgment in favor of the . alleged infringer where its expert report was “not too conclusory”); Unwired Planet, LLC v. Apple Inc., No. 13-CV-04134-VC, 2017 WL 1175379, at *5 (N.D. Cal. Feb. 14, 2017) (holding “[a]t most, the infringer bears some initial burden of plausibly identifying products subject to the marking requirement” and granting summary judgment in favor of the alleged infringer where it submitted a declaration and attached exhibits identifying particular products). This district court *1368agreed with that approach, concluding that if the defendant did not at least have the burden of identifying unmarked products it believed fell within the claims, “a defendant’s general allegations could easily instigate a fishing expedition for the paten-tee.” J.A. 59.

Other courts have required the patentee prove that none of its unmarked goods practice. the asserted claims. See, e.g., Adrea, LLC v. Barnes & Noble, Inc., No. 13-CV-4137 JSR, 2015 WL 4610465, at *1-2 (S.D.N.Y. July 24, 2015). Courts adopting this approach reason the patentee is in a better position to know whether his goods practice the patents-in-suit. Id. at *2 (citing Dunlap, 152 U.S. at 248, 14 S.Ct. 576); see also, e.g., DR Sys., Inc. v. Eastman Kodak Co:, No. 08-CV-0669H(BLM), 2009 WL 2632685, at *4 (S.D. Cal. Aug. 24, 2009) (“Just as a patentee’s compliance with the marking statute is a matter particularly within its knowledge, so are the details of its own product line.”); Kimberly-Clark Worldwide, Inc. v. First Quality Baby Prods., LLC, No. 1.-09-CV-1685, 2013 WL 1821593, at *3 (M.D. Pa. Apr. 30, 2013).

We hold an alleged infringer who challenges the patentee’s compliance with § 287 bears an initial burden of production to articulate the products it believes are unmarked “patented articles” subject to § 287. To be clear, this is a low bar. The alleged infringer need only put the patentee on notice that he or his authorized licensees sold specific unmarked products which the alleged infringer believes practice the patent. The alleged in-fringer’s burden is a burden of production, not one of persuasion or proof. Without some notice of what market products BRP believes required marking, Arctic Cat’s universe of products for which it would have to establish compliance would be unbounded. See Fortinet, 2015 WL 5971585, at *5 (‘Without some notice of what marketed products may practice the invention, AMI’s universe of products for which it would have to establish compliance with, or inapplicability of, the marking statute would be unbounded.” (quoting Sealant, 2014 WL 1008183, at *31)). Permitting in-fringers to allege failure to mark without identifying any products could lead to a large scale fishing expedition and gamesmanship. Once the alleged infringer meets its burden of production, however, the pat-entee bears the burden to prove the products identified do not practice the patented invention.

We do not here determine the minimum showing needed to meet the initial burden of production, but we hold in this case it was satisfied by BRP. At trial BRP introduced the licensing agreement between Honda and Arctic Cat showing Honda’s license to practice “Arctic Cat patents that patently cover Arctic Cat’s Controlled Thrust Steering methods, systems and developments.” J.A. 7830 § 1.01. BRP identified fourteen Honda PWCs from three versions- of its Aquatrax series sold between 2002 and 2009. J.A. 3540-41 ¶ II. BRP’s expert testified that he “review[ed] information regarding those models” and believed if BRP’s OTAS system practiced the patents, so did Honda’s throttle reapplication system in the Aquatrax PWCs. J.A. 2447-49; J.A. 2482. This was sufficient to satisfy BRP’s initial burden of production.

At summary judgment, the district court found BRP identified Honda PWCs and “presented an array of evidence” alleging they practice the asserted patents, but concluded BRP failed to meet its burden because it did not conduct a claim analysis of the products. J.A. 59-61. It later denied BRP’s motion for judgment as a matter of law because BRP “failed as a matter óf law to meet its burden in proving that Honda sold patent*1369ed articles.” J.A. 75 (emphasis added). The district court erred when it placed this burden on the alleged infringer. BRP shouldered only a burden of production to identify unmarked products that it alleges should have been marked. It was Arctic Cat’s burden to prove those products— once identified—do not practice the patent-at-issue. The alleged infringer need not produce claim charts to meet its initial burden pf identifying products. It is the patentee who bears the burden of proving that it satisfied the marking requirements and thus the patentee who would have to prove that the unmarked products identified by the infringer do not fall within the patent claims. The district court erred in placing this burden upon BRP and thus we vacate and remand on marking.

Because the district court adopted this legal approach at the summary judgment stage, it made clear to the parties that it would be BRP’s burden to prove that the unmarked products fell within the patent claims. Arctic Cat, therefore, did not haye a fair opportunity to develop its case regarding the Honda PWCs at trial. Because Arctic Cat was not on notice regarding its burden, and in fact labored under the assumption that BRP had the burden-of proof, reversal would be improper. We thus vacate the district court’s judgment as to marking and remand so that Arctic Cat has an opportunity to proffer evidence related to the identified Honda PWCs.2 Because we conclude BRP has met its initial burden of production, Arctic Cat must now establish the Honda PWCs do not practice the asserted patents to recover damages under the constructive notice provisions of § 287. -

III. Damages

BRP appeals the district court’s denial of judgment as a matter of law that the jury’s royalty award of $102.54 per infringing unit should be vacated based on inadmissible expert testimony. Prior to trial, the district court denied BRP’s Daubert motion to exclude this testimony of Arctic Cat’s expert, Walter Bratic, regarding the calculation of a reasonable royalty rate. J.A. 24-28. BRP also appeals the district court’s grant of an ongoing royalty of $205.08 per infringing unit. J.A. 137-44.

A. Reasonable Royalty Rate

The Eleventh Circuit reviews Daubert decisions for abuse of discretion. Quiet Tech. DC-8, Inc. v. Hurel-Dubois UK Ltd., 326 F.3d 1333, 1339-40 (11th Cir. 2003). “We review the jury’s determination of the amount of damages, an issue of fact, for substantial evidence.” Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1310 (Fed. Cir. 2009).

BRP raises the same arguments to exclude the testimony of Mr. Bratic rejected by the district court. BRP argues Mr. Bratic erroneously used BRP’s later-developed “Intelligent Brake and Reverse” (“iBR”) system as a value benchmark for BRP’s allegedly infringing and earlier-developed OTAS system. It argues that Mr. Bratic failed to establish that iBR is sufficiently comparable to the technology and value of OTAS, and thus his benchmark cannot serve as the basis for the jury’s royalty award. The' district court found that Mr. Bratic properly relied on the opinion of another Arctic Cat expert, Dr. Cuzzillo, who opined that OTAS and iBR are of comparable technological and safety value. J.A 24-26. The district court noted that Dr. Cuzzillo’s opinion was not vague *1370or conclusory but based on “his own investigation of the OTAS and iBR brake technologies, how they work, and the benefits provided as well as discussions with [another expert and review of his report].” J.A. 25-26. To the extent BRP found the comparison problematic, the district court suggested “that is a line of attack more appropriately addressed through cross-examination.” J.A. 26. BRP was given this chance at trial. E.g., j.Á, 1738-43.

. We agree with the district court’s analysis and conclude it did not abuse its discretion in admitting Mr. Bratic’s damages testimony. Cf. Apple Inc. v. Motorola, Inc,, 757 F.3d 1286, 1316-20 (Fed. Cir. 2014) (factually attacking the accuracy of a benchmark goes to evidentiary weight, not admissibility), overruled on other grounds by Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015) (en banc). Because BRP does not argue the royalty rate is not otherwise supported by substantial evidence,3 we affirm the district court’s denial of judgment as a matter of law. as to the jury’s reasonable royalty rate.

B. Ongoing Royalty Rate

We review for abuse of discretion a district court’s grant of an ongoing royalty. Whitserve, LLC v. Computer Packages, Inc., 694 F.3d 10, 35 (Fed. Cir. 2012). Ongoing royalties may be based on a post-judgment hypothetical negotiation using the Georgia-Pacific factors. Fresenius USA, Inc. v. Baxter Int’l, Inc., 582 F.3d 1288, 1303 (Fed. Cir. 2009) (citing Georgia-Pac. Corp. v. U.S. Plywood Corp., 318 F.Supp. 1116, 1120 (S.D.N.Y. 1970)).

The district court did not abuse its discretion in awarding an ongoing royalty rate. The district court weighed the relevant Georgia-Pacific factors and determined that Arctic Cat is entitled to an ongoing royalty amount higher than the jury rate. J.A. 137-42. While BRP argues the rate impermissibly covers its profits, we have affirmed rates at or near the infringer’s alleged profit margin. See, e.g., Golight, Inc. v. Wal-Mart Stores, Inc., 355 F.3d 1327, 1338-39 (Fed. Cir. 2004) (affirming district court’s grant of a reasonable royalty the defendant argued covered its profits); Bos. Sci. Corp. v. Cordis Corp., 497 Fed.Appx. 69 (Fed. Cir. 2013) (same for ongoing royalties); cf. Douglas Dynamics, LLC v. Buyers Prods. Co., 717 F.3d 1336, 1346 (Fed. Cir. 2013) (vacating and remanding a royalty award where the district court “clearly erred by limiting the ongoing royalty rate based on [the defendant’s] profit margins”). And we have explained that “[o]nce a judgment of validity and infringement has been entered ... the calculus is markedly different because different economic factors are involved.” Amado v. Microsoft Corp., 517 F.3d 1353, 1362 (Fed. Cir. 2008); see also Paice LLC v. Toyota Motor Corp., 504 F.3d 1293, 1317 (Fed. 'Cir. 2007) (Rader, J., concurring) (“[P]re-suit and post-judgment acts of infringement are distinct, and may warrant different royalty rates given the change in the parties’ legal relationship and other factors.”). We see no abuse .of discretion in the district court’s analysis and affirm its order awarding an ongoing royalty.

IV. Willfulness & Enhanced Damages

We review enhanced damages under 35 U.S.C. § 284 for abuse of discretion. Halo Elecs., Inc., v. Pulse Elecs., Inc., *1371— U.S. -, 136 S.Ct. 1923, 1934, 196 L.Ed.2d 278 (2016), A,party seeking enhanced damages under § 284 bears the burden of proof by, a preponderance of the evidence. Id. (citing Octane Fitness,. LLC v. ICON Health & Fitness Inc., -—: U.S. -, 134 S.Ct. 1749, 1758, 188 L.Ed.2d 816 (2014)). The Eleventh Circuit reviews de novo previously raised objections to jury instructions and gives district courts wide discretion in wording so long as the instructions accurately state the law. SEC v. Big Apple Consulting USA, Inc., 783 F.3d 786, 802 (11th Cir. 2015).

BRP appeals the district court’s denial of judgment as a matter' of law that it did not willfully infringe, the asserted patents because the jury’s willfulness finding is not supported by substantial evidence and the district court erred in instructing the jury. It also argues the district court abused its discretion by trebling damages.'

The jury’s willfulness finding is supported by substantial evidence. In denying BRP’’s motion for judgment as a matter of law on willfulness, the district court found substantial evidence demonstrated that BRP knew about the patents before they issued, conducted only a cursory analysis of the patents, waited years before seeking advice of qualified and competent counsel, and unsuccessfully tried to buy the asserted patents through a third party. J.A. 70-72. The district court denied BRP’s renewed motion for judgment as a matter of law on willfulness, stating it “will not second-guess the jury or substitute [the court’s] judgment for its judgment” where the verdict is supported by substantial evidence. JA. 124, Neither will we.

We reject BRP’s argument that the district, court’s jury instruction ,was erroneous. The district court instructed the jury that as to willful infringement, “Arctic Cat must prove by clear and convincing evidence that BRP actually knew or should have known that its actions constituted an unjustifiably high risk of infringement of a valid and enforceable patent.” ' J.A. 3037 (emphasis added). BRP argues this “should have known” standard contradicts Halo. Appellant’s Br. 61 (citing Halo, 136 S.Ct. at 1933). But this Court addressed this issue and concluded:

Halo did not disturb- the- substantive standard for the second prong of Sea-gate, subjective willfulness. Rather, Halo emphasized that subjective willfulness alone—i.e., proof that the defendant acted despite a risk of infringement that was “ ‘either known or so obvious that it should have been known to the accused infringer,’” Halo, 136 S.Ct. at 1930 (quoting [In re Seagate Technology, LLC], 497 F.3d. [1360] at 1371 [ (Fed. Cir. 2007)])—can support an award of enhanced damages.

WesternGeco L.L.C. v. ION Geophysical Corp., 837 F.3d 1358, 1362 (Fed. Cir. 2016). Thus, the court did not err in instructing the jury as BRP argues.

Finally, the district court did not abuse its discretion by trebling damages. While the district court initially trebled damages without much explanation, J.A. 97-98, it explained its decision in a subsequent thorough and well-reasoned opinion. See J.A. 99-116 (applying the factors outlined in Read Corp. v. Portec, Inc., 970 F.2d 816. (Fed. Cir. 1992)). Although the district court did not allow the parties to brief the issue, we will not adopt a blanket rule that a district court abuses its discretion by deciding an issue without receiving briefing from the parties. That is especially true where, as here, BRP attacks the district -court’s procedure but does not explain how additional briefing would have changed the outcome. In short, BRP has not shown- that the district *1372court’s failure to allow briefing amounts to an abuse of discretion. We affirm the district court’s denial of judgment as a matter of law on willfulness and its order trebling damages.

Conclusion

For the foregoing reasons, we affirm the district court’s denial of judgment as a matter of law that the asserted claims would have been obvious, that the jury-awarded royalty rate should be vacated, and that BRP did not willfully infringe the asserted claims. We also affirm the district court’s orders granting an ongoing royalty and trebling damages. We vacate the district court’s denial of judgment as a matter of law as to marking and remand for a new trial on this issue.

AFFIRMED IN PART, VACATED IN PART, AND REMANDED

Costs

No costs.

Amit P. Mehta, United States District Judge

In her now iconic song "Sorry," Beyoncé Knowles references the universally understood "middle finger"1 to express her hostility and feelings of betrayal over what *152listeners widely believe to be her husband's infidelity.2 Plaintiff Wesley Scott Ashton makes use of the middle finger for a different reason: He has placed a "pictogram" of an extended middle finger on a coffee mug, along with the phrase "People Pleaser in Recovery," to communicate a satirical message about overcoming obsequious behavior-quite literally an act of "mean mugging."3 Plaintiff seeks to copyright the combination of words and image, claiming it is a "highly creative poetic literary work." Defendant United States Copyright Office4 viewed Plaintiff's work differently. The agency determined that the combination of the words "People Pleaser in Recovery" and "Refill," with an image of a raised middle finger, is not sufficiently creative to constitute a copyrightable literary work.

Plaintiff filed this suit pro se to challenge Defendant's denial of his copyright application. He contends that Defendant failed to consider certain evidence he presented in his application and, furthermore, that Defendant's decision that his work is not sufficiently creative to warrant copyright protection is wrong as a matter of law. Both parties have moved for summary judgment. Finding no violation in Defendant's decision-making process or the substance of its decision, the court denies Plaintiff's Motion for Summary Judgment and grants Defendant's Cross-Motion for Summary Judgment.

I. BACKGROUND

A. Administrative Proceedings

In March 2012, Plaintiff Wesley Scott Ashton submitted an application to the United States Copyright Office ("Defendant") to register a work he titled "People Pleaser" ("the Work" or "mug"). Admin. R., ECF No. 19 [hereinafter AR], at 2. Plaintiff's Work is a hand-painted coffee mug that contains text and images. See AR at 3, 12-13. The exterior cylindrical surface of the mug reads "People Pleaser in Recovery" and is adorned with three butterflies, a yellow flower, and a lizard. See AR at 3-4, 7, 35. The bottom exterior of the mug contains an image of the back of a fist with a middle finger extended. See AR at 13. Below it is the hand-written text: "2008 (c) WSA." The word "Refill" is written on the bottom interior of the mug. See AR at 12. Images of the Work appear below.

*153On his application, Plaintiff indicated that he wanted two different copyrights: one for "2-Dimensional artwork" and another for "Text." See AR at 1. Defendant rejected Plaintiff's application on July 3, 2013, reasoning that the Work "lacks the authorship necessary to support a copyright claim." AR at 43-44. The decision stated that the text of Plaintiff's Work "will not support a claim to copyright under the [legal] standards" because it lacked the minimum amount of text needed for protection. AR at 43. Defendant noted that Plaintiff's "artwork on the cup is copyrightable," but because Plaintiff had "refused to have the claim in text" removed, Defendant was denying Plaintiff's application in full. AR at 43.

Plaintiff sought reconsideration of the denial, but this, too, was unsuccessful. In a letter dated September 20, 2013, Plaintiff asked Defendant to revisit its decision, contending that it was "flawed on its face" because it did not provide a rational explanation for its rejection of his application. AR at 47. Plaintiff seemingly sought both a copyright for the Work's text and a copyright for the artwork. See AR at 49 (arguing that his Work is "copyrightable as a combined composition of art and text."). In addition, Plaintiff argued that the "text" on the mug constituted a "literary work," as that term is defined in the Copyright Act. AR at 51; see also 17 U.S.C. § 101 (defining "literary work"). He identified the "text" of his Work as "three phrases": two "script phases"-"People Pleaser in Recovery"

*154and "Refill"-and a "pictogram of the raised middle finger." AR at 51-52. Plaintiff argued that "the three phrases together include the requisite minimum degree of creativity to merit copyright protection because even shorter, less expressive one line sentences" have received such protection. For support, he cited to Universal City Studios, Inc. v. Kamar Industries, Inc. , No. H-82-2377, 1982 WL 1278 (S.D. Tex. Sept. 20, 1982). AR at 52. And to buttress his position that a "raised middle finger" is copyrightable text, he cited and submitted a copy of a 82-page law review article, Ira P. Robbins, Digitus Impudicus: The Middle Finger and the Law , 41 U.C. Davis L. Rev. 1403 (2008), for the proposition that the gesture has "particularly vulgar connotations." AR at 52.

These arguments did not persuade Defendant, who denied Plaintiff's request for reconsideration in a letter dated January 9, 2014, which expanded on its initial decision. AR at 149-50. This second decision stated that Defendant had "carefully reviewed this work in light of the points raised in your letter" and acknowledged some specific points Plaintiff had raised in his reconsideration request, e.g. AR at 149 ("[Y]ou note that ..."), but concluded that "the simple combination and arrangement of the short phrase and word simply does not contain any authorship that would support a copyright registration in text" under the law. AR at 150. In reaching this decision, Defendant excluded from the "text" it considered for copyrightability the middle finger drawing; Defendant said that it considered that image to be "2D artwork, not text." AR at 150. At the end of the decision, Defendant invited Plaintiff to proceed with registering the artwork on his mug. AR at 150.

Plaintiff again sought reconsideration in a letter dated April 4, 2014. In this final request to the agency, Plaintiff asserted that the middle finger is "text," because it is a word, indicia, or symbol and, therefore, meets the definition of a "literary work" in the Copyright Act, 17 U.S.C. § 101, and possesses "a sufficient amount of originality ... to merit copyright protection as text." AR at 155-56. Again, he directed Defendant to the law review article Digitus Impudicus. AR at 155. Building on his argument that the middle finger drawing constituted "literary work," Plaintiff asserted that his Work's "text" contained "three phrases working together ... to express an idea in a poetic manner that possesses a sufficient amount of creativity to merit copyright protection as a text." AR at 157. Defendant explained that the "text" of his mug, "when property construed," constitutes a three-line poem:

The two script phrases are "People Pleaser in Recovery" located on the side wall of the work and "Refill" located inside the cup on its upper floor surface. The middle finger pictogram is arranged on the bottom of the cup so that it is flashed whenever the cup is tilted to drink from, for example. And, more specifically, the middle finger pictogram is juxtaposed with the "People Pleaser in Recovery" phrase so that the pictogram may be flashed while the audience is reading the "People Pleaser in Recovery" phrase. The word "Refill" located inside the mug is there to remind the reader that one must "refill" one's cup to recover from people pleasing.

AR at 157. These phrases, words, and images, when combined with their arrangement on the mug, Plaintiff asserted, constitute sufficient creativity to warrant copyright protection. AR at 157. Moreover, he once again argued that "even shorter, less expressive one line sentences have merited copyright protection," citing Universal City Studios , and also contended that important literary works, such as William Carlos Williams' sixteen-word poem "The Red Wheelbarrow," "may consist of a *155single short phrase." AR at 157-58. Plaintiff concluded by asking for both text and artwork copyrights. AR at 160.

Defendant denied, in part, this second request for reconsideration on September 27, 2016. AR at 209-13. The five-page decision ("Final Decision") stated that the work "is not copyrightable as a literary work," but that the two-dimensional artwork elements "remain copyrightable." AR at 209. The Final Decision focused on the requirements for a text copyright. After describing the work and the administrative history of Plaintiff's request, Defendant set forth the requirements for a copyright-namely, that the work must be "original." AR at 209-11. The "originality" requirement, Defendant explained, has two components: the work must "have been independently created by the author" and it "must possess sufficient creativity." AR at 210 (citing Feist Publ'ns, Inc. v. Rural Tel. Serv. Co. , 499 U.S. 340, 345, 111 S.Ct. 1282, 113 L.Ed.2d 358 (1991) ). It elaborated on these components by describing case law concerning the "sufficient creativity" requirement and citing the Copyright Office's regulations, 37 C.F.R. §§ 202.1(a), 202.10(a), containing the relevant standards. See AR at 210-12.

In a section titled "Analysis of the Work" that followed a summary of the governing law, Defendant examined Plaintiff's Work under the applicable rules. AR at 212-13. Defendant concluded that "the individual textual elements of the Work and the selection, coordination, and arrangement of those elements fail to satisfy the requirement of creative authorship necessary to sustain a claim to copyright in text." AR at 212.

In reaching this decision, Defendant made three findings. First, Defendant concluded that the Work's individual textual elements were not sufficiently creative because " '[w]ords and short phrases' and 'familiar symbols' are not copyrightable." AR at 212 (quoting 37 C.F.R. § 202.1(a) ). The accompanying discussion gave special attention to the middle finger, noting that, "[a]s for the pictogram that appears on the bottom of the mug, your argument that it constitutes a literary work, even if correct , does not persuade the [Defendant] Board that it is individually copyrightable." AR at 212 (emphasis added). Defendant continued by finding, "[t]o the degree the pictogram is considered an expressive symbol, and thus a literary work ... the content of the expression is certainly de minimis. " AR at 212. Next, Defendant addressed Plaintiff's argument that the three elements-the two text phrases and the middle finger drawing-as a whole formed a literary work that was eligible for a copyright. See AR at 212. It concluded that "even assuming these three elements could qualify as a literary work-which is doubtful-taken as a whole they fail to meet the creativity threshold set forth in [the Supreme Court's decision] in Feist for a work of authorship of any kind, literary or otherwise." AR at 212 (citing Feist Publ'ns , 499 U.S. at 359, 111 S.Ct. 1282 ). Defendant elaborated by noting, in response to Plaintiff's argument that his work embodied the requisite creativity because of the elements' arrangement, that "[t]here is ... nothing inherently original about placing text or symbols on a mug" and that "any effect that these elements are intended to create or actually do create in an observer is wholly irrelevant to the intrinsic creativity of the Work itself." AR at 213. Third, Defendant concluded that Plaintiff's mug's artwork elements-including the flower, the butterflies, and the middle finger-met the relevant creativity standards and thus could be registered, as it had before. AR at 213. This second denial of Plaintiff's request for reconsideration constituted final agency action. See 37 C.F.R. § 202.5.

*156B. Procedural History

Plaintiff filed this action on November 18, 2016, challenging the denial of his application for a text copyright under the Administrative Procedure Act ("APA"), 5 U.S.C. §§ 701 - 706. See Compl., ECF No. 1; see also 17 U.S.C. § 701(e) (providing that actions taken by the Register of Copyrights under the Copyright Act are subject to the provisions of the APA). Plaintiff argues that the decision-making process Defendant used in rejecting his application for a text copyright was flawed because: (1) Defendant failed to consider Plaintiff's evidence, and (2) its denial is contrary to law. See generally Compl. Because the court's review under the APA is limited to the administrative record, the parties conducted no discovery and filed cross-motions for summary judgment. See 5 U.S.C. § 706 (in reviewing an agency action under the APA, "the court shall review the whole record or those parts of it cited by a party"). Now before the court are Plaintiff's Motion for Summary Judgment, ECF No. 13, and Defendant's Cross-Motion for Summary Judgment, ECF No. 14.

II. LEGAL STANDARD

Under Federal Rule of Civil Procedure 56, summary judgment is appropriate when the moving party demonstrates that "there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law." Fed. R. Civ. P. 56(a). However, in a case involving review of a final agency action, the standard set forth in Rule 56 does not apply because of the limited role of the court in reviewing the administrative record. See AFL-CIO v. Chao , 496 F.Supp.2d 76, 81 (D.D.C. 2007). The court's role, in an APA action, "is to determine whether or not as a matter of law the evidence in the administrative record permitted the agency to make the decision it did." Charter Operators of Alaska v. Blank , 844 F.Supp.2d 122, 127 (D.D.C. 2012) (quoting Occidental Eng'g Co. v. INS , 753 F.2d 766, 769-70 (9th Cir. 1985) ). Summary judgment "serves as a mechanism for deciding, as a matter of law, whether the agency action is supported by the administrative record and is otherwise consistent with the APA standard of review." Id.

Under the APA, a court must set aside an agency action if it is "arbitrary, capricious, an abuse of the discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A). When deciding whether an agency's action violates this standard, the court "must consider whether the [agency's] decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment." Marsh v. Oregon Nat'l Res. Council , 490 U.S. 360, 378, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989) (internal quotation marks omitted). Agency action is arbitrary or capricious if, as is relevant in this case, the agency "entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise." Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co. , 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983). So long as the agency has examined the relevant data and identified a basis for its action, the court will uphold the agency's action. See Bowman Transp., Inc. v. Ark.-Best Freight Sys., Inc. , 419 U.S. 281, 286, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974). Judicial review in APA cases is "[h]ighly deferential" and "presumes the validity of agency action," permitting reversal only when "the agency's decision is not supported by substantial evidence, or if the agency has made a clear error in judgment."

*157Hagelin v. Fed. Election Comm'n , 411 F.3d 237, 242 (D.C. Cir. 2005) (quoting AT&T Corp. v. FCC , 220 F.3d 607, 616 (D.C. Cir. 2000) ).

III. DISCUSSION

A. Procedural Challenge

Plaintiff first contends that Defendant's decision was arbitrary and capricious because Defendant failed to consider various arguments he made and to evaluate evidence that he presented. Specifically, Plaintiff asserts that Defendant "made no intelligent account for the ruling as it provides no analysis of creativity, and provides no reasons for its decision, when it comes to the creative content of the poem as a whole." Pl.'s Mot. for Summ. J., ECF No. 13, Pl.'s Mem. in Supp. of Mot., ECF No. 13-1 [hereinafter Pl.'s Mem.], at 16; see also id. at 17 (arguing that Defendant "never considered" the "creative aspect" of the text of his mug, which he argues incorporates a "pictogram"); id. at 19 (arguing that Defendant "never proffer[ed] any explanation as to why the 3-D arrangement of the pictogram text sandwiched between English text in order to emphasize the pictogram is not substantially creative"). In addition, Plaintiff takes issue with the fact that Defendant did not expressly address various case citations and arguments he raised. Id. at 18. The court finds no error in the agency's decision-making process.

An agency is required to consider relevant data and articulate an explanation that bears "a rational connection between the facts found and the choice made." Bowen v. Am. Hosp. Ass'n , 476 U.S. 610, 626, 106 S.Ct. 2101, 90 L.Ed.2d 584 (1986) (internal quotation marks omitted). An agency action is arbitrary and capricious if the agency relied on inappropriate considerations, failed to consider "an important aspect of the problem," explained its decision in a way that "runs counter to the evidence before the agency," or is "so implausible that it [cannot] be ascribed to a difference in view or the product of agency expertise." State Farm Auto Ins. Co. , 463 U.S. at 43, 103 S.Ct. 2856. In an arbitrary and capricious challenge, the court will "uphold a decision of less than ideal clarity if the agency's path may reasonably be discerned." Id. (quoting Bowman Transp. , 419 U.S. at 286, 95 S.Ct. 438 ).

Defendant's Final Decision did not violate the process requirements of the APA. The court begins with Plaintiff's argument that Defendant provided neither "reasons" nor an "intelligent account" of its decision that the text of his mug lacked the requisite creativity. The reasons and intelligent account that Plaintiff claims are missing are clearly set forth in the section titled "Analysis of the Work" in the Final Decision. See AR at 212. There, Defendant stated that it "finds that the individual textual elements of the Work and the selection, coordination, and arrangement of those elements fail to satisfy the requirement of creative authorship to sustain a claim to copyright in text." Id. Further to this point, Defendant explicitly found that: (1) "none of [Plaintiff's] Work's constituent textual elements, considered individually, is sufficiently creative" to receive protection; and (2) that "even assuming" the two lines of text and "pictogram" formed a literary work, as Plaintiff had argued, the elements "taken as a whole ... fail to meet the creativity threshold set forth in Feist ." AR at 212-13. Put another way, Defendant concluded that the text of Plaintiff's Work-whether composed solely of the word and the phrase, or whether it also included the middle finger "pictogram"-was too short to merit copyright protection under the creativity standards for copyrightability. It was on this basis that Defendant denied Plaintiff a copyright. Thus, contrary to what Plaintiff alleges, *158Defendant unquestionably made an "intelligent account" for its decision.

To the extent Plaintiff is challenging Defendant's decision on grounds that it failed to address every component of his argument, Plaintiff's claim also falls flat. Plaintiff identifies a litany of items and arguments he cited in his correspondence in favor of his application, all of which he contends Defendant failed to consider, namely: (1) Universal City Studios , No. H-82-2377, 1982 WL 1278, Pl.'s Mem. at 17; (2) the law review article that discusses criminal penalties for the use of the middle finger, id. 16-17; (3) the "creative contribution of the pictogram to [Plaintiff's] literary work," id. at 22; and (4) William Carlos Williams's poem, "The Red Wheelbarrow," which Plaintiff submitted as an exhibit, id. at 18. But Plaintiff's insistence that Defendant's decision is arbitrary and capricious because Defendant did not expressly address these (and other) specific points has no support in the law. An agency is required to consider every "important aspect of the problem" and explain its decision in a way that is in line with the evidence. State Farm Auto Ins. Co. , 463 U.S. at 43, 103 S.Ct. 2856. This command implicitly sets limits on the agency's obligation: while the agency must weigh the evidence and provide a reasoned decision, it is under no obligation to consider all non-dispositive or tangential aspects of the issue before it. The court has no trouble concluding, based on its review of the administrative record, that even though the Final Decision does not specifically reference each of the above-listed points, Defendant satisfied its duty under the APA to provide a reasoned decision on every important aspect of the problem.

In any event, none of the items Plaintiff contends Defendant skipped over were material to the creativity inquiry that Defendant undertook to address Plaintiff's second request for reconsideration. A brief review of the items cited by Plaintiff in his Motion makes this plain. First, the case Plaintiff cited in his second reconsideration request, Universal City Studios , 1982 WL 1278, does not discuss whether the three- or four-word phrases at issue in that case were eligible for copyright protection, see id. at *4-5, and is further distinguished by the fact that, in that copyright infringement case, there was evidence and a presumption that the defendant in that case had copied the plaintiff's work, id. at *4. Next up is the 16-word poem "The Red Wheelbarrow," which Plaintiff cited for proposition that "important literary works may consist of a short phrase," AR 158. But if Plaintiff is suggesting that Defendant did not appreciate that proposition, he is clearly wrong. See, e.g. , AR at 211 (stating that words and short phrases are not eligible for protection, but acknowledging that some combinations of otherwise uncopyrightable material may be registered). Finally, the meaning and the offensive nature of the middle finger is well-known. Thus, the law review article submitted by Plaintiff presents no new information that Defendant should have considered in adjudicating this dispute. Moreover, as Defendant noted, the "espoused intentions of the author" and "any effect that [Plaintiff's Work's] elements are intended to create" are not relevant to the creativity inquiry. See AR at 211, 213. Finding no error in Defendant's decision-making process, the court turns to Plaintiff's substantive challenge to Defendant's Final Decision.

B. Substantive Challenge

The second part of Plaintiff's challenge concerns the merits of Defendant's decision. Plaintiff argues that his Work "is sufficiently creative to merit copyright registration." Pl.'s Mem. at 14. Defendant's decision to the contrary, he contends, is flawed because Defendant: (1) applied an *159inappropriate standard for determining "originality," id. at 15-16; (2) did not consider the "middle finger pictogram" to be a literary element, id. at 20-22; (3) wrongfully concluded the middle finger's creativity was "de minimis," id. at 22; and (4) inappropriately ignored the arrangement of the words on his Work, id. at 18-20. None of these arguments are well taken.

Only works that are "original" receive the protections of copyright. 17 U.S.C. § 102(a) ; Feist , 499 U.S. at 345, 111 S.Ct. 1282 (stating that originality is the "sine qua non" of copyright). The originality requirement has two components: (1) the work must have been independently created by the author; and (2) the work must possess a "minimal degree of creativity." Feist , 499 U.S. at 345, 111 S.Ct. 1282. The "requisite level" of creativity "is "extremely low; even a slight amount will suffice." Id. Though the bar is low, not every work that exhibits some degree of creativity is copyrightable. Regulations accompanying the Copyright Act reflect this rule. As is relevant here, the regulations provide that "[w]ords and short phrases such as names, titles, and slogans" and "familiar symbols or designs" are not eligible for a copyright. See 37 C.F.R. § 202.1(a). Courts regularly apply this rule. See, e.g., Boisson v. Banian, Ltd. , 273 F.3d 262, 269-71 (2d Cir. 2001) ; Ward v. Andrews McMeel Publ'g, LLC , 963 F.Supp.2d 222, 232-34 (S.D.N.Y. 2013) ; Prunte v. Universal Music Grp., Inc. , 699 F.Supp.2d 15, 25-26 (D.D.C. 2010).

The court quickly disposes of Plaintiff's first contention that Feist does not supply the relevant standard for "creativity," because that case "pertains to whether a compilation of data possesses sufficient creativity to merit copyright protection," whereas his work is a poem. Pl.'s Mem. at 16.5 Plaintiff is correct that Feist concerned a compilation of facts-residential entries in a telephone book-but that does not render its standard inapplicable here. Indeed, courts routinely apply the creativity standard articulated in Feist to a variety of works, not just factual compilations. See Southco, Inc. v. Kanebridge Corp. , 390 F.3d 276, 281-82, 285-86 (3d Cir. 2004) (citing Feist and applying the case's standard for originality to determine whether product part numbers were eligible for copyright protection); Urantia Found. v. Maaherra , 114 F.3d 955, 958-59 (9th Cir. 1997) (applying Feist to analyze whether a book of revelations allegedly received from non-human spiritual beings was eligible for copyright protection); Am. Dental Ass'n v. Delta Dental Plans Ass'n , 126 F.3d 977, 979 (7th Cir. 1997) (applying Feist to analyze the copyrightability of a taxonomy of dental procedures which included descriptions and a numbering system); CMM Cable Rep, Inc. v. Ocean Coast Properties, Inc. , 97 F.3d 1504, 1516 (1st Cir. 1996) (applying Feist and its originality standard when examining the copyrightability of a phrase in commercial promotional materials in an enforcement action). Without question, the creativity standard articulated in Feist is applicable here.

Next up is Plaintiff's claim that Defendant's refusal to treat the "pictogram" of the middle finger as a "literary work" constitutes error. That argument, however, misconstrues Defendant's Final Decision. The Final Decision actually assumes that *160the pictogram is a literary work, but nevertheless rejects its copyright-worthiness because it and the rest of the Work's text, considered together, is insufficiently creative. As Defendant wrote: "To the degree the pictogram is considered an expressive symbol, and thus a literary work, ... the content of the expression is certainly de minimis. " AR at 212. The court therefore need not consider the legal question whether the pictogram of the middle finger meets the statutory definition of "literary work." See 17 U.S.C. § 101 (" 'Literary works' are works, other than audiovisual works, expressed in words, numbers, or other verbal or numerical symbols or indicia....").

Relatedly, Plaintiff argues that Defendant misapplied the law when it found that the three elements of the Work lacked sufficient creativity. Pl.'s Mem. at 18-22. But under the deferential review the court must give Defendant in this APA action, see OddzOn Prods., Inc. v. Oman , 924 F.2d 346, 348 (D.C. Cir. 1991), the court cannot conclude that Defendant's decision was arbitrary and capricious. It has long been the position of the Copyright Office that some works are too short to receive the protections of copyright. See Kitchens of Sara Lee, Inc. v. Nifty Foods Corp. , 266 F.2d 541, 544 (2d Cir. 1959) ("The Copyright Office does not regard as sufficient to warrant copyright registration 'familiar symbols or designs, mere variations of typographic ornamentation, lettering or coloring, and mere listings of ingredients or contents.' .... [This statement] is a fair summary of the law."). Indeed, courts consistently have applied that rule to short words or phrases. See, e.g., Prunte , 699 F.Supp.2d at 22-23, 26-30 (applying the originality standard articulated in Feist and concluding that various phrases in plaintiff's songs, including "I'm a maniac" and "that's what's up," are unprotectable); Peters v. West , 776 F.Supp.2d 742, 749-50 (N.D. Ill. 2011) (applying the originality standard articulated in Feist in a copyright infringement action against rapper Kanye West, and concluding that the well-known phrase "[t]hat which does not kill me makes me stronger" is not protectable). In view of this well-settled law, and according weight to the Copyright Office's application of its own regulations in this case, the court finds no "clear error of judgment" in Defendant's decision that combined elements of Plaintiff's Work are not sufficiently creative to receive copyright protection. Marsh , 490 U.S. at 378, 109 S.Ct. 1851.

Finally, as to Plaintiff's contention that Defendant afforded no weight to the "arrangement" of the components on the mug, once more, he is wrong. Defendant properly recognized that "a combination of unprotectable elements may be eligible for copyright protection if the elements are 'selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work.' " AR at 213 (quoting 17 U.S.C. § 101 ). Applying that rule, Defendant ruled that "[t]here is, of course, nothing inherently original about placing text or symbols on a mug, and the simplistic relations of the elements to one another in the Work does not transform this arrangement into something copyrightable." Id. The court finds nothing arbitrary about that conclusion.

IV. CONCLUSION

Accordingly, Plaintiff's Motion for Summary Judgment, ECF No. 13, is denied, and Defendant's Cross-Motion for Summary Judgment, ECF No. 14, is granted.6

*161A separate Order accompanies this Memorandum Opinion.

NEWMAN, Circuit Judge:

This appeal from the grant of a preliminary injunction concerns primarily the availability of copyright protection for the visual images electronically displayed by a coin-operated video game of the sort currently enjoying widespread popularity throughout the country. Omni Video Games, Inc., its distributor, and two of its officers appeal from an order entered May 22, 1981 in the District Court for the Eastern District of New York (Eugene H. Nick-erson, Judge),v preliminarily enjoining them from infringing the copyright of Stern Electronics, Inc. in the audiovisual work entitled “Scramble” and from making further use of the trademark “SCRAMBLE” in connection with electronic video games. 523 F.Supp. 635. Appellants contend that the visual images and accompanying sounds of the video game fail to satisfy the fixation and originality requirements of the Copyright Act, 17 U.S.C.App. § 102(a) (1976), and that they, rather than appellees, have superior rights to the mark “SCRAMBLE”. We reject these contentions and affirm the preliminary injunction.

Video games like “Scramble” can roughly be described as computers programmed to create on a television screen cartoons in which some of the action is controlled by the player. In Stern’s “Scramble,” for example, the video screen displays a spaceship moving horizontally through six different scenes in which obstacles are encountered. With each scene the player faces increasing difficulty in traversing the course and scoring points. The first scene depicts mountainous terrain, missile bases, and fuel depots. The player controls the altitude and speed of the spaceship, decides when to release the ship’s supply of bombs, and fires lasers that can destroy attacking missiles and aircraft. He attempts to bomb the missile bases (scoring points for success), bomb the fuel depots (increasing his own diminishing fuel supply with each hit), *854avoid the missiles being fired from the ground, and avoid crashing his ship into the mountains. And that is only scene one. In subsequent scenes the hazards include missile-firing enemy aircraft and tunnel-like airspaces. The scenes are in color, and the action is accompanied by battlefield sounds.

The game is built into a cabinet containing a cathode ray tube, a number of electronic circuit boards, a loudspeaker, and hand controls for the player. The electronic circuitry includes memory storage devices called PROMs, an acronym for “programmable read only memory.”1 The PROM stores the instructions and data from a computer program in such a way that when electric current passes through the circuitry, the interaction of the program stored in the PROM with the other components of the game produces the sights and sounds of the audiovisual display that the player sees and hears. The memory devices determine not only the appearance and movement of the images but also the variations in movement in response to the player’s operation of the hand controls.

Stern manufactures amusement equipment, including video games, for distribution worldwide. In January 1981 at a London trade exhibit Stern became aware of “Scramble,” an electronic video game developed in late 1980 by a Japanese corporation, Konami Industry Co., Ltd. The audiovisual display constituting what Stern alleges is the copyrightable work was first published in Japan on January 8,1981. Stern secured an exclusive sub-license to distribute the “Scramble” game in North and South America from Konami’s exclusive licensee, and began selling the game in the United States on March 17, 1981. Even in the fast-paced world of video games, “Scramble” quickly became a big success. Approximately 10,000 units were sold at about $2,000 each in the first two months for an initial sales volume of about $20 million.

On April 14, 1981, a Certificate of Copyright Registration for the audiovisual work “Scramble” was issued to Konami by the United States Copyright Office, and shortly thereafter documents were filed with the Copyright Office reflecting the license and sub-license to Stern. To satisfy the statutory requirement for deposit of copies of a work to be copyrighted, 17 U.S.C.App. § 408(b) (1976), Konami submitted video tape recordings of the “Scramble” game, both in its “attract mode” and in its “play mode.” 2

Omni alleges that, concurrently with Stern’s sales of the “Scramble” game and even earlier, it was endeavoring to sell a line of video game products so constructed that each unit could be equipped for playing different games by substituting a PROM containing the program for a particular game. Omni contends that it planned to market this line of interchangeable games with the label “Scramble” affixed to the headboard of each unit; the name of the particular game was also to be prominently displayed. On December 1, 1980, Omni’s president ordered ten silk screen name plates bearing the name “Scramble.” Between that date and March 17, 1981, the date of Stern’s first sale of its “Scramble” game, Omni sold five units of video games bearing the name “Scramble” on the headboard. In April 1981 Omni began to sell a video game called “Scramble” that not only bears the same name as the “Scramble” game Stern was then marketing, but also is *855virtually identical in both sight and sound. It sold this copy of Stern’s “Scramble” game, known in the trade as a “knock-off,” for several hundred dollars less than Stern’s game.

1. Copyright Issues

In challenging the preliminary injunction that bars distribution of its “Scramble” game, Omni does not dispute that Konami and its sub-licensee Stern are entitled to secure some copyright protection for their “Scramble” game. Omni contends that Ko-nami was entitled to copyright only the written computer program that determines the sights and sounds of the game’s audiovisual display.3 While that approach would have afforded some degree of protection, it would not have prevented a determined competitor from manufacturing a “knockoff” of “Scramble” that replicates precisely the sights and sounds of the game’s audiovisual display. This could be done by writing a new computer program that would interact with the hardware components of a video game to produce on the screen the same images seen in “Scramble,” accompanied by the same sounds. Such replication is possible because many different computer programs can produce the same “results,” whether those results are an analysis of financial records or a sequence of images and sounds. A program is simply “a set of statements [/.e., data] or instructions to be used directly or indirectly in a computer in order to bring about a certain result,” Pub. L.No. 96-517, § 10(a), 94 Stat. 3015, 3028 (1980) (amending 17 U.S.C.App. § 101 (1976)). To take an elementary example, the result of displaying a “4” can be achieved by an instruction to add 2 and 2, subtract 3 from 7, or in a variety of other ways. Obviously, writing a new program to replicate the play of “Scramble” requires a sophisticated effort, but it is a manageable task.

To secure protection against the risk of a “knock-off” of “Scramble” based upon an original program, Konami eschewed registration of its program as a literary work and chose instead to register the sights and sounds of “Scramble” as an audiovisual work. See 17 U.S.C.App. § 102(a)(6) (1976). The Act defines “audiovisual works” as “works that consist of a series of related images which are intrinsically intended to be shown by the use of machines, or devices such as projectors, viewers, or electronic equipment, together with accompanying sounds, if any, regardless of the nature of the material objects, such as films or tapes, in which the works are embodied.” 17 U.S.C.App. § 101 (1976). Omni contends that Konami is not entitled to secure a copyright in the sights and sounds of its “Scramble” game because the audiovisual work is neither “fixed in any tangible medium of expression” nor “original” within the meaning of § 102(a). Both contentions arise from the fact that the sequence of some of the images appearing on the screen during each play of the game will vary depending upon the actions taken by the player. For example, if he fails to avoid enemy fire, his spaceship will be destroyed; if he fails to destroy enough fuel depots, his own fuel supply will run out, and his spaceship will crash; if he succeeds in destroying missile sites and enemy planes, those images will disappear from the screen; and the precise course travelled by his spaceship will depend upon his adjustment of the craft’s altitude and velocity.

If the content of the audiovisual display were not affected by the participation of the player, there would be no doubt that the display itself, and not merely the written computer program, would be eligible for copyright. The display satisfies the statutory definition of an original “audiovisual work,” and the memory devices of the game satisfy the statutory requirement of a “copy” in which the work is “fixed.”4 The *856Act defines “copies” as “material objects ... in which a work is fixed by any method now known or later developed, and from which the work can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device” and specifies that a work is “fixed” when “its embodiment in a copy ... is sufficiently permanent or stable to permit it to be perceived, reproduced, or otherwise communicated for a period of more than transitory duration.” 17 U.S.C.App. § 101 (1976). The audiovisual work is permanently embodied in a material object, the memory devices, from which it can be perceived with the aid of the other components of the game.

We agree with the District Court that the player’s participation does not withdraw the audiovisual work from copyright eligibility. No doubt the entire sequence of all the sights and sounds of the game are different each time the game is played, depending upon the route and speed the player selects for his spaceship and the timing and accuracy of his release of his craft’s bombs and lasers. Nevertheless, many aspects of the sights and the sequence of their appearance remain constant during each play of the game. These include the appearance (shape, color, and size) of the player’s spaceship, the enemy craft, the ground missile bases and fuel depots, and the terrain over which (and beneath which) the player’s ship flies, as well as the sequence in which the missile bases, fuel depots, and terrain appears. Also constant are the sounds heard whenever the player successfully destroys an enemy craft or installation or fails to avoid an enemy missile or laser. It is true, as appellants contend, that some of these sights and sounds will not be seen and heard during each play of the game in the event that the player’s spaceship is destroyed before the entire course is traversed. But the images remain fixed, capable of being seen and heard each time a player succeeds in keeping his spaceship aloft long enough to permit the appearances of all the images and sounds of a complete play of the game. The repetitive sequence of a substantial portion of the sights and sounds of the game qualifies for copyright protection as an audiovisual work.

Appellants’ claim that the work lacks originality proceeds along two lines. Repeating their attack on fixation, they assert that each play of the game is an original work because of the player’s participation. The videotape of a particular play of the game, they assert, secured protection only for that one “original” display. However, the repeated appearance of the same sequence of numerous sights and sounds in each play of the game defeats this branch of the argument. Attacking from the opposite flank, appellants contend that the audiovisual display contains no originality because all of its reappearing features are determined by the previously created computer program. This argument is also without merit. The visual and aural features of the audiovisual display are plainly original variations sufficient to render the display copyrightable even though the underlying written program has an independent existence and is itself eligible for copyright. Nor is copyright defeated because the audiovisual work and the computer program are both embodied in the same components of the game. The same thing occurs when an audio tape embodies both a musical composition and a sound recording. Moreover, the argument overlooks the sequence of the creative process. Someone first conceived what the audiovisual display would look like and sound like. Originality occurred at that point. Then the program was written. Finally, the program was imprinted into the memory devices so that, in operation with the components of the game, *857the sights and sounds could be seen and heard. The resulting display satisfies the requirement of an original work.

We need not decide at what point the repeating sequence of images would form too insubstantial a portion of an entire display to warrant a copyright, nor the somewhat related issue of whether a sequence of images (e.g., a spaceship shooting down an attacking plane) might contain so little in the way of particularized form of expression as to be only an abstract idea portrayed in noncopyrightable form, see Nichols v. Universal Pictures Corp., 45 F.2d 119, 121 (2d Cir. 1930), cert. denied, 282 U.S. 902, 51 S.Ct. 216, 75 L.Ed. 795 (1931). Assessing the entire effect of the game as it appears and sounds, we conclude that its repetitive sequence of images is copyrightable as an audiovisual display. See Atari, Inc. v. Amusement World, Inc., No. 81-803 (D.Md. Nov. 27, 1981); Midway Manufacturing Co. v. Drikschneider, No. 81-0-243 (D.Neb. July 15, 1981); Williams Electronics, Inc. v. Artic International, Inc., Civ. No. 81-1852 (D.N.J. June 24, 1981); Cinematronics, Inc. v. K. Noma Enterprise Co., Civ. No. 81-439 (D.Ariz. May 22, 1981).

2. Trademark Issue

Appellants contend that they, rather than appellee, have superior common law rights in the mark “SCRAMBLE” based on their use of the mark in the United States in early 1981, prior to appellants’ use. The District Court found that appellants’ prior use was not bona fide and that appellee’s use entitled it to injunctive relief against appellants’ use. We agree.

Though there is no direct evidence that Omni learned of the “SCRAMBLE” mark because of the planned use of the mark by Konami and Stern, that inference was available to the District Court, at least for purposes of considering Stern’s entitlement to a preliminary injunction. Omni appears to have attempted only a preemptive use of the mark by ordering just ten name plates with the mark and affixing just five of them to headboards of units of separately named games sold in early 1981. Cf. La Societe Anonyme Des Parfum LeGalion v. Jean Patou, Inc., 495 F.2d 1265 (2d Cir. 1974) . The District Court was entitled to conclude that Omni was simply attaching a secondary label in a bad faith attempt to reserve a mark. Blue Bell, Inc. v. Farah Manufacturing Co., 508 F.2d 1260 (5th Cir. 1975); Blue Bell, Inc. v. Jaymar-Ruby, Inc., 497 F.2d 433 (2d Cir. 1974). Omni’s claim that it sought to establish a trade name for its line of separately named, interchangeable games is undermined by the fact that once it secured a “knock-off” of Stern’s “Scramble” game, it affixed only the mark “SCRAMBLE”, thereby designating the name of the game, and not the line of which it was a part. As Judge Nickerson observed, “[I]t would be a truly remarkable coincidence if defendants independently thought of the name ‘Scramble’ and then, only a few months later, produced a video game virtually identical to the one bearing the same name. It is more likely that defendants sought to appropriate the trademark with the expectation that they would later imitate the audiovisual display.”

Moreover, the equities abundantly justified issuance of an injunction against Omni’s use of the mark. Stern has a substantial investment in the mark, having achieved success in the marketplace with its sales of a large number of units bearing the mark. By contrast, Omni has placed the mark on the headboard of five units of games that are not “Scramble” and has used the mark for a “Scramble” game that is a pirated “knock-off” of Stern’s game.

The preliminary injunction is affirmed.

SYKES, Circuit Judge.

Chapman Kelley is a nationally recognized artist known for his representational paintings of landscapes and flowers — in particular, romantic floral and woodland interpretations set within ellipses. In 1984 he received permission from the Chicago Park District to install an ambitious wildflower display at the north end of Grant Park, a prominent public space in the heart of downtown Chicago. “Wildflower Works” was thereafter planted: two enormous elliptical flower beds, each nearly as big as a football field, featuring a variety of native wildflowers and edged with borders of gravel and steel.

Promoted as “living art,” Wildflower Works received critical and popular acclaim, and for a while Kelley and a group of volunteers tended the vast garden, pruning and replanting as needed. But by 2004 Wildflower Works had deteriorated, and the City’s goals for Grant Park had changed. So the Park District dramatically modified the garden, substantially reducing its size, reconfiguring the oval flower beds into rectangles, and changing some of the planting material.

Kelley sued the Park District for violating his “right of integrity” under the Visual Artists Rights Act of 1990 (“VARA”), 17 U.S.C. § 106A, and also for breach of contract. The contract claim is insubstantial; the main event here is the VARA claim, which is novel and tests the boundaries of copyright law. Congress enacted this statute to comply with the nation’s obligations under the Berne Convention for the Protection of Literary and Artistic Works. VARA amended the Copyright Act, importing a limited version of the civil-law concept of the “moral rights of the artist” into our intellectual-property law. In brief, for certain types of visual art — paintings, drawings, prints, sculptures, and exhibition photographs — VARA confers upon the artist certain rights of attribution and integrity. The latter include the right of the artist to prevent, during his lifetime, *292any distortion or modification of his work that would be “prejudicial to his ... honor or reputation,” and to recover for any such intentional distortion or modification undertaken without his consent. See 17 U.S.C. § 106A(a)(3)(A).

The district court held a bench trial and entered a split judgment. The court rejected Kelley’s moral-rights claim for two reasons. First, the judge held that although Wildflower Works could be classified as both a painting and a sculpture and therefore a work of visual art under VARA, it lacked sufficient originality to be eligible for copyright, a foundational requirement in the statute. Second, following the First Circuit’s decision in Phillips v. Pembroke Real Estate, Inc., 459 F.3d 128 (1st Cir.2006), the court concluded that site-specific art like Wildflower Works is categorically excluded from protection under VARA. The court then held for Kelley on the contract claim, but found his evidence of damages uncertain and entered a nominal award of $1. Both sides appealed.

We affirm in part and reverse in part. There is reason to doubt several of the district court’s conclusions: that Wildflower Works is a painting or sculpture; that it flunks the test for originality; and that all site-specific art is excluded from VARA. But the court was right to reject this claim; for reasons relating to copyright’s requirements of expressive authorship and fixation, a living garden like Wildflower Works is not copyrightable. The district court’s treatment of the contract claim is another matter; the Park District is entitled to judgment on that claim as well.

I. Background

Kelley is a painter noted for his use of bold, elliptical outlines to surround scenes of landscapes and flowers. In the late-1970s and 1980s, he moved from the canvas to the soil and created a series of large outdoor wildflower displays that resembled his paintings. He planted the first in 1976 alongside a runway at the Dallas-Fort Worth International Airport and the second in 1982 outside the Dallas Museum of Natural History. The wildflower exhibit at the museum was temporary; the one at the airport just “gradually petered out.”

In 1983 Kelley accepted an invitation from Chicago-based oil executive John Swearingen and his wife, Bonnie — collectors of Kelley’s paintings — to come to Chicago to explore the possibility of creating a large outdoor wildflower display in the area. He scouted sites by land and by air and eventually settled on Grant Park, the city’s showcase public space running along Lake Michigan in the center of downtown Chicago. This location suited Kelley’s artistic, environmental, and educational mission; it also provided the best opportunity to reach a large audience. Kelley met with the Park District superintendent to present his proposal, and on June 19,1984, the Park District Board of Commissioners granted him a permit to install a “permanent Wild Flower Floral Display” on a grassy area on top of the underground Monroe Street parking garage in Daley Bicentennial Plaza in Grant Park. Under the terms of the permit, Kelley was to install and maintain the exhibit at his own expense. The Park District reserved the right to terminate the installation by giving Kelley “a 90 day notice to remove the planting.”

Kelley named the project “Chicago Wildflower Works I.” The Park District issued a press release announcing that “a new form of ‘living’ art” was coming to Grant Park — “giant ovals of multicolored wildflowers” created by Kelley, a painter and “pioneer in the use of natural materials” who “attracted national prominence for his efforts to incorporate the landscape *293in artistic creation.” The announcement explained that “[o]nce the ovals mature, the results mil be two breathtaking natural canvases of Kelley-designed color patterns.”

In the late summer of 1984, Kelley began installing the two large-scale elliptical flower beds at the Grant Park site; they spanned 1.5 acres of parkland and were set within gravel and steel borders. A gravel walkway bisected one of the ovals, and each flower bed also accommodated several large, preexisting air vents that were flush with the planting surface, providing ventilation to the parking garage below. For planting material Kelley selected between 48 and 60 species of self-sustaining wildflowers native to the region. The species were selected for various aesthetic, environmental, and cultural reasons, but also to increase the likelihood that the garden could withstand Chicago’s harsh winters and survive with minimal maintenance. Kelley designed the initial placement of the wildflowers so they would blossom sequentially, changing colors throughout the growing' season and increasing in brightness towards the center of each ellipse. He purchased the initial planting material — between 200,000 and 300,000 wildflower plugs — at a cost of between $80,000 and $152,000. In September of 1984, a battery of volunteers planted the seedlings under Kelley’s direction.

When the wildflowers bloomed the following year, Wildflower Works was greeted with widespread acclaim. Chicago’s mayor, the Illinois Senate, and the Illinois Chapter of the American Society of Landscape Artists issued commendations. People flocked to see the lovely display— marketed by the Park District as “living landscape art” — and admiring articles appeared in national newspapers. Wildflower Works was a hit. Here’s a picture:

*294For the next several years, Kelley’s permit was renewed and he and his volunteers tended the impressive garden. They pruned and weeded and regularly planted new seeds, both to experiment with the garden’s composition and to fill in where initial specimen had not flourished. Of course, the forces of nature — the varying bloom periods of the plants; their spread habits, compatibility, and life cycles; and the weather — produced constant change. Some wildflowers naturally did better than others. Some spread aggressively and encroached on neighboring plants. Some withered and died. Unwanted plants sprung up from seeds brought in by birds and the wind. Insects, rabbits, and weeds settled in, eventually taking a toll. Four years after Wildflower Works was planted, the Park District decided to discontinue the exhibit. On June 3, 1988, the District gave Kelley a 90-day notice of termination.

Kelley responded by suing the Park District in federal court, claiming the termination of his permit violated the First Amendment. The parties quickly settled; in exchange for dismissal of the suit, the Park District agreed to extend Kelley’s permit for another year. On September 14, 1988, the Park District issued a “Temporary Permit” to Kelley and Chicago Wildflower Works, Inc., a nonprofit organization formed by his volunteers. This permit authorized them “to operate and maintain a two ellipse Wildflowers Garden Display ... at Daley Bicentennial Plaza in Grant Park” until September 1, 1989. The permit stipulated that Kelley “will have responsibility and control over matters relating to the aesthetic design and content of Wildflower Works I,” and Wildflower Works, Inc. “shall maintain the Wildflower Works I at no cost to the Chicago Park District including, without limitation, weeding and application of fertilizer.” Although it did not contain a notice-of-termination provision, the permit did state that “[t]he planting material is the property of Mr. Chapman Kelley” and that Kelley “may remove the planting material” if the permit was not extended. Finally, the permit provided that “[t]his agreement does not create any proprietary interest for Chicago Wildflower Works, Inc., or Mr. Chapman Kelley in continuing to operate and maintain the Wildflower Garden Display after September 1,1989.”

The Park District formally extended this permit each succeeding year through 1994. After that point Kelley and his volunteers continued to cultivate Wildflower Works without a permit, and the Park District took no action, adverse or otherwise, regarding the garden’s future. In March 2004 Kelley and Jonathan Dedmon, president of Wildflower Works, Inc., attended a luncheon to discuss the 20th anniversary of Wildflower Works. At the luncheon Dedmon asked Park District Commissioner Margaret Burroughs if Wildflower Works needed a new permit. Commissioner Burroughs responded, ‘You’re still there, aren’t you? That’s all you need to do.”

Three months later, on June 10, 2004, Park District officials met with Kelley and Dedmon to discuss problems relating to inadequate maintenance of the garden and forthcoming changes to Grant Park necessitated by the construction of the adjacent Millennium Park. The officials proposed reconfiguring Wildflower Works — decreasing its size from approximately 66,000 square feet to just under 30,000 square feet and remaking its elliptical flower beds into rectangles. The District’s director of development invited Kelley’s views on this proposal but made it clear that the District planned to go forward with the reconfiguration with or without Kelley’s approval. Kelley objected to the proposed changes, but did not request an opportunity to remove his planting material before the reconfiguration took place. A week later the *295Park District proceeded with its plan and reduced Wildflower Works to less than half its original size. The elliptical borders became rectilinear, weeds were removed, surviving wildflowers were replanted in the smaller-scale garden, and some new planting material was added. Dedmon sent a letter of protest to the Park District.

Kelley then sued the Park District for violating his moral rights under VARA. He claimed that Wildflower Works was both a painting and a sculpture and therefore a “work of visual art” under VARA, and that the Park District’s reconfiguration of it was an intentional “distortion, mutilation, or other modification” of his work and was “prejudicial to his ... honor or reputation.” See 17 U.S.C. § 106A(a)(3)(A). He also alleged breach of contract; he claimed that Commissioner Burroughs’s remark created an implied contract that the Park District had breached when it altered Wildflower Works without providing reasonable notice.1 On the VARA claim Kelley sought compensation for the moral-rights violation, statutory damages, and attorney’s fees; on the contract claim he sought the fair-market value of the planting material removed in the reconfiguration. He later quantified his damages, estimating the value of the plants at $1.5 million and requesting a staggering $25 million for the VARA violation.

The case proceeded to a bench trial, and the district court entered judgment for the Park District on the VARA claim and for Kelley on the contract claim. See Kelley v. Chi. Park Dist., No. 04 C 07715, 2008 WL 4449886 (N.D.Ill. Sept. 29, 2008). The judge first concluded that Wildflower Works could be classified as both a painting and a sculpture and therefore qualified as' a work of visual art under VARA. Id. at *4-5. But he also held that Wildflower Works was insufficiently original for copyright, a prerequisite to moral-rights protection under VARA. Id. at *6. Alternatively, the judge concluded that Wildflower Works was site-specific art, and following the First Circuit’s decision in Phillips, held that VARA did not apply to this category of art. Id. at *6-7. On the contract claim the court construed the Chicago Park District Act, 70 III. Comp. Stat. 1505/7.01, to permit individual commissioners to enter into binding contracts on the Park District’s behalf. Id. at *7-8. The judge found that Commissioner Burroughs’s statement — “You’re still there, aren’t you? That’s all you need to do.”— created an implied contract that the Park District had breached by failing to give Kelley reasonable notice before altering Wildflower Works. Id. But the judge also concluded that Kelley had failed to prove damages to a reasonable certainty and awarded $1 in nominal damages. Id. at *9.

Kelley appealed, challenging the adverse judgment on the VARA claim and the district court’s treatment of the damages issue on the contract claim. The Park District cross-appealed from the judgment on the contract claim.

II. Discussion

This case comes to us from a judgment entered after a bench trial; we review the district court’s factual findings for clear error and its conclusions of law de novo. Spurgin-Dienst v. United States, 359 F.3d 451, 453 (7th Cir.2004). In this circuit, questions of copyright eligi*296bility are issues of law subject to independent review. Schrock v. Learning Curve Int’l, Inc., 586 F.3d 513, 517 (7th Cir.2009).

A. Kelley’s Moral-Rights Claim Under the Visual Artists Rights Act of 1990

1. A brief history of moral rights

That artists have certain “moral rights” in their work is a doctrine long recognized in civil-law countries but only recently imported into the United States. Moral rights are generally grouped into two categories: rights of attribution and rights of integrity. “Rights of attribution” generally include the artist’s right to be recognized as the author of his work, to publish anonymously and pseudonymously, to prevent attribution of his name to works he did not create, and to prevent his work from being attributed to other artists. Carter v. Helmsley-Spear, Inc., 71 F.3d 77, 81 (2d Cir.1995) (citing Ralph E. Lerner & Judith Bresler, Art Law 419-20 (1989)). “Rights of integrity” include the artist’s right to prevent the modification, mutilation, or distortion of his work, and in some cases (if the work is of recognized stature), to prevent its destruction. Id. at 81-82 (citing Art Law at 420-21).

Originating in nineteenth-century France, moral rights — le droit moral2— are understood as rights inhering in the artist’s personality, transcending property and contract rights and existing independently of the artist’s economic interest in his work. See 3 Melville B. Nimmer & David Nimmer, Nimmer on Copyright § 8D.01[A] (2010); 5 William F. Patry, Patry on Copyright §§ 16:1, 16:3 (2010); John Henry Merryman, The Refrigerator of Bernard Buffet, 27 Hastings L.J. 1023, 1023-28 (1976). American copyright law, on the other hand, protects the economic interests of artists; Article I of the Constitution authorizes Congress “To Promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” U.S. Const, art. 1, § 8, cl. 8. Unlike other intellectual-property rights, moral rights are unrelated to the artist’s pecuniary interests and are grounded in philosophical ideas about the intrinsic nature and cultural value of art rather than natural-property or utility justifications.8 See Carter, 71 F.3d at 81 (describing moral rights as “rights of a spiritual, non-economic and personal nature [that] ... spring from a belief that an artist in the process of creation injects his spirit into the work and that the artist’s personality, as well as the integrity of the work, should therefore be protected and preserved”). VARA introduced a limited version of this European doctrine into American law, but it is not an easy fit.3 4

*297VARA was enacted as a consequence of the United States’ accession to the Berne Convention for the Protection of Literary and Artistic Works. After many years of resistance, the Senate ratified the treaty in 1988, bringing the United States into the Berne Union effective the following year. See 4 Nimmer § 17.01[C][2] (2010); 5 Pa-try §§ 16:1, 16:3. The Berne Convention dates to 1886, when seven European nations (plus Haiti and Tunisia) joined together to extend copyright protection across their borders. See 4 Nimmer § 17.01[B][1] nn. 10 & 17 (2002). During the course of the next century, many other nations joined, and the treaty underwent periodic revisions, most notably for our purposes in 1928 when Article 6bis was added, incorporating the concept of moral rights. See 3 id. § 8D.01[B] (2004); 5 Patry §§ 16:1, 16:3. Article 6bis provides:

(1) Independently of the author’s economic rights, and even after the transfer of the said rights, the author shall have the right to claim authorship of the work and to object to any distortion, mutilation or other modification of, or other derogatory action in relation to, the said work, which would be prejudicial to his honor or reputation.

(3) The means of redress for safeguarding the rights granted by this Article shall be governed by the legislation of the country where protection is claimed.

Berne Convention for the Protection of Literary and Artistic Works, art. 6bis, Sept. 9, 1886, as revised at Paris on July 24, 1971, S. Treaty Doc. No. 99-27 (1986).

When the United States joined the Berne Union in 1989, the concept of moral rights was largely unknown in American law. See Lee v. A.R.T. Co., 125 F.3d 580, 582 (7th Cir.1997) (“[I]t was accepted wisdom [before VARA] that the United States did not enforce any claim of moral rights.”); see also Weinstein v. Univ. of Ill., 811 F.2d 1091, 1095 n. 3 (7th Cir.1987) (The Continental principle of le droit moral is a doctrine that “no American jurisdiction follows as a general matter.”); Merryman, The Refrigerator of Bernard Buffet, 27 Hastings L.J. at 1035-36 (“The moral right of the artist, and in particular that component called the right of integrity of the work of art, simply does not exist in our law.”). Article 6bis was a major obstacle to Berne ratification. See Martin v. City of Indianapolis, 192 F.3d 608, 611 (7th Cir.1999) (The treaty’s moral-rights concept “was controversial in this country” and was embraced post-Berne only “in a very limited way.”); Carter, 71 F.3d at 82-83 (“The issue of federal protection of moral rights was a prominent hurdle in the debate over whether the United States should join the Berne Convention....”); see also 3 Nimmer § 8D.02[A]-[D] (2004); 5 Patry §§ 16:1, 16:3; Roberta Rosenthal Kwall, How Fine Art Fares Post VARA, 1 Marq. Intell. Prop. L.Rev. 1, 1-4 (1997).

American unease with European moral-rights doctrine — more particularly, the obligations imposed by Article 6bis — persisted beyond Berne ratification. Indeed, Congress initially took the position that domestic law already captured the concept in existing copyright and common-law doctrines and in the statutory law of some states. See Berne Convention Implementation Act of 1988, Pub.L. No. 100-568, §§ 2(2), (3), 102 Stat. 2853; 3 Nimmer § 8D.02[D][1] (2009); 5 Patry § 16:3. This was seen as an implausible claim. See 3 Nimmer § 8D.02[D][1] (“Th[e] Congressional finding flies in the *298face of numerous judicial and scholarly-pronouncements.5 Patry § 16:3 (The American position that existing federal and state laws satisfied minimum Berne obligations created “a web of fictional compliance.”). “[A] question of international credibility existed,” and “some Berne co-Unionists ... expressed doubts regarding the accuracy or sincerity of the U.S. declaration that its law already afforded a degree of moral rights protection equivalent to Berne standards.” Jane C. Ginsburg, Copyright in the 101st Congress: Commentary on the Visual Artists Rights Act and the Architectural Works Copyright Protection Act of 1990, 14 Colum.-VLA J.L. & Arts 477, 478-79 (1990). VARA was enacted to fill this perceived gap, but its moral-rights protection is quite a bit narrower than its European counterpart.

2. VARA’s scope

VARA amended the Copyright Act and provides a measure of protection for a limited set of moral rights falling under the rubric of “rights of attribution” and “rights of integrity” — but only for artists who create specific types of visual art. 17 U.S.C. § 106A(a). The statutory coverage is limited to paintings, drawings, prints, sculptures, and photographs created for exhibition existing in a single copy or a limited edition of 200 or less. See id. § 101 (defining “work of visual art”). The rights conferred by the statute exist independently of property rights; the artist retains them even after he no longer holds title to his work. Id. § 106A(a).

More specifically, VARA’s attribution and integrity rights are as follows:

(a) Rights of attribution and integrity. Subject to section 107 and independent of the exclusive rights provided in section 106, the author of a work of visual art—

(1) shall have the right—

(A) to claim authorship of that work, and

(B) to prevent the use of his or her name as the author of any work of visual art which he or she did not create;

(2) shall have the right to prevent the use of his or her name as the author of the work of visual art in the event of distortion, mutilation, or other modification of the work which would be prejudicial to his or her honor or reputation; and

(3) subject to the limitations set forth in section 113(d), shall have the right—

(A) to prevent any intentional distortion, mutilation, or other modification of that work which would be prejudicial to his or her honor or reputation, and any intentional distortion, mutilation, or modification of that work is a violation of that right, and

(B) to prevent any destruction of a work of recognized stature, and any intentional or grossly negligent destruction of that work is a violation of that right.

17 U.S.C. § 106A(a) (emphasis added). At issue here is the right of integrity conferred by subsection (a)(3)(A), which precludes any intentional modification or distortion of a work of visual art that “would be prejudicial to [the artist’s] honor or reputation.”

A qualifying “work of visual art” is defined as:

(1) a painting, drawing, print, or sculpture, existing in a single copy, in a limited edition of 200 or fewer that are signed and consecutively numbered by the author, or, in the case of a sculpture, in multiple cast, carved, or fabricated sculptures of 200 or fewer that are con*299secutively numbered by the author and bear the signature or other identifying mark of the author; or

(2) a still photographic image produced for exhibition purposes only, existing in a single copy that is signed by the author, or in a limited edition of 200 copies or fewer that are signed and consecutively number by the author.

Id. § 101 (emphasis added). This definition also contains a number of specific exclusions: e.g., posters, maps, and globes; books, newspapers, magazines, and other periodicals; “motion picture[s] or other audiovisual work[s]”; merchandising and promotional materials; “any work made for hire”; and “any work not subject to copyright protection under this title.” Id.

This last exclusion simply reinforces the point that YARA supplements general copyright protection; to qualify for moral rights under VARA, a work must first satisfy basic copyright standards. Under the Copyright Act of 1976, copyright subsists in “original works of authorship fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced, or otherwise communicated.” Id. § 102(a). “Works of authorship” include “pictorial, graphic, and sculptural works.” Id. § 102(a)(5). VARA’s definition of “work of visual art” is limited to a narrow subset of this broader universe of “pictorial, graphic, and sculptural works” that are otherwise eligible for copyright; only a select few categories of art get the extra protection provided by the moral-rights concept. 5 Patry § 16:7 (2010) (“Protected ‘works of visual art’ is a narrower subcategory of ‘pictorial, graphic, and sculptural works,’ protected in section 102(a)(5).”).

Several exceptions limit the scope of the rights granted under the statute:

(c) Exceptions. (1) The modification of a work of visual art which is a result of the passage of time or the inherent nature of the materials is not a distortion, mutilation, or other modification described in subsection (a)(3)(A).

(2) The modification of a work of visual art which is the result of conservation, or of the public presentation, including lighting and placement, of the work is not a destruction, distortion, mutilation, or other modification described in subsection (a)(3) unless the modification is caused by gross negligence.

17 U.S.C. § 106A(c) (emphasis added). The second of these — the “public presentation” exception — is at issue here. Another exception invoked by the Park District is found in a different section of the Copyright Act that defines the scope of a copyright owner’s rights:

(d)(1) In a case in which—

(A) a work of visual art has been incorporated in or made part of a building in such a way that removing the work from the building will cause the destruction, distortion, mutilation, or other modification of the work as described in section 106A(a)(3), and

(B) the author consented to the installation of the work in the building either before the effective date set forth in section 610(a) of the Visual Artists Rights Act of 1990, or in a written instrument executed on or after such effective date that is signed by the owner of the building and the author and that specifies that installation of the work may subject the work to destruction, distortion, mutilation, or other modification, by reason of its removal,

then the rights conferred by paragraphs (2) and (3) of section 106A(a) shall not apply. *300 Id. § 113 (emphasis added). This is known as the “building exception.”

VARA rights cannot be transferred or assigned, but they can be waived in a writing signed by the artist and “specifically identifying] the work, and uses of that work, to which the waiver applies.” Id. § 106A(e)(l). Absent a written waiver, the artist retains VARA rights during his lifetime even if he transfers ownership of the work or assigns his copyright.5 Id. § 106A(d)(l), (e)(2).

3. Is Wildflower Works a painting or sculpture?

The district court held that Wildflower Works was both a painting and a sculpture but was insufficiently original to qualify for copyright. Alternatively, the court concluded that it was site-specific art and held that all site-specific art is implicitly excluded from VARA. Other arguments — in particular, whether Wildflower Works satisfies additional threshold requirements for copyright and whether VARA’s public-presentation or building exceptions applied — were not reached.

On appeal Kelley contests the district court’s conclusions regarding originality and site-specific art. The Park District defends these holdings and also reiterates the other arguments it made in the district court, except one: The Park District has not challenged the district court’s conclusion that Wildflower Works is a painting and a sculpture.

This is an astonishing omission. VARA’s definition of “work of visual art” operates to narrow and focus the statute’s coverage; only a “painting, drawing, print, or sculpture,” or an exhibition photograph will qualify. These terms are not further defined, but the overall structure of the statutory scheme clearly illuminates the limiting effect of this definition. Copyright’s broad general coverage extends to “original works of authorship,” and this includes “pictorial, graphic, and sculptural works.” 17 U.S.C. § 102(a)(5). The use of the adjectives “pictorial” and “sculptural” suggests flexibility and breadth in application. In contrast VARA uses the specific nouns “painting” and “sculpture.” To qualify for moral-rights protection under VARA, Wildflower Works cannot just be “pictorial” or “sculptural” in some aspect or effect, it must actually be a “painting” or a “sculpture.” Not metaphorically or by analogy, but really.

That Kelley considered the garden to be both a painting and a sculpture — only rendered in living material — is not dispositive. He also characterized it as an experiment in environmental theory, telling a reporter he was trying to “figure out the economic and ecological impact of introducing wildflowers into cities.” In promoting Wildflower Works, Kelley variously described the project as a “living wildflower painting,” a “study on wildflower landscape and management,” and “a new vegetative management system that beautifies [the] landscape economically with low-maintenance wildflowers.”

Kelley’s expert, a professor of art history, reinforced his view that Wildflower Works was both a painting and a sculpture, but the district court largely disregarded her testimony as unhelpful.6 Kel *301 ley, 2008 WL 4449886, at *5. For its part the Park District initially marketed Wildflower Works as “living art,” but this adds little to the analysis. VARA plainly uses the terms “painting” and “sculpture” as words of limitation. Even assuming a generous stance on what qualifies, see 5 Patry § 16:7 (suggesting a “liberal attitude toward what may be considered a painting, drawing, print, or sculpture”), the terms cannot be read coextensively with the broader categories of “pictorial” and “sculptural” works that are generally eligible for copyright under § 102(a)(5). If a living garden like Wildflower Works really counts as both a painting and a sculpture, then these terms do no limiting work at all.

The district judge worried about taking “too literalist an approach to determining whether a given object qualifies as a sculpture or painting.” Kelley, 2008 WL 4449886, at *4. His concern was the “tension between the law and the evolution of ideas in modern or avant garden art; the former requires legislatures to taxonomize artistic creations, whereas the latter is occupied with expanding the definition of what we accept to be art.” Id. We agree with .this important insight. But there’s a big difference between avoiding a literalistic approach and embracing one that is infinitely malleable. The judge appears to have come down too close to the latter extreme.7

*302In short, this case raises serious questions about the meaning and application of VARA’s definition of qualifying works of visual art — questions with potentially decisive consequences for this and other moral-rights claims. But the Park District has not challenged this aspect of the district court’s decision, so we move directly to the question of copyrightability, which is actually where the analysis should start in the first place.

4. Is Wildñower Works copyrightable?

To merit copyright protection, Wildflower Works must be an “original work[ ] of authorship fixed in a[ ] tangible medium of expression ... from which [it] can be perceived, reproduced, or otherwise communicated.” 17 U.S.C. § 102(a). The district court held that although Wildflower Works was both a painting and a sculpture, it was ineligible for copyright because it lacked originality. There is a contradiction here. As we have explained, VARA supplements general copyright protection and applies only to artists who create the specific subcategories of art enumerated in the statute. VARA-eligible paintings and sculptures comprise a discrete subset of otherwise copyrightable pictorial and sculptural works; the statute designates these works of fine art as worthy of special protection. If a work is so lacking in originality that it cannot satisfy the basic requirements for copyright, then it can hardly qualify as a painting or sculpture eligible for extra protection under VARA. See Cronin, Dead on the Vine, 12 Vand. J. Ent. & Tech. L. at 239 (“[I]f a work does not evince sufficient original expression to be copyrightable, the work should belong in a category other than ‘visual art’ as this term is contemplated under VARA.”).

That point aside, the district court’s conclusion misunderstands the originality requirement. Originality is “the touchstone of copyright protection today,” an implicit constitutional and explicit statutory requirement. Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 347, 346, 111 S.Ct. 1282, 113 L.Ed.2d 358 (1991) (“Originality is a constitutional requirement.”); id. at 355 (The Copyright Act of 1976 made the originality requirement explicit.); see also Schrock, 586 F.3d at 518-19 (“As a constitutional and statutory matter, ‘[t]he sine qua non of copyright is originality.’ ” (quoting Feist, 499 U.S. at 345, 111 S.Ct. 1282)). Despite its centrality in our copyright regime, the threshold for originality is minimal. See Feist, 499 U.S. at 345, 111 S.Ct. 1282; Am. Dental Ass’n v. Delta Dental Plans Ass’n, 126 F.3d 977, 979 (7th Cir.1997) (“The necessary degree of ‘originality’ is low....”). The standard requires “only that the work was independently created by the author (as opposed to copied from other works), and that it possesses at least some minimal degree of creativity.” Feist, 499 U.S. at 345, 111 S.Ct. 1282 (citation omitted). The “requisite level of creativity is extremely low; even a slight amount will suffice. The vast majority of works make the grade quite easily, as they possess some creative spark.” Id. (citation omitted).

The district court took the position that Wildflower Works was not original because Kelley was not “the first person to ever conceive of and express an arrangement of growing wildflowers in ellipse-shaped enclosed area[s].” Kelley, 2008 WL 4449886, at *6. This mistakenly equates originality with novelty; the law is clear that a work can be original even if it is not novel. Feist, 499 U.S. at 345, 111 S.Ct. 1282 (“Originality does not signify *303novelty; a work may be original even though it closely resembles other works so long as the similarity is fortuitous, not the result of copying.”). No one argues that Wildflower Works was copied; it plainly possesses more than a little creative spark.

The judge was also at a loss to discover “what about the exhibit is original. Is it the elliptical design? The size? The use of native instead of non-native plants? The environmentally-sustainable gardening method to which ‘vegetative management system’ apparently refers?” Kelley, 2008 WL 4449886, at *6." It is true that common geometric shapes cannot be copyrighted. See U.S. Copyright Office, Compendium II: Copyright Office Practices § 503.02(a)-(b) (1984); 2 Patry § 4:17 (2010). And “[i]n no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such a work.” 17 U.S.C. § 102(b).

The Park District suggests that Wildflower Works is an uncopyrightable “method” or “system,” and is also ineligible because its design uses simple elliptical shapes. The first of these arguments is not well-developed; the second is misplaced. Although Wildflower Works was designed to be largely self-sustaining (at least initially), it’s not really a “method” or “system” at all. It’s a garden. And Kelley is seeking statutory protection for the garden itself, not any supposed “system” of vegetative management encompassed within it. Regarding the use of elliptical shapes, an author’s expressive combination or arrangement of otherwise noncopyrightable elements (like geometric shapes) may satisfy the originality requirement. Roulo v. Russ Berrie & Co., 886 F.2d 931, 939 (7th Cir.1989); 2 Patry § 4:17 (Geometric shapes or symbols cannot themselves be protected, but an original creative arrangement of them can be.).

The real impediment to copyright here is not that Wildflower Works fails the test for originality (understood as “not copied” and “possessing some creativity”) but that a living garden lacks the kind of authorship and stable fixation normally required to support copyright. Unlike originality, authorship and fixation are explicit constitutional requirements; the Copyright Clause empowers Congress to secure for “authors” exclusive rights in their “writings.” U.S. Const, art 1, § 8, cl. 8; see also 2 Patry § 3:20 (2010) (“[T]he Constitution uses the terms ‘writings’ and ‘authors;’ ‘originality’ is not used.”); id. § 3:22 (2010); 1 Nimmer § 2.03[A]-[B] (2004). The originality requirement is implicit in these express limitations on the congressional copyright power. See Feist, 499 U.S. at 346, 111 S.Ct. 1282 (The constitutional reference to “authors” and “writings” “presuppose[s] a degree of originality.”). The Supreme Court has “repeatedly construed all three terms in relation to one another [or] perhaps has collapsed them into a single concept”; therefore, “[w]ritings are what authors create, but for one to be an author, the writing has to be original.” 2 Patry § 3:20.

‘Without fixation,” moreover, “there cannot be a ‘writing.’” Id. § 3:22. The Nimmer treatise elaborates:

Fixation in tangible form is not merely a statutory condition to copyright. It is also a constitutional necessity. That is, unless a work is reduced to tangible form it cannot be regarded as a “writing” within the meaning of the constitutional clause authorizing federal copyright legislation. Thus, certain works of conceptual art stand outside of copyright protection.

*3041 Nimmer § 2.03[B]. A work is “fixed” in a tangible medium of expression “when its embodiment in a copy or phonorecord ... is sufficiently permanent or stable to permit it to be perceived, reproduced, or otherwise communicated for a period of more than transitory duration.” 17 U.S.C. § 101. As William Patry explains:

Fixation serves two basic roles: (1) easing problems of proof of creation and infringement, and (2) providing the dividing line between state common law protection and protection under the federal Copyright Act, since works that are not fixed are ineligible for federal protection but may be protected under state law. The distinction between the intangible intellectual property (the work of authorship) and its fixation in a tangible medium of expression (the copy) is an old and fundamental and important one. The distinction may be understood by examples of multiple fixations of the same work: A musical composition may be embodied in sheet music, on an audio-tape, on a compact disc, on a computer hard drive or server, or as part of a motion picture soundtrack. In each of the fixations, the intangible property remains a musical composition.

2 Patry § 3:22 (internal quotation marks omitted).

Finally, “authorship is an entirely human endeavor.” Id. § 3:19 (2010). Authors of copyrightable works must be human; works owing their form to the forces of nature cannot be copyrighted. Id. § 3:19 n. 1; see also U.S. Copyright Office, Compendium II: Copyright Office Practices § 503.03(a) (“[A] work must be the product of human authorship” and not the forces of nature.) (1984); id. § 202.02(b).

Recognizing copyright in Wildflower Works presses too hard on these basic principles. We fully accept that the artistic community might classify Kelley’s garden as a work of postmodern conceptual art. We acknowledge as well that copyright’s prerequisites of authorship and fixation are broadly defined. But the law must have some limits; not all conceptual art may be copyrighted. In the ordinary copyright ease, authorship and fixation are not contested; most works presented for copyright are unambiguously authored and unambiguously fixed. But this is not an ordinary case. A living garden like Wildflower Works is neither “authored” nor “fixed” in the senses required for copyright. See Toney v. L’Oreal USA, Inc., 406 F.3d 905, 910 (7th Cir.2005) (“A person’s likeness — her persona — is not authored and it is not fixed.”); see also Cronin, Dead on the Vine, 12 Vand. J. Ent. & Tech. L. at 227-39.

Simply put, gardens are planted and cultivated, not authored. A garden’s constituent elements are alive and inherently changeable, not fixed. Most of what we see and experience in a garden — the colors, shapes, textures, and scents of the plants — originates in nature, not in the mind of the gardener. At any given moment in time, a garden owes most of its form and appearance to natural forces, though the gardener who plants and tends it obviously assists. All this is true of Wildflower Works, even though it was designed and planted by an artist.

Of course, a human “author” — whether an artist, a professional landscape designer, or an amateur backyard gardener— determines the initial arrangement of the plants in a garden. This is not the kind of authorship required for copyright. To the extent that seeds or seedlings can be considered a “medium of expression,” they originate in nature, and natural forces— not the intellect of the gardener — determine their form, growth, and appearance. Moreover, a garden is simply too changea*305ble to satisfy the primary purpose of fixation; its appearance is too inherently variable to supply a baseline for determining questions of copyright creation and infringement. If a garden can qualify as a “work of authorship” sufficiently “embodied in a copy,” at what point has fixation occurred? When the garden is newly planted? When its first blossoms appear? When it is in full bloom? How — and at what point in time — is a court to determine whether infringing copying has occurred?

In contrast, when a landscape designer conceives of a plan for a garden and puts it in -writing — records it in text, diagrams, or drawings on paper or on a digital-storage device — we can say that his intangible intellectual property has been embodied in a fixed and tangible “copy.” This writing is a sufficiently permanent and stable copy of the designer’s intellectual expression and is vulnerable to infringing copying, giving rise to the designer’s right to claim copyright. The same cannot be said of a garden, which is not a fixed copy of the gardener’s intellectual property. Although the planting material is tangible and can be perceived for more than a transitory duration, it is not stable or permanent enough to be called “fixed.” Seeds and plants in a garden are naturally in a state of perpetual change; they germinate, grow, bloom, become dormant, and eventually die. This life cycle moves gradually, over days, weeks, and season to season, but the real barrier to copyright here is not temporal but essential. The essence of a garden is its vitality, not its fixedness. It may endure from season to season, but its nature is one of dynamic change.

We are not suggesting that copyright attaches only to works that are static or fully permanent (no medium of expression lasts forever), or that artists who incorporate natural or living elements in their work can never claim copyright. Kelley compares Wildflower Works to the Crown Fountain, a sculpture by Spanish artist Jaume Piensa that sits nearby in Chicago’s Millennium Park. The surfaces of Plensa’s fountain are embedded with LED screens that replay recorded video images of the faces of 1,000 Chicagoans. See http://www.explorechicago.org/city/en/ things_see_do/attractions/dca_tourism/ Crown_Fountain.html (last visited Feb. 10, 2011). But the Copyright Act specifically contemplates works that incorporate or consist of sounds or images that are broadcast or transmitted electronically, such as telecasts of sporting events or other live performances, video games, and the like. See 17 U.S.C. § 101 (defining “fixed” as including a “work consisting of sounds, images, or both, that are being transmitted ... if a fixation of the work is being made simultaneously with its transmission”); see also Balt. Orioles, Inc. v. Major League Baseball Players Ass’n, 805 F.2d 663, 675 (7th Cir.1986); Midway Mfg. Co. v. Artic Int'l, Inc., 704 F.2d 1009, 1013-14 (7th Cir.1983). Wildflower Works does not fit in this category; the Crown Fountain is not analogous.

Though not addressing the requirement of fixation directly, the district court compared Wildflower Works to “[t]he mobiles of Alexander Calder” and “Jeff Koons’ ‘Puppy,’ a 43-foot flowering topiary.” Kelley, 2008 WL 4449886, at *4. These analogies are also inapt. Although the aesthetic effect of a Calder mobile is attributable in part to its subtle movement in response to air currents, see http://en.wikipedia.org/ wiki/Alexander_Calder (last visited Feb. 10, 2011), the mobile itself is obviously fixed and stable. In “Puppy” the artist assembled a huge metal frame in the shape of a puppy and covered it with thousands of blooming flowers sustained by an irrigation system within the frame. See http://en.wikipedia.org/wiki/Jeff_Koons (last visited Feb. 10, 2011). This may be *306sufficient fixation for copyright (we venture no opinion on the question), but Wildflower Works is quite different. It is quintessentially a garden; “Puppy” is not.

In short, Wildflower Works presents serious problems of authorship and fixation that these and other examples of conceptual or kinetic art do not. Because Kelley’s garden is neither “authored” nor “fixed” in the senses required for basic copyright, it cannot qualify for moral-rights protection under VARA.

5. Site-specifíc art, and the public-presentation and building exceptions

This case also raises some important questions about the application of VARA to site-specific art, as well as the statute’s public-presentation and building exceptions. Though we need not decide these questions, we do have a few words of caution about the district court’s treatment of the issue of VARA and site-specific art. The court classified Wildflower Works as a form of site-specific art; we see no reason to upset this factual finding. The court then adopted the First Circuit’s holding in Phillips that site-specific art is categorically excluded from VARA. This legal conclusion is open to question.

Phillips involved a VARA claim brought by artist David Phillips in a dispute over a display of 27 of his sculptures in Boston’s Eastport Park across from Boston Harbor. Phillips v. Pembroke Real Estate, Inc., 459 F.3d 128, 130 (1st Cir.2006). A planned redesign of the park called for the removal and relocation of Phillips’s sculptures; he sought an injunction under VARA, claiming the removal of his sculptures would violate his right of integrity. Id. at 131. The district court held that although the sculptures qualified as a single integrated work of visual art, park administrators were entitled to remove them under VARA’s public-presentation exception. Id. at 138-39. The First Circuit affirmed on alternative grounds, holding that VARA does not apply to any site-specific art.

The court based this holding on a perceived irreconcilable tension between the public-presentation exception and the purpose of site-specific art: “By definition, site-specific art integrates its location as one of its elements. Therefore, the removal of a site-specific work from its location necessarily destroys that work of art.” Id. at 140. Under the public-presentation exception, a modification of a work of visual art stemming from a change in its “public presentation, including lighting or placement,” is not actionable unless it is caused by gross negligence. If VARA applied to site-specific art, the First Circuit reasoned, then the statute would “purport! ] to protect site-specific art” but also “permit its destruction by the application” of the public-presentation exception. Id. The court held that “VARA does not protect site-specific art and then permit its destruction by removal from its site pursuant to the statute’s public presentation exception. VARA does not apply to site-specific art at all.” Id. at 143.

There are a couple of reasons to question this interpretation of VARA. First, the term “site-specific art” appears nowhere in the statute. Nothing in the definition of a “work of visual art” either explicitly or by implication excludes this form of art from moral-rights protection. Nor does application of the public-presentation exception operate to eliminate every type of protection VARA grants to creators of site-specific art; the exception simply narrows the scope of the statute’s protection for all qualifying works of visual art. The exception basically provides a safe harbor for ordinary changes in the public presentation of VARA-qualifying artworks; the artist has no cause of action *307unless through gross negligence the work is modified, distorted, or destroyed in the process of changing its public presentation.

Second, Phillips’s all-or-nothing approach to site-specific art may be unwarranted. Site-specific art is not necessarily destroyed if moved; modified, yes, but not always utterly destroyed. Moreover, some of YARA’s protections are unaffected by the public-presentation exception. An artist’s right of integrity can be violated in ways that do not implicate the work’s location or manner of public presentation; site-specific art — like any other type of art — can be defaced and damaged in ways that do not relate to its public display. And the public-presentation exception does nothing to limit the right of attribution, which prevents an artist’s name from being misappropriated.

Then there is the matter of the building exception, which applies to works “incorporated in or made part of a building in such a way that removing the work from the building will cause the destruction, distortion, mutilation, or other modification of the work.” 17 U.S.C. § 113(d)(1)(A). These works do not get moral-rights protection if the artist: (1) consented to the installation of his work in the building (if pre-VARA); or (2) executed a written acknowledgment that removal of the work may subject it to destruction, distortion, mutilation, or modification (if post-VARA). Id. § 113(d)(1)(B). On its face this exception covers a particular kind of site-specific art. Its presence in the statute suggests that site-specific art is not categorically excluded from VARA.8

These observations are of course general and not dispositive. Because we are resolving the VARA claim on other grounds, we need not decide whether VARA is inapplicable to site-specific art.

B. The Park District’s Cross-Appeal on the Contract Claim

The Park District challenges the judgment against it for breach of contract even though damages were assessed at a nominal $1. The district court held that Commissioner Burroughs’s casual remark- — -“You’re still there, aren’t you? That’s all you need to do.” — created an implied-in-fact contract requiring the Park District to give Kelley reasonable notice before reconfiguring Wildflower Works. Although factual findings about the existence of a contract are reviewed for clear error, ReMapp Int’l Corp. v. Comfort Keyboard Co., 560 F.3d 628, 633 (7th Cir.2009), there is a' threshold legal question here about the commissioner’s unilateral authority to bind the Park District to a contract. Our review is de novo. See Manning v. United States, 546 F.3d 430, 432 (7th Cir.2008).

Two statutes guide our analysis. The first is the Chicago Park District Act, which provides in relevant part that “[t]he commissioners of [the Park District] constitute the corporate authorities thereof, and have full power to manage and control all the officers and property of the district, and all parks, driveways, boulevards and parkways maintained by such district or committed to its care and custody.” 70 III. Comp. Stat. 1505/7.01. The district court noted the statute’s use of the plural “commissioners” and “authorities” and concluded from this that each individual commissioner was a separate corporate *308“authority” with the power to unilaterally bind the Park District.

This conclusion strains the statutory language and ignores how public bodies customarily operate. It also contradicts another provision in the Illinois Park District Code, which applies to all Illinois park districts and must be read in conjunction with the Chicago Park District Act. The Illinois Park District Code states:

No member of the board of any park district ... shall have power to create any debt, obligation, claim or liability, for or on account of said park district ... except with the express authority of said board conferred at a meeting thereof and duly recorded in a record of its proceedings.

70 III. Comp. Stat. 1205/4-6 (emphasis added). When read together, these statutes confirm that there is only one corporate authority of the Chicago Park District — its Board of Commissioners — and that individual commissioners cannot unilaterally bind the Park District’s Board to a contract without express Board approval.

There is no evidence that the Park District’s Board of Commissioners authorized Commissioner Burroughs to enter into a contract with Kelley. Moreover, Illinois law provides that ultra vires contracts entered into by municipal corporations are invalid, see, e.g., McMahon v. City of Chicago, 339 Ill.App.3d 41, 273 Ill.Dec. 447, 789 N.E.2d 347, 350 (2003), so Commissioner Burroughs’s offhand remark cannot have created a valid implied-in-fact contract. The judgment for Kelley on the contract claim was premised on legal error; the Park District was entitled to judgment on this claim.

For the foregoing reasons, we Affirm the judgment in favor of the Park District on the VARA claim; we Reverse the judgment in favor of Kelley on the contract claim and Remand with instructions to enter judgment for the Park District.

EASTERBROOK, Circuit Judge.

In 1985 Cosmair, Inc., concluded that young women craved pink and blue hair. To meet the anticipated demand, Cosmair developed a line of “hair cosmetics” — hair coloring that is easily washed out. These inexpensive products, under the name Zazu, were sold in the cosmetic sections of mass merchandise stores. Apparently the teenagers of the late 1980s had better taste than Cosmair’s marketing staff thought. The product flopped, but its name gave rise to this trademark suit. Cosmair is the United States licensee of L’Oréal, S.A., a French firm specializing in perfumes, beauty aids, and related products. Cosmair placed L’Oréal’s marks on the bottles and ads. For reasons the parties have not explained, L’Oréal rather than Cosmair is the defendant even though the events that led to the litigation were .orchestrated in New York rather than Paris. L’Oréal does not protest, so for simplicity, we refer to Cos-mair and L’Oréal collectively as “L’Oréal.”

L’Oréal hired Wordmark, a consulting firm, to help it find a name for the new line of hair cosmetics. After checking the United States Trademark Register for conflicts, Wordmark suggested 250 names. L’Oréal narrowed this field to three, including Zazu, and investigated their availability. This investigation turned up one federal registration of Zazu as a mark for clothing and two state service mark registrations including that word. One of these is Zazú Hair Designs; the other was defunct.

Zazú Hair Designs is a hair salon in Hinsdale, Illinois, a suburb of Chicago. We call it “zhd” to avoid confusion with the Zazú mark, (zhd employs an acute accent and L’Oréal did not; no one makes anything of the difference.) The salon is a partnership between Raymond R. Koubek and Salvatore J. Segretto, hairstylists who joined forces in 1979. zhd registered Zazú with Illinois in 1980 as a trade name for its salon. L’Oréal called the salon to find out if zhd was selling its own products. The employee who answered reported that the salon was not but added, “we’re working on it”. L’Oréal called again; this time it was told that zhd had no products available under the name Zazú.

L’Oréal took the sole federal registration, held by Riviera Slacks, Inc., as a serious obstacle. Some apparel makers have migrated to cosmetics, and if Riviera were about to follow Ralph Lauren (which makes perfumes in addition to shirts and skirts) it might have a legitimate complaint against a competing use of the mark. Sands, Taylor & Wood Co. v. Quaker Oats Co., 24 U.S.P.Q.2d 1001, 1011 (7th Cir.1992). Riviera charged L’Oréal $125,000 for a covenant not to sue if L’Oréal used the Zazu mark on cosmetics. In April 1986, covenant in hand and satisfied that zhd’s state trade name did not prevent the introduction of a national product, L’Oréal made a small interstate shipment of. hair cosmetics under the Zazu name. It used this shipment as the basis of an application for federal registration, filed on June 12, 1986. By August L’Oréal had advertised and sold its products nationally.

Unknown to L’Oréal, Koubek and Segret-to had for some time aspired to emulate Vidal Sassoon by marketing shampoos and *502conditioners under their salon’s trade name. In 1985 Koubek began meeting with chemists to develop zhd’s products. Early efforts were unsuccessful; no one offered a product that satisfied zhd. Eventually zhd received acceptable samples from Gift Cosmetics, some of which Segret-to sold to customers of the salon in plain bottles to which he taped the salon’s business card. Between November 1985 and February 1986 zhd made a few other sales. Koubek shipped two bottles to a friend in Texas, who paid $13. He also made two shipments to a hair' stylist friend in Florida—40 bottles of shampoo for $78.58. These were designed to interest the Floridian in the future marketing of the product line. These bottles could not have been sold to the public, because they lacked labels listing the ingredients and weight. See 21 U.S.C. § 362(b); 15 U.S.C. §§ 1452, 1453(a); 21 C.F.R. §§ 701.3, 701.13(a). After L’Oréal’s national marketing was under way, its representatives thrice visited zhd and found that the salon still had no products for sale under the Zazú name. Which is not to say that zhd was supine. Late in 1985 zhd had ordered 25,000 bottles silk-screened with the name Zazú. Later it ordered stick-on labels listing the ingredients of its products. . In September 1986 zhd began to sell small quantities of shampoo in bottles filled (and labeled) by hand in the salon. After the turn of the year zhd directed the supplier of the shampoo and conditioner to fill some bottles;, the record does not reveal how many.

After a bench trial the district court held that zhd’s sales gave it an exclusive right to use the Zazú name nationally for hair products. 9 U.S.P.Q.2d 1972 (N.D.Ill.1988). The court enjoined L’Oréal from using the mark (a gesture, since the product had bombed and L’Oréal disclaimed any interest in using Zazu again). It also awarded zhd $100,000 in damages on account of lost profits and $1 million more to pay for corrective advertising to restore luster to the Zazú mark. Finding that L’Oréal had infringed zhd’s mark intentionally and used “oppressive and deceitful” tactics in the litigation, the court awarded an additional $1 million in punitive damages, topped off with $76,000 to cover zhd’s legal expenses. (L’Oréal has changed law firms for the appeal; its current counsel did not participate in the events that the district judge found to be unethical.)

Between the filing of the opinion and the entry of judgment, L’Oréal made a motion under Fed.R.Civ.P. 52(b) for an additional hearing. It wanted to present evidence that it believed would show the judge that he misunderstood the actions and motives of its lawyers. The court denied this motion without comment and entered judgment. Not dissuaded, L’Oréal served an all-but-identical motion within ten days of the judgment. Two years later the court granted this motion. After taking additional evidence the court modified some of its findings, but not its judgment. 1991 WL 128694, 1991 U.S.Dist. Lexis 9433. zhd won an additional $92,000 in legal fees to compensate for the cost of the new hearing. L’Oréal at last appealed, and the parties have ignored a potential problem: a second post-trial motion does not extend the time for appeal. Charles v. Daley, 799 F.2d 343 (7th Cir.1986). But because only one of these motions was filed after the judgment, it suspended the judgment’s finality and makes the appeal timely, even though the denial of the first motion shows that the duplicate was unlikely to succeed. United States v. Ibarra, — U.S. -, 112 S.Ct. 4, 116 L.Ed.2d 1 (1991).

I

Federal law permits the registration of trademarks and the enforcement of registered marks. Through § 43(a) of the Lan-ham Act, 15 U.S.C. § 1125(a), a provision addressed to deceit, it also indirectly allows the enforcement of unregistered marks. But until 1988 federal law did not specify how one acquired the rights that could be registered or enforced without registration.† That subject fell into the domain of *503state law, plus federal common law elaborating on the word “use” in § 43(a). Two Pesos, Inc. v. Taco Cabana, Inc., — U.S. -, -, 112 S.Ct. 2753, 2757, 120 L.Ed.2d 615 (1992); id. — U.S. at -, 112 S.Ct. at 2766-67 (Thomas, J., concurring). See also 15 U.S.C. § 1127 (“ ‘trademark’ includes any word ... used by a person”). At common law, “use” meant sales to the public of a product with the mark attached. Trade-Mark Cases, 100 U.S. 82, 94-95, 25 L.Ed. 550 (1879); Menendez v. Holt, 128 U.S. 514, 520-21, 9 S.Ct. 143, 143-44, 32 L.Ed. 526 (1888). See also Hanover Star Milling Co. v. Metcalf, 240 U.S. 403, 414, 36 S.Ct. 357, 360, 60 L.Ed. 713 (1916); United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90, 97, 39 S.Ct. 48, 50, 63 L.Ed. 141 (1918).

“Use” is neither a glitch in the Lanham Act nor a historical relic. By insisting that firms use marks to obtain rights in them, the law prevents entrepreneurs from reserving brand names in order to make their rivals’ marketing more costly. Public sales let others know that they should not invest resources to develop a mark similar to one already used in the trade. Blue Bell, Inc. v. Farah Manufacturing Co., 508 F.2d 1260, 1264-65 (5th Cir.1975); see also William M. Landes and Richard A. Posner, Trademark Law: An Economic Perspective, 30 J.L. & Econ. 265, 281-84 (1987). Only active use allows consumers to associate a mark with particular goods and notifies other firms that the mark is so associated.

Under the common law, one must win the race to the marketplace to establish the exclusive right to a mark. Blue Bell v. Farah; La Societe Anonyme des Parfums LeGalion v. Jean Patou, Inc., 495 F.2d 1265, 1271-74 (2d Cir.1974). Registration modifies this system slightly, allowing slight sales plus notice in the register to substitute for substantial sales without notice. 15 U.S.C. § 1051(a). (The legislation in 1988 modifies the use requirement further, but we disregard this.) zhd’s sales of its product are insufficient use to establish priority over L’Oréal. A few bottles sold over the counter in Hinsdale, and a few more mailed to friends in Texas and Florida, neither link the Zazú mark with zhd’s product in the minds of consumers nor put other producers on notice. As a practical matter zhd had no product, period, until months after L’Oréal had embarked on its doomed campaign.

In finding that zhd’s few sales secured rights against the world, the district court relied on cases such as Department of Justice v. Calspan Corp., 578 F.2d 295 (C.C.P.A.1978), which hold that a single sale, combined with proof of intent to go on selling, permit the vendor to register the mark. See also Axton-Fisher Tobacco Co. v. Fortune Tobacco Co., 82 F.2d 295 (C.C.P.A.1936); Maternally Yours, Inc. v. Your Maternity Shop, Inc., 234 F.2d 538, 542 (2d Cir.1956); Community of Roquefort v. Santo, 443 F.2d 1196 (C.C.P.A.1971). But use sufficient to register a mark that soon is widely distributed is not necessarily enough to acquire rights in the absence of registration. The Lanham Act allows only trademarks “used in commerce” to be registered. 15 U.S.C. § 1051(a). Courts have read “used” in a way that allows firms to seek protection for a mark before investing substantial sums in promotion. See Fort Howard Paper Co. v. Kimberly-Clark Corp., 390 F.2d 1015 (C.C.P.A.1968); cf. Jim Dandy Co. v. Martha White Foods, Inc., 458 F.2d 1397, 1399 (C.C.P.A.1972) (party may rely on advertising to show superior registration rights); but see Weight Watchers International, Inc. v. I. Rokeach & Sons, Inc., 211 U.S.P.Q. 700, 709 (T.M.T.A.B.1981) (more than minimal use is required to register because the statute allows only “owner[s]” to register, and ownership of a mark depends on commer*504cial use). Liberality in registering marks is not problematic, because the registration gives notice to latecomers, which token use alone does not. Firms need only search the register before embarking on development. Had zhd registered Zazú, the parties could have negotiated before L’Oréal committed large sums to marketing.

zhd applied for registration of Zazú after L’Oréal not only had applied to register the mark but also had put its product on the market nationwide. Efforts to register came too late. At oral argument zhd suggested that L’Oréal’s knowledge of zhd’s plan to enter the hair care market using Zazú establishes zhd’s superior right to the name. Such an argument is unavailing. Intent to use a mark, like a naked registration, establishes no rights at all. Hydro-Dynamics, Inc. v. George Putnam & Co., 811 F.2d 1470, 1472 (Fed.Cir.1987). Even under the 1988 amendments (see note †), which allow registration in advance of contemplated use, an unregistered plan to use a mark creates no rights. Just as an intent to buy a choice parcel of land does not prevent a rival from closing the deal first, so an intent to use a mark creates no rights a competitor is bound to respect. A statute granting no rights in bare registrations cannot plausibly be understood to grant rights in “intents” divorced from either sales or registrations. Registration itself establishes only a rebuttable presumption of use as of the filing date. Rolley, Inc. v. Younghusband, 204 F.2d 209, 211 (9th Cir.1953). zhd made first use of Zazú in connection with hair services in Illinois, but this does not translate to a protectable right to market hair products nationally. The district court construed L’Oréal’s knowledge of zhd’s use of Zazú for salon services as knowledge “of [zhd’s] superior rights in the mark.” 9 U.S.P.Q.2d at 1978. zhd did not, however, have superi- or rights in the mark as applied to hair products, because it neither marketed such nor registered the mark before L’Oréal’s use. Because the mark was not registered for use in conjunction with hair products, any knowledge L’Oréal may have had of zhd’s plans is irrelevant. Cf. Weiner King, Inc. v. Wiener King Corp., 615 F.2d 512 (C.C.P.A.1980).

Imagine the consequences of zhd's approach. Businesses that knew of an intended use would not be entitled to the mark even if they made the first significant use of it. Businesses with their heads in the sand, however, could stand on the actual date they introduced their products, and so would have priority over firms that intended to use a mark but had not done so. Ignorance would be rewarded — and knowledgeable firms might back off even though the rivals’ “plans” or “intent” were unlikely to come to fruition. Yet investigations of the sort L’Oréal undertook prevent costly duplication in the development of trademarks and protect consumers from the confusion resulting from two products being sold under the same mark. See Natural Footwear Ltd. v. Hart, Shaffner & Marx, 760 F.2d 1383, 1395 (3d Cir.1985). L’Oréal should not be worse off because it made inquiries and found that, although no one had yet used the mark for hair products, zhd intended to do so. Nor should a potential user have to bide its time until it learns whether other firms are serious about marketing a product. The use requirement rewards those who act quickly in getting new products in the hands of consumers. Had L'Oréal discovered that zhd had a product on the market under the Zazú mark or that zhd had registered Zazú for hair products, L’Oréal could have chosen another mark before committing extensive marketing resources. Knowledge that zhd planned to use the Zazú mark in the future does not present an obstacle to L’Oréal’s adopting it today. Selfway, Inc. v. Travelers Petroleum, Inc., 579 F.2d 75, 79 (C.C.P.A.1978).

Occasionally courts suggest that “bad faith” adoption of a mark defeats a claim to priority. See California Cedar Products Co. v. Pine Mountain Corp., 724 F.2d 827, 830 (9th Cir.1984); Stern Electronics, Inc. v. Kaufman, 669 F.2d 852, 857 (2d Cir.1982); Blue Bell v. Farah, 508 F.2d at 1267.. Although zhd equates L'Oréal’s knowledge of its impending use with “bad faith,” the cases use the term differently. In each instance the court applied the label *505“bad faith” to transactions designed merely to reserve a mark, not to link the name to a product ready to be sold to the public. In California Cedar Products, for example, two firms sprinted to acquire the abandoned Duraflame mark. One shipped some of its goods in the abandoning company’s wrapper with a new name pasted over it. Two days later the other commenced bona fide sales under the Duraflame mark. The court disregarded the first shipment, calling it “both premature and in bad faith”, 724 F.2d at 830, and held that the first firm to make bona fide sales to customers was the prior user. “Bad faith” was no more than an epithet stapled to the basic conclusion: that reserving a mark is forbidden, so that' the first producer to make genuine sales gets the rights. If these eases find a parallel in our dispute, zhd occupies the place of the firm trying to reserve a mark for “intended” exploitation. zhd doled out a few samples in bottles lacking labeling necessary for sale to the public. Such transactions are the sort of pre-marketing maneuvers that these cases hold insufficient to establish rights in a trademark.

The district court erred in equating a use sufficient to support registration with a use sufficient to generate nationwide rights in the absence of registration. Although whether zhd’s use is sufficient to grant it rights in the Zazú mark is a question of fact on which appellate review is deferential, California Cedar Products, 724 F.2d at 830; cf. Scandia Down Corp. v. Euroquilt, Inc., 772 F.2d 1423, 1428-29 (7th Cir.1985), the extent to which zhd used the mark is not disputed, zhd’s sales of hair care products were insufficient as a matter of law to establish national trademark rights at the time L’Oréal put its electric hair colors on the market.

II

A second, and independently sufficient, error requires us to set aside the judgment. zhd did not establish that L’Oréal’s sales injured it in the slightest, let alone that it is entitled to $2.1 million plus hefty attorneys’ fees.

The district court awarded zhd $100,000 as compensation for lost profits. Because zhd did not have any track record of sales preceding L’Oréal’s campaign, the district judge assumed that zhd would sell shampoo or conditioner in all of the 25,000 bottles. Yet ownership of bottles hardly establishes that zhd could sell a single quart of shampoo, let alone that it would make a profit of $4 per bottle, zhd ordered 25,000 bottles not because it had any prospect of selling them but because the vendor offered a quantity discount, zhd did not present evidence showing that other salons that have tried to sell hair products nationally made profits approximating $100,000. As far as the record reveals, zhd approached only four other salons about selling its product. What these salons could have sold is unknown, zhd’s salon sold about $12,000 of other firms’ products per year; nothing in the record enables us to tell the extent to which these sales would have been displaced by products with a Zazú mark, or how much more' profit per bottle zhd would have made selling its own shampoo rather than Vidal Sassoon’s. In the years since L’Oréal cried uncle, zhd has done nothing to enter the market, using Zazú or any other mark. The 25,000 bottles are in storage, most of them empty. If there are indeed profits to be made, zhd has been singularly lax in reaping them.

Nothing but conjecture supports a conclusion that zhd could have made $4 per bottle on 25,000 bottles. The district court brushed aside the lack of proof on the rationale that “the speculation which is inherent in the award of damages to plaintiff was knowingly and intentionally brought upon defendant by itself.” 9 U.S.P.Q.2d at 1979. Although uncertainty created by wrongful acts does not insulate the wrongdoer from liability, “people who want damages have to prove them, using methodologies that need not be intellectually sophisticated but must not insult the intelligence.” Schiller & Schmidt, Inc. v. Nordisco Corp., 969 F.2d 410, 415 (7th Cir.1992). See Bigelow v. RKO Radio Pictures, Inc., 327 U.S. 251, 264, 66 S.Ct. 574, 579, 90 L.Ed. 652 (1946); Sands, Taylor & Wood *506 Co., 24 U.S.P.Q.2d at 1013-14; FDIC v. W.R. Grace & Co., 877 F.2d 614, 624 (7th Cir.1989); Boxhorn’s Big Muskego Gun Club, Inc. v. Electrical Workers, 798 F.2d 1016, 1021-23 (7th Cir.1986). Compensatory damages must rest on “a just and reasonable estimate of the damage based on relevant data”. Bigelow, 327 U.S. at 264, 66 S.Ct. at 580.. Allowance for uncertainty is one thing, see Grove Fresh Distributors, Inc. v. New England Apple Products Co., 969 F.2d 552 (7th Cir.1992), and rank speculation another, zhd offered no evidence to allow a reasonable estimate that it would have made $100,000 from its nascent products had L'Oréal not sold zany hair colors under the Zazu mark.

The district court added $1 million to the $100,000, for a total compensatory award of $1.1 million, on the theory that zhd needed to counteract the effect of L'Oréal's promotional campaign. Although the court called this an award “for corrective advertising,” the judgment allows zhd to do as it pleases with the million; it need not run a single ad. The court arrived at $1 million by taking 20% of the $5 million L’Oréal had spent advertising Zazu hair coloring. The court’s only explanation was that “[t]he evidence shows that defendant injured the mark. In these circumstances, plaintiff is entitled to try to resurrect it.” 1991 WL 128694 at *2, 1991 U.S.Dist. Lexis 9433 at *4-*5.

Like the $100,000, the $1 million is an arbitrary number; an award of $6% would be neither more nor less supportable. Using a percentage of some number unrelated to the plaintiff's injury is an unacceptable way to estimate damages, a point we recently made in another trademark ease. Sands, Taylor & Wood Co., 24 U.S.P.Q.2d at 1014.

Any compensatory award depends on loss, and in treating the need for advertising as a “loss” the court overlooked the principle that a trademark cannot be worth less than zero. “Corrective advertising” is a method of repair. Defendant diminishes the value of plaintiff’s trademark, and advertising restores that mark to its original value. E.g., Big O Tire Dealers, Inc. v. Goodyear Tire & Rubber Co., 561 F.2d 1365, 1374-76 (10th Cir.1977); Comment, Money Damages and Corrective Advertising: An Economic Analysis, 55 U.Chi.L.Rev. 629 (1988). Expenses for repair cannot be justified when they exceed the value of the asset. If a car worth $4,000 is crushed in a collision and repair would cost $10,000, the court awards damages of $4,000, not $10,000 (or, as seems to have happened in our case, $14,-000).

To justify damages to pay for corrective advertising a plaintiff must show that the confusion caused by the defendant’s mark injured the plaintiff and that “repair” of the old trademark, rather than adoption of a new one, is the least expensive way to proceed, zhd established neither element. Although both zhd and the district court refer to L’Oréal’s product as “garish,” zhd did not offer any evidence that consumers (even those who found orange hair revolting) would be less willing to use shampoo tastefully marketed under the same name. Indeed, zhd did not offer evidence that a single consumer remembered L’Oréal’s products by the time of trial. If as the district court found L'Oréal “injured the mark to such an extent that it may never be useful to plaintiff in conjunction with the merchandising of plaintiff’s hair care products”, 9 U.S.P.Q.2d at 1979, the appropriate course is to forget about Zazú and award damages fully compensating for the mark’s value at the time L’Oréal pounced. See Coursey v. Broadhurst, 888 F.2d 338, 344 (5th Cir.1989); Phillips Petroleum Co. v. Stokes Oil Co., Inc., 863 F.2d 1250, 1257 (6th Cir.1988). Choosing another name under which to market its product will cost zhd less than $1 million— zhd spent far less than this on the imprinted bottles, never advertised its product, and did not come within an order of magnitude of this sum in promoting the salon throughout its existence, zhd made no claim that the salon had suffered as a result of L’Oréal’s use of Zazu, and few people were aware of zhd’s plans to market hair products under the Zazú name. A new name thus can be chosen at a low cost. *507zhd has made no corrective advertising expenditures — suggesting either that LiOréal’s use of Zazu did not injure zhd or that it is cheaper for zhd to use a new name than to correct the confusion caused by L’Oréal’s use.

The final $1 million of the award was explicitly punitive. The district court found that L’Oréal had wilfully infringed zhd’s mark and that “its conduct before and during the litigation ha[d] been oppressive and deceitful.” 9 U.S.P.Q.2d at 1979. Believing at the time that L’Oréal’s net worth was $20 million, the court opined that an award of 5% of this sum was necessary to deter such conduct in the future. Punitive damages are problematic because the Lanham Act, although providing for the trebling of compensatory damages, forbids other penalties. 15 U.S.C. § 1117(a)'; Getty Petroleum Corp. v. Barto Petroleum Corp., 858 F.2d 103 (2d Cir.1988); cf. Fleischmann Distilling Corp. v. Maier Brewing Co., 386 U.S. 714, 87 S.Ct. 1404, 18 L.Ed.2d 475 (1967). The district court found a punitive award authorized by the law of Illinois without explaining where one finds such authorization. The parties have been of no greater assistance. Because some old cases say that Illinois law supports punitive awards in trademark cases, Aladdin Mfg. Co. v. Mantle Lamp Co., 116 F.2d 708, 716-17 (7th Cir.1941); Keller Products, Inc. v. Rubber Linings Corp., 213 F.2d 382, 387 (7th Cir.1954), and L’Oréal has not asked us to revisit the subject, we press forward.

In stating that L’Oréal’s conduct was wilful, the district judge seems to have equated this term with knowing or intentional. Wilful is a word with many uses in the law, but even the weaker connotations imply knowledge of unlawfulness or reckless indifference to the law, and not simply intentional acts. E.g., McLaughlin v. Richland Shoe Co., 486 U.S. 128, 108 S.Ct. 1677, 100 L.Ed.2d 115 (1988); TWA v. Thurston, 469 U.S. 111, 105 S.Ct. 613, 83 L.Ed.2d 523 (1985). L’Oréal found out that zhd was using Zazu as a service mark, not that zhd had protected rights in Zazú as a mark for hair care products. The district court did not conclude that L’Oréal actually knew that its use of Zazu infringed, or didn’t care; L’Oréal’s whole course of investigation suggests the opposite.

Bad conduct during the litigation is a more secure footing for a penalty. Counsel told the judge that L’Oréal sold a substantial volume of Zazu hair cosmetics in northern Illinois, leading the court to set an injunction bond too high for zhd to post. Later the judge learned that these were wholesale sales; local retail sales, which the judge wanted to use as the measure of the bond for an injunction keeping Zazu hair cosmetics out of zhd’s home territory, were much smaller. The judge accused counsel of deceit. Although counsel protest, saying that they were misunderstood (and that they told the judge, when asked unexpectedly about sales, that they had only wholesale numbers), we accept the district judge’s finding. Counsel committed a clearer misrepresentation about L’Oréal’s net worth, which zhd sought in order to make a pitch for hefty punitive damages. L’Oréal’s lawyers opposed the request for information, representing that under French law the firm, as a privately held business, had a right to keep such matters secret. The district judge then used $20 million as a rough-and-ready estimate. Later the court learned that L’Oréal is a publicly traded firm, whose glossy annual report ballyhoos a net worth exceeding $1 billion. Counsel did not bother to ascertain the status of their client and regaled the judge with a fable. The court also concluded that L’Oréal’s attorneys made false statements to the Trademark Office and unethically tried to negotiate directly with zhd’s partners rather than communicating through its lawyer.

Contempt of court, Fed.R.Civ.P. 11, and 28 U.S.C. § 1927 provide the principal vehicles for penalizing misconduct in litigation, and before imposing sanctions a judge ought to ask and answer the questions these bodies of law pose. Still, calling the penalty “punitive damages” is inconsequential given Chambers v. NASCO, Inc., — U.S. -, 111 S.Ct. 2123, 115 L.Ed.2d 27 (1991), which holds that judges *508may impose sanctions without express authority in statute or rule. See also Business Guides, Inc. v. Chromatic Communications Enterprises, Inc., 498 U.S. 533, 111 S.Ct. 922, 112 L.Ed.2d 1140 (1991). Is $1 million the right penalty for counsel’s acts? A judge sanctioning misconduct may not draw a number from the ¿Ether but must explain the choice by reference to its role in compensating the wronged party or deterring conduct that injures the judicial system. In re Oil Spill by the Amoco Cadiz, 954 F.2d 1279, 1334 (7th Cir.1992); cf. Frantz v. United States Powerlifting Federation, 836 F.2d 1063, 1066 (7th Cir.1987). After Pacific Mutual Life Insurance Co. v. Haslip, — U.S. -, 111 S.Ct. 1032, 113 L.Ed.2d 1 (1991), courts should be especially careful to justify the level of punitive awards.

One million dollars cannot be justified as necessary to either compensation or deterrence. The judge discussed neither. Instead he calculated the award as a percentage of L’Oréal’s (supposed) net worth — as if having a large net worth were the wrong to be deterred! Punitive damages are appropriate when some wrongful conduct evades detection; a multiplier then both compensates and deters. Punitive awards also are appropriate when the conduct in question is wholly antisocial, for then excessive awards cannot deter borderline conduct that may be beneficial. FDIC v. W.R. Grace & Co., 877 F.2d 614, 623 (7th Cir.1989). Some trademark infringements are concealed; for example, passing off a lower quality product under a copy of another’s mark. L’Oréal’s conduct, however, was not of this type. L’Oréal did not try to palm off its product as zhd’s, and there was little chance that L’Oréal’s national advertisements and sales would escape notice. L’Oréal’s use was not only easily detected but also in the class of potentially beneficial uses that could be discouraged by excessive awards. So, too, its conduct of the litigation was well known to zhd. Employing L’Oréal’s net worth as the starting point did not help the court identify concealable offenses and may have misled it about the appropriate size of the award.

Courts take account of a defendant’s wealth when “[a]n amount sufficient to punish or to deter one individual may be trivial to another.” Black v. Iovino, 219 Ill.App.3d 378, 394, 162 Ill.Dec. 513, 524, 580 N.E,2d 139, 150 (1st Dist.1991). See also Douglass v. Hustler Magazine, Inc., 769 F.2d 1128, 1145 (7th Cir.1985). For natural persons the marginal utility of money decreases as wealth increases, so that higher fines may be needed to deter those possessing great wealth. (“May be” is an important qualifier; the entire penalty includes extra-judicial consequences, such as loss of business and other future income, that is likely to be greater for wealthier defendants.) Corporations, however, are not wealthy in the sense that persons are. Corporations are abstractions; investors own the net worth of the business. These investors pay any punitive awards (the value of their shares decreases), and they may be of average wealth. Pension trusts and mutual funds, aggregating the investments of millions of average persons, own the bulk of many large corporations. Seeing the corporation as wealthy is an illusion, which like other mirages frequently leads people astray.

Corporate assets finance ongoing operations and are unrelated to either the injury done to the victim or the size of the award needed to cause corporate managers to obey the law. Net worth is a measure of profits that have not yet been distributed to the investors. Why should damages increase because the firm reinvested its earnings? Absolute size, like net worth, also is a questionable reason to extract more per case. If L’Oréal introduces 10,000 products, perhaps 10 of these infringe someone else’s trademark. Awards of ordinary damages in all 10 cases deter wrongful conduct. The net judgment bill of the firm will increase with its wrongs, without the need for punitive damages. If a larger firm is more likely to commit a wrong on any given transaction, then its total damages will increase more than proportionally to its size without augmentation in any given case; if a larger firm is equally or less likely to commit a tort per transaction, then the court ought to praise the manag*509ers rather than multiply the firm’s penalty. Consider: General Motors is much larger than Chrysler, and so makes more defective cars, but the goals of compensation and deterrence are achieved for both firms by awarding as damages the injury produced per defective car. Corporate size is a reason to magnify damages only when the wrongs of larger firms are less likely to be punished; yet judges rarely have any reason to suppose this, and the court in this case had none.

None of the awards — not the $100,000 for lost profits, not the $1 million for corrective advertising, and not the $1 million in punitive damages — rests on an adequate foundation. Given our conclusion in Part I, L’Oréal prevails on the merits and zhd is not entitled to damages of any , size, or to attorneys’ fees under the' Lanham Act. The case is remanded so that the district court may consider whether a sanction is appropriate for L’Oréal’s misconduct during the litigation.

Reversed and Remanded.

CUDAHY, Circuit Judge,

dissenting.

On the important issue of good faith, L’Oréal’s conduct here merits a very hard look. In the case of Riviera, a men’s clothing retailer, L’Oréal was careful to pay $125,000 for an agreement not to sue. Yet men’s clothing and hair cosmetics marketed to women hardly seem related at all. On the other hand, a women’s hair salon developing a line of hair care products is a purveyor of goods and services that seem closely related to hair cosmetics. Therefore, L’Oréal’s knowledge of ZHD’s use defeats any claim L’Oréal may have to priority.

One of the keys here seems to be the use of ZAZU as a service mark connected with the provision of salon services by ZHD. A service mark can be infringed by its use on a closely related product.1 See, e.g., Re Hyper Shoppes (Ohio), Inc., 837 F.2d 463 (Fed.Cir.1988) (affirming refusal to register two “closely related” marks: one for furniture and the other for general merchandise services that included the sale of furniture); Steelcase, Inc. v. Steelcare, Inc., 219 U.S.P.Q. 433 (TMTAB 1983) (mark used for furniture refinishing service likely to be confused with mark used for office furniture); see also 2 J. Thomas McCarthy, Trademark and Unfair Competition § 24:6, at 71 (2d ed. 1984 & Supp.1991) (stating that “[wjhere the services consist of retail sales services, likelihood of confusion is found when another mark is used on goods which are commonly sold through such a retail outlet”). A service and a product are related if buyers are likely to assume a common source or sponsorship. Tisch Hotels, Inc. v. Americana Inn, Inc., 350 F.2d 609, 612 (7th Cir.1965); 2 McCarthy, supra, § 24:6 at 183 (“The ‘related goods’ test is merely a facet of the ultimate and final test of ‘likelihood of confusion.’ ”). The salon services and hair products at issue in this case, which are nearly as kindred as a service and product can be, offer the paradigmatic illustration of things that are closely related. Thus the majority’s disregard for ZHD’s substantial use of ZAZU in connection with salon services is unfounded.

As the majority correctly notes, the standard for granting federal registration is somewhat less exacting than that for establishing common law trademark rights. See La Societe Anonyme des Parfums le Galion v. Jean Patou, Inc., 495 F.2d 1265 (2d Cir.1974). But that distinction is so slight as to be inconsequential here. While any bona fide transaction that is more than a “mere sham” will, when combined with intent to continue use, suffice to support federal registration, Avakoff v. Southern *510 Pacific Co., 765 F.2d 1097, 1098 (Fed.Cir.1985); 1 McCarthy, supra, § 19:37 at 967, any use greater than de minimis will still warrant trademark protection in the absence of registration. See SweeTarts v. Sunline, Inc., 380 F.2d 923, 929 (8th Cir.1967) (market penetration “need not be large” but must be “significant enough to pose the real likelihood of confusion”); 1 McCarthy, supra, § 16:2(B) at 723; § 16.-2(C) at 726. Thus, bona fide test marketing or small experimental sales — indeed, any use that is not nominal or token — can satisfy the test. E.I. du Pont de Nemours & Co. v. G.C. Murphy Co., 199 U.S.P.Q. 807, 812 (TMTAB 1978); 1 McCarthy, supra, § 16:2(B) at 723.

In this case, ZHD’s use of the ZAZU mark, both in its highly successful salon service business, which drew some out-of-state clients, and in its local and interstate product sales to customers and to a potential marketer, surely is more than de min-imis. The extensive evidence of ZHD’s intent to step up hair product sales — such as its order for 25,000 ZAZU-emblazoned bottles and its inquiry about advertising rates in a national magazine — bolsters this assessment. Even if ZHD did fail to demonstrate more than a de minimis market penetration nationally, at the very least it successfully established exclusive rights within its primary area of operation. The salon’s substantial advertising, increasing revenue and staff and preliminary product sales indicate sufficient market penetration to afford trademark protection in that region. See Natural Footwear Ltd. v. Hart, Schaffner & Marx, 760 F.2d 1383 (3d Cir.) (senior user can establish common law rights in geographic areas where it achieved market penetration), cert. denied, 474 U.S. 920, 106 S.Ct. 249, 88 L.Ed.2d 257 (1985); SweeTarts, 380 F.2d at 928 (plaintiff’s prior use of trademark within a given market area entitles it to exclusive use of that mark within that area); V & V Food Products, Inc. v. Cacique Cheese Co., 683 F.Supp. 662 (N.D.Ill.1988).

L’Oréal concedes that ZHD has exclusive rights to use ZAZU for salon services in the Hinsdale area. Those exclusive rights also preclude L’Oréal from using the mark on hair products in the local area because of the likelihood of confusion between those products and ZHD’s salon services, even apart from any confusion between the two parties’ products. Given the deferential standard of review on the factual question of use, therefore, I think it clear that ZHD has achieved market penetration and exclusive rights to the ZAZU mark at the very least in the Chicago area.

ZHD’s contention that its rights in the ZAZU mark extend beyond the local area is enhanced by evidence that L’Oréal did not, as we have noted, act in good faith. The majority’s consideration of the good faith issue minimizes the important role good faith plays in trademark disputes, particularly disputes involving unregistered marks. See 2 McCarthy, supra,' § 26:4 at 292 (noting thát good' faith • has “significant, if not independent, importance” in disputes involving unregistered marks). Bad faith can defeat claims to priority. See, e.g., A.J. Canfield Co. v. Honickman, 808 F.2d 291 (3rd Cir.1986) (stating the doctrine that a senior user “has enforceable rights against any junior user who adopted the mark with knowledge of its senior use”); Johanna Farms, Inc. v. Citrus Bowl, Inc., 468 F.Supp. 866, 876 (E.D.N.Y.1978) (stating that junior users who do not adopt the mark in good faith “will not be accorded the right to exploit the mark, even in areas previously untouched by the senior user”); 2 McCarthy, supra, § 26:3 at 289; § 26:4 at 292. Contrary to the majority’s, narrow characterization of bad faith as a concept employed solely to.deter attempts to reserve marks prior to genuine sales, courts have examined junior users’ good faith in a variety of contexts. In fact, this court has held that a good faith junior user is simply one that begins using a mark without knowledge that another party already is using it. The Money Store v. Harriscorp Finance, Inc., 689 F.2d 666, 674 (7th Cir.1982); see 2 McCarthy, supra, § 26:4 at 292 (equating good faith to “the junior user’s lack of knowledge”). And while such knowledge may not automatically negate good faith, only the most unusual situations encom*511pass both knowledge and good faith. See GTE Corp. v. Williams, 904 F.2d 536, 541 (10th Cir.), cert. denied, - U.S. -, 111 S.Ct. 557, 112 L.Ed.2d 564 (1990).

In sum, I believe ZHD’s sale of ZAZU services and products constituted sufficient use to establish exclusive trademark rights in the Chicago area, if not nationally. In any event, L’Oréal’s pursuit of its line of hair cosmetics in spite of its knowledge of ZHD’s use defeats L’Oréal’s position. Therefore, I respectfully dissent.