delivered the opinion of the Court.
This ease presents an important question of statutory interpretation arising under the patent laws. The issue before us is whether the owner of a patent on a chemical process is guilty of patent misuse, and therefore is barred from seeking relief against contributory infringement of its patent rights, if it exploits the patent only in conjunction with the sale of an unpatented article that constitutes a material part of the invention and is not suited for commercial use outside the scope of the patent claims. The answer will determine whether respondent, the owner of a process patent on a chemical herbicide, may maintain an action for contributory infringement against other manufacturers of the chemical used in the process. To resolve this issue, we must construe the various provisions of 35 U. S. C. §271, which Congress enacted in 1952 to codify certain aspects of the doctrines of contributory infringement and patent misuse that previously had been developed by the judiciary.
I
The doctrines of contributory infringement and patent misuse have long and interrelated histories. The idea that a patentee should be able to obtain relief against those whose *180acts facilitate infringement by others has been part of our law since Wallace v. Holmes, 29 F. Cas. 74 (No. 17,100) (CC Conn. 1871). The idea that a patentee should be denied relief against infringers if he has attempted illegally to extend the scope of his patent monopoly is of somewhat more recent origin, but it goes back at least as far as Motion Picture Patents Co. v. Universal Film Mfg. Co., 243 U. S. 502 (1917). The two concepts, contributory infringement and patent misuse, often are juxtaposed, because both concern the relationship between a patented invention and unpatented articles or elements that are needed for the invention to be practiced.
Both doctrines originally were developed by the courts. But in its 1952 codification of the patent laws Congress endeavored, at least in part, to substitute statutory precepts for the general judicial rules that had governed prior to that time. Its efforts find expression in 35 U. S. C. § 271:
“(a) Except as otherwise provided in this title, whoever without authority makes, uses or sells any patented invention, within the United States during the term of the patent therefor, infringes the patent.
“(b) Whoever actively induces infringement of a patent shall be liable as an infringer.
“(c) Whoever sells a component of a patented machine, manufacture, combination or composition, or a material or apparatus for use in practicing a patented process, constituting a material part of the invention, knowing the same to be especially made or especially adapted for use in an infringement of such patent, and not a staple article or commodity of commerce suitable for substantial noninfringing use, shall be liable as a contributory infringer.
“(d) No patent owner otherwise entitled to relief for infringement or contributory infringement of a patent shall be denied relief or deemed guilty of misuse or illegal extension of the patent right by reason of his having *181done one or more of the following: (1) derived revenue from acts which if performed by another without his consent would constitute contributory infringement of the patent; (2) licensed or authorized another to perform acts which if performed without his consent would constitute contributory infringement of the patent; (3) sought to enforce his patent rights against infringement or contributory infringement.”
Of particular import to the present controversy are subsections (c) and (d). The former defines conduct that constitutes contributory infringement; the latter specifies conduct of the patentee that is not to be deemed misuse.
A
The catalyst for this litigation is a chemical compound known to scientists as “3, 4-diehloropropionanilide” and referred to in the chemical industry as “propanil.” In the late 1950’s, it was discovered that this compound had properties that made it useful as a selective, “post-emergence” herbicide particularly well suited for the cultivation of rice. If applied in the proper quantities, propanil kills weeds normally found in rice crops without adversely affecting the crops themselves. It thus permits spraying of general areas where the crops are already growing, and eliminates the necessity for hand weeding or flooding of the rice fields. Propanil is one of several herbicides that are commercially available for use in rice cultivation.
Efforts to obtain patent rights to propanil or its use as a herbicide have been continuous since the herbicidal qualities of the chemical first came to light. The initial contender for a patent monopoly for this chemical compound was the Monsanto Company. In 1957, Monsanto filed the first of three successive applications for a patent on propanil itself. After lengthy proceedings in the United States Patent Office, a patent, No. 3,382,280, finally was issued in 1968. It was de-*182dared invalid, however, when Monsanto sought to enforce it by suing Rohm and Haas Company (Rohm & Haas), a competing manufacturer, for direct infringement. Monsanto Co. v. Rohm & Haas Co., 312 F. Supp. 778 (ED Pa. 1970), aff’d, 456 F. 2d 592 (CA3), cert. denied, 407 U. S. 934 (1972). The District Court held that propanil had been implicitly revealed in prior art dating as far back as 1902, even though its use as a herbicide had been discovered only recently. 312 F. Supp., at 787-790. Monsanto subsequently dedicated the patent to the public, and it is not a party to the present suit.
Invalidation of the Monsanto patent cleared the way for Rohm & Haas, respondent here, to obtain a patent on the method or process for applying propanil. This is the patent on which the present lawsuit is founded. Rohm & Haas’ efforts to obtain a propanil patent began in 1958. These efforts finally bore fruit when, on June 11, 1974, the United States Patent Office issued Patent No. 3,816,092 (the Wilson patent) to Harold F. Wilson and Dougal H. McRay.1 The patent contains several claims covering a method for applying propanil to inhibit the growth of undesirable plants in areas containing established crops.2' Rohm & Haas has been the sole owner of the patent since its issuance.
*183Petitioners, too, are chemical manufacturers. They have manufactured and sold propanil for application to rice crops since before Rohm & Haas received its patent. They market the chemical in containers on which are printed directions for application in accordance with the method claimed in the Wilson patent. Petitioners did not cease manufacture and sale of propanil after that patent issued, despite knowledge that farmers purchasing their products would infringe on the patented method by applying the propanil to their crops. Accordingly, Rohm & Haas filed this suit, in the United States District Court for the Southern District of Texas, seeking in-junctive relief against petitioners on the ground that their manufacture and sale of propanil interfered with its patent rights.
The complaint alleged not only that petitioners contributed to infringement by farmers who purchased and used petitioners’ propanil, but also that they actually induced such infringement by instructing farmers how to apply the herbicide. See 35 U. S. C. §§ 271 (b) and (c). Petitioners responded to the suit by requesting licenses to practice the patented method. When Rohm & Haas refused to grant such licenses, however, petitioners raised a defense of patent misuse and counterclaimed for alleged antitrust violations by respondent. The parties entered into a stipulation of facts, and petitioners moved for partial summary judgment. They argued that Rohm & Haas has misused its patent by conveying the right to practice the patented method only to purchasers of its own propanil.
The District Court granted summary judgment for petitioners. 191 USPQ 691 (1976). It agreed that Rohm & Haas was barred from obtaining relief against infringers of its patent because it had attempted illegally to extend its patent monopoly. The District Court recognized that 35 U. S. C. *184§ 271 (d) specifies certain conduct which is not to be deemed patent misuse. The court ruled, however, that “[t]he language of § 271 (d) simply does not encompass the totality of [Rohm & Haas'] conduct in this case.” 191 USPQ, at 704. It held that respondent's refusal to grant licenses, other than the “implied” licenses conferred by operation of law upon purchasers of its propanil, constituted an attempt by means of a “tying” arrangement to effect a monopoly over an un-patented component of the process. The District Court concluded that this conduct would be deemed patent misuse under the judicial decisions that preceded § 271 (d), and it held that “[n] either the legislative history nor the language of § 271 indicates that this rule has been modified.” 191 USPQ, at 707.3
The United States Court of Appeals for the Fifth Circuit reversed. 599 F. 2d 685 (1979). It emphasized the fact that propanil, in the terminology of the patent law, is a “nonstaple” article, that is, one that has no commercial use except in connection with respondent’s patented invention. After a thorough review of the judicial developments preceding enactment of § 271, and a detailed examination of the legislative history of that provision, the court concluded that the legislation restored to the patentee protection against contributory infringement that decisions of this Court theretofore had undermined. To secure that result, Congress found it necessary to cut back on the doctrine of patent misuse. The Court of Appeals determined that, by specifying in § 271 (d) conduct that is not to be deemed misuse, “Congress *185did clearly provide for a patentee’s right to exclude others and reserve to itself, if it chooses, the right to sell nonstaples used substantially only in its invention.” 599 F. 2d, at 704 (emphasis in original). Since Rohm & Haas’ conduct was designed to accomplish only what the statute contemplated, the court ruled that petitioners’ misuse defense was of no avail.
We granted certiorari, 444 U. S. 1012 (1980), to forestall a possible conflict in the lower courts4 and to resolve an issue of prime importance in the administration of the patent law.
B
For present purposes certain material facts are not in dispute. First, the validity of the Wilson patent is not in question at this stage in the litigation.5 We therefore must assume that respondent is the lawful owner of the sole and exclusive right to use, or to license others to use, propanil as a herbicide on rice fields in accordance with the methods claimed in the Wilson patent. Second, pétitioners do not dispute that their manufacture and sale of propanil together with instructions for use as a herbicide constitute contributory infringement of the Rohm & Haas patent. Tr. of Oral Arg. 14. Accordingly, they admit that propanil constitutes “a material part of [respondent’s] invention,” that it is “espe*186cially made or especially adapted for use in an infringement of [the] patent,” and that it is “not a staple article or commodity of commerce suitable for substantial noninfringing use,” all within the language of 35 U. S. C. § 271 (c).6 They also concede that they have produced and sold propanil with knowledge that it would be used in a manner infringing on respondent’s patent rights. To put the same matter in slightly different terms, as the litigation now stands, petitioners admit commission of a tort and raise as their only defense to liability the contention, that respondent, by engaging in patent misuse, comes into court with unclean hands.7
As a result of these concessions, our chief focus of inquiry must be the scope of the doctrine of patent misuse in light of the limitations placed upon that doctrine by §271 (d). On this subject, as well, our task is guided by certain stipulations and concessions. The parties agree that Rohm & Haas makes and sells propanil; that it has refused to license petitioners or any others to do the same; that it has not granted express licenses either to retailers or to end users of the product; and that farmers who buy propanil from Rohm & Haas may use it, without fear of being sued for direct infringement, by virtue of an “implied license” they obtain when Rohm & Haas relinquishes its monopoly by selling the propanil. See App. 35-39. See also United States v. Univis Lens Co., 316 U. S. 241, 249 (1942); cf. Adams v. Burke, 17 Wall. 453 (1873). The parties further agree that §§271 (d)(1) and (3) permit respondent both to sell propanil itself and to sue *187others who sell the same product without a license, and that under § 271 (d) (2) it would be free to demand royalties from others for the sale of propanil if it chose to do so.
The parties disagree over whether respondent has engaged in any additional conduct that amounts to patent misuse. Petitioners assert that there has been misuse because respondent has “tied” the sale of patent rights to the purchase of propanil, an unpatented and indeed unpatentable article, and because it has refused to grant licenses to other producers of the chemical compound. They argue that § 271 (d) does not permit any sort of tying arrangement, and that resort to such a practice excludes respondent from the category of patentees “otherwise entitled to relief” within the meaning of § 271 (d). Rohm & Haas, understandably, vigorously resists this characterization of its conduct. It argues that its acts have been only those that § 271 (d), by express mandate, excepts from characterization as patent misuse. It further asserts that if this conduct results in an extension of the patent right to a control over an unpatented commodity, in this instance the extension has been given express statutory sanction.
II
Our mode of analysis follows closely the trail blazed by the District Court and the Court of Appeals. It is axiomatic, of course, that statutory construction must begin with the language of the statute itself. But the language of § 271 is generic and freighted with a meaning derived from the deci-sional history that preceded it. The Court of Appeals appropriately observed that more than one interpretation of the statutory language has a surface plausibility. To place § 271 in proper perspective, therefore, we believe that it is helpful first to review in detail the doctrines of contributory infringement and patent misuse as they had developed prior to Congress’ attempt to codify the governing principles.
As we have noted, the doctrine of contributory infringe*188ment had its genesis in an era of simpler and less subtle technology. Its basic elements are perhaps best explained with a classic example drawn from that era. In Wallace v. Holmes, 29 F. Cas. 74 (No. 17,100) (CC Conn. 1871), the patentee had invented a new burner for an oil lamp. In compliance with the technical rules of patent claiming, this invention was patented in a combination that also included the standard fuel reservoir, wick tube, and chimney necessary for a properly functioning lamp. After the patent issued, a competitor began to market a rival product including the novel burner but not the chimney. Id., at 79. Under the sometimes scholastic law of patents, this conduct did not amount to direct infringement, because the competitor had not replicated every single element of the patentee’s claimed combination. Cf., e. g., Prouty v. Ruggles, 16 Pet. 336, 341 (1842). Yet the court held that there had been “palpable interference” with the patentee’s legal rights, because purchasers would be certain to complete the combination, and hence the infringement, by adding the glass chimney. 29 F. Cas., at 80. The court permitted the patentee to enforce his rights against the competitor who brought about the infringement, rather than requiring the patentee to undertake the almost insuperable task of finding and suing all the innocent purchasers who technically were responsible for completing the infringement. Ibid. See also Bowker v. Dows, 3 F. Cas. 1070 (No. 1,734) (CC Mass. 1878).
The Wallace case demonstrates, in a readily comprehensible setting, the reason for the contributory infringement doctrine. It exists to protect patent rights from subversion by those who, without directly infringing the patent themselves, engage in acts designed to facilitate infringement by others. This protection is of particular importance in situations, like the oil lamp case itself, where enforcement against direct infringers would be difficult, and where the technicalities of patent law make it relatively easy to profit from another’s invention without risking a charge of direct infringement. *189See Thomson-Houston Electric Co. v. Ohio Brass Co., 80 F. 712, 721 (CA6 1897) (Taft, Circuit Judge); Miller, Some Views on the Law of Patent Infringement by Inducement, 53 J. Pat. Off. Soc. 86, 87-94 (1971).
Although the propriety of the decision in Wallace v. Holmes seldom has been challenged, the contributory infringement doctrine it spawned has not always enjoyed full adherence in other contexts. The difficulty that the doctrine has encountered stems not so much from rejection of its core concept as from a desire to delimit its outer contours. In time, concern for potential anticompetitive tendencies inherent in actions for contributory infringement led to retrenchment on the doctrine. The judicial history of contributory infringement thus may be said to be marked by a period of ascendancy, in which the doctrine was expanded to the point where it became subject to abuse, followed by a somewhat longer period of decline, in which the concept of patent misuse was developed as an increasingly stringent antidote to the perceived excesses of the earlier period.
The doctrine of contributory infringement was first addressed by this Court in Morgan Envelope Co. v. Albany Paper Co., 152 U. S. 425 (1894). That case was a suit by a manufacturer of a patented device for dispensing toilet paper against a supplier of paper rolls that fit the patented invention. The Court accepted the contributory infringement doctrine in theory but held that it could not be invoked against a supplier of perishable commodities used in a patented invention. The Court observed that a contrary outcome would give the patentee “the benefit of a patent” on ordinary articles of commerce, a result that it determined to be unjustified on the facts of that case. Id., at 433.
Despite this wary reception, contributory infringement actions continued to flourish in the lower courts.8 Eventually *190the doctrine gained more wholehearted acceptance here. In Leeds & Catlin Co. v. Victor Talking Machine Co., 213 U. S. 325 (1909), the Court upheld an injunction against contributory infringement by a manufacturer of phonograph discs specially designed for use in a patented disc-and-stylus combination. Although the disc itself was not patented, the Court noted that it was essential to the functioning of the patented combination, and that its method of interaction with the stylus was what “mark[ed] the advance upon the prior art.” Id., at 330. It also stressed that the disc was capable of use only in the patented combination, there being no other commercially available stylus with which it would operate. The Court distinguished the result in Morgan Envelope on the broad grounds that “'[n]ot one of the determining factors there stated exists in the case at bar,” and it held that the attempt to link the two cases “is not only to confound essential distinctions made by the patent laws, but essential distinctions between entirely different things.” 213 U. S., at 335.
The contributory infringement doctrine achieved its high-water mark with the decision in Henry v. A. B. Dick Co., 224 U. S. 1 (1912). In that case a divided Court extended contributory infringement principles to permit a conditional licensing arrangement whereby a manufacturer of a patented printing machine could require purchasers to obtain all supplies used in connection with the invention, including such staple items as paper and ink, exclusively from the patentee. The Court reasoned that the market for these supplies was created by the invention, and that sale of a license to use the *191patented product, like sale of other species of property, could be limited by whatever conditions the property owner wished to impose. Id., at 31-32. The A. B. Dick decision and its progeny in the lower courts led to a vast expansion in conditional licensing of patented goods and processes used to control markets for staple and nonstaple goods alike.9
This was followed by what may be characterized through the lens of hindsight as an inevitable judicial reaction. In Motion Picture Patents Co. v. Universal Film Mfg. Co., 243 U. S. 502 (1917), the Court signalled a new trend that was to continue for years thereafter.10 The owner of a patent on projection equipment attempted to prevent competitors from selling film for use in the patented equipment by attaching to the projectors it sold a notice purporting to condition use of the machine on exclusive use of its film. The film previously had been patented but that patent had expired. The Court addressed the broad issue whether a patentee possessed the right to condition sale of a patented machine on the purchase of articles “which are no part of the patented machine, and which are not patented.” Id., at 508. Relying upon the rule that the scope of a patent “must be limited to the invention described in the claims,” id., at 511, the Court held that the attempted restriction on use of unpatented supplies was improper:
“Such a restriction is invalid because such a film is obviously not any part of the invention of the patent in suit; because it is an attempt, without statutory warrant, to continue the patent monopoly in this particular ehar-*192acter of film after it has expired, and because to enforce it would be to create a monopoly in the manufacture and use of moving picture films, wholly outside of the patent in suit and of the patent law as we have interpreted it.” Id., at 518.
By this reasoning, the Court focused on the conduct of the patentee, not that of the alleged infringer. It noted that as a result of lower court decisions, conditional licensing arrangements had greatly increased, indeed, to the point where they threatened to become “perfect instrument [s] of favoritism and oppression.” Id., at 515. The Court warned that approval of the licensing scheme under consideration would enable the patentee to “ruin anyone unfortunate enough to be dependent upon its confessedly important improvements for the doing of business.” Ibid. This ruling was. directly in conflict with Henry v. A. B. Dick Co., supra, and the Court expressly observed that that decision “must be regarded as overruled.” 243 U. S., at 518.
The broad ramifications of the Motion Picture case apparently were not immediately comprehended, and in a series of decisions over the next three decades litigants tested its limits. In Carbice Corp. v. American Patents Corp., 283 U. S. 27 (1931), the Court denied relief to a patentee who, through its sole licensee, authorized use of a patented design for a refrigeration package only to purchasers from the licensee of solid carbon dioxide (“dry ice”), a refrigerant that the licensee manufactured.11 The refrigerant was a well-known and widely used staple article of commerce, and the patent in question claimed neither a machine for making it nor a process for using it. Id., at 29. The Court held that the patent holder and its licensee were attempting to exclude *193competitors in the refrigerant business from a portion of the market, and that this conduct constituted patent misuse. It reasoned:
“Control over the supply of such unpatented material is beyond the scope of the patentee’s monopoly; and this limitation, inherent in the patent grant, is not dependent upon the peculiar function or character of the unpatented material or on the way in which it is used. Belief is denied because the {licensee] is attempting, without sanction of law, to employ the patent to secure a limited monopoly of unpatented material used in applying the invention.” Id., at 33-34.
The Court also rejected the patentee’s reliance on the Leeds & Gatlin decision. It found “no suggestion” in that case that the owner of the disc-stylus combination patent had attempted to derive profits from the sale of unpatented supplies as opposed to a patented invention. 283 U. S., at 34.
Other decisions of a similar import followed. Leitch Mfg. Co. v. Barber Co., 302 U. S. 458 (1938), found patent misuse in an attempt to exploit a process patent for the curing of cement through the sale of bituminous emulsion, an unpat-ented staple article of commerce used in the process. The Court eschewed an attempt to limit the rule of Carbice and Motion Picture to cases involving explicit agreements extending the patent monopoly, and it stated the broad proposition that “every use of a patent as a means of obtaining a limited monopoly of unpatented material is prohibited.” 302 U. S., at 463. Morton Salt Co. v. G. S. Suppiger Co., 314 U. S. 488, 492-494 (1942), which involved an attempt to control the market for salt tablets used in a patented dispenser, explicitly linked the doctrine of patent misuse to the “unclean hands” doctrine traditionally applied by courts of equity. Its companion case, B. B. Chemical Co. v. Ellis, 314 U. S. 495, 495-498 (1942), held that patent misuse barred relief even where infringement had been actively induced, and that practical *194difficulties in marketing a patented invention could not justify patent misuse.12
Although none of these decisions purported to cut back on the doctrine of contributory infringement itself, they were generally perceived as having that effect, and how far the developing doctrine of patent misuse might extend was a topic of some speculation among members of the patent bar. *195The Court’s decisions had not yet addressed the status of contributory infringement or patent misuse with respect to nonstaple goods, and some courts and commentators apparently took the view that control of nonstaple items capable only of infringing use might not bar patent protection against contributory infringement.13 This view soon received a serious, if not fatal, blow from the Court’s controversial decisions in Mercoid Corp. v. Mid-Continent Investment Co., 320 U. S. 661 (1944) (Mercoid I), and Mercoid Corp. v. Minneapolis-Honeywell Regulator Co., 320 U. S. 680 (1944) (Mercoid II). In these cases, the Court definitely held that any attempt to control the market for unpatented goods would constitute patent misuse, even if those goods had no use outside a patented invention. Because these cases served as the point of departure for congressional legislation, they merit more than passing citation.
Both cases involved a single patent that claimed a combination of elements for a furnace heating system. Mid-Continent was the owner of the patent, and Honeywell was its licensee. Although neither company made or installed the furnace system, Honeywell manufactured and sold stoker switches especially made for and essential to the system’s operation. The right to build and use the system was granted to purchasers of the stoker switches, and royalties owed the patentee were calculated on the number of stoker switches sold. Mercoid manufactured and marketed a competing stoker switch that was designed to be used only in the patented combination. Mercoid had been offered a sublicense *196by the licensee but had refused to take one. It was sued for contributory infringement by both the patentee and the licensee, and it raised patent misuse as a defense.
In Mercoid I the Court barred the patentee from obtaining relief because it deemed the licensing arrangement with Honeywell to be an unlawful attempt to extend the patent monopoly. The opinion for the Court painted with a very broad brush. Prior patent misuse decisions had involved attempts “to secure a partial monopoly in supplies consumed ... or unpatented materials employed” in connection with the practice of the invention. None, however, had involved an integral component necessary to the functioning of the patented system. 320 U. S., at 665. The Court refused, however, to infer any “difference in principle” from this distinction in fact. Ibid. Instead, it stated an expansive rule that apparently admitted no exception:
“The necessities or convenience of the patentee do not justify any use of the monopoly of the patent to create another monopoly. The fact that the patentee has the power to refuse a license does not enable him to enlarge the monopoly of the patent by the expedient of attaching conditions to its use. . . . The method by which the monopoly is sought to be extended is immaterial. . . . When the patentee ties something else to his invention, he acts only by virtue of his right as the owner of property to make contracts concerning it and not otherwise. He then is subject to all the limitations upon that right which the general law imposes upon such contracts. The contract is not saved by anything in the patent laws because it relates to the invention. If it were, the mere act of the patentee could make the distinctive claim of the patent attach to something which does not possess the quality of invention. Then the patent would be diverted from its statutory purpose and become a ready instrument for economic control in domains where the *197anti-trust acts or other laws not the patent statutes define the public policy.” Id., at 666.
The Court recognized that its reasoning directly conflicted with Leeds & Catlin Co. v. Victor Talking Machine Co., supra, and it registered disapproval, if not outright rejection, of that case. 320 U. S., at 668. It also recognized that “[t]he result of this decision, together with those which have preceded it, is to limit substantially the doctrine of contributory infringement.” Id., at 669. The Court commented, rather cryptically, that it would not “stop to consider” what “residuum” of the contributory infringement doctrine “may be left.” Ibid.
Mercoid II did not add much to the breathtaking sweep of its companion decision. The Court did reinforce, however, the conclusion that its ruling made no exception for elements essential to the inventive character of a patented combination. “However worthy it may be, however essential to the patent, an unpatented part of a combination patent is no more entitled to monopolistic protection than any other un-patented device.” 320 U. S., at 684.
What emerges from this review of judicial development is a fairly complicated picture, in which the rights and obligations of patentees as against contributory infringers have varied over time. We need not decide how respondent would have fared against a charge of patent misuse at any particular point prior to the enactment of 36 U. S. C. § 271. Nevertheless, certain inferences that are pertinent to the present inquiry may be drawn from these historical developments.
First, we agree with the Court of Appeals that the concepts of contributory infringement and patent misuse “rest on antithetical underpinnings.” 699 F. 2d, at 697. The traditional remedy against contributory infringement is the injunction. And an inevitable concomitant of the right to enjoin another from contributory infringement is the capacity to suppress competition in an unpatented article of commerce. See, e. g.. *198Thomson-Houston Electric Co. v. Kelsey Electric R. Specialty Co., 72 F. 1016, 1018-1019 (CC Conn. 1896). Proponents of contributory infringement defend this result on the grounds that it is necessary for the protection of the patent right, and that the market for the unpatented article flows from the pat-entee’s invention. They also observe that in many instances the article is “unpatented” only because of the technical rules of patent claiming, which require the placement of an invention in its context. Yet suppression of competition in unpat-ented goods is precisely what the opponents of patent misuse decry.14 If both the patent misuse and contributory infringement doctrines are to coexist, then, each must have some separate sphere of operation with which the other does not interfere.
Second, we find that the majority of eases in which the patent misuse doctrine was developed involved undoing the damage thought to have been done by A. B. Dick. The desire to extend patent protection to control of staple articles of commerce diqd slowly, and the ghost of the expansive contributory infringement era continued to haunt the courts. As a result, among the historical precedents in this Court, only the Leeds <& Catlin and Mercoid cases bear significant factual similarity to the present controversy. Those cases involved questions of control' over unpatented articles that were essential to the patented inventions, and that were unsuited for any commercial noninfringing use. In this case, we face similar questions in connection with a chemical, pro-*199panil, the herbicidal properties of which are essential to the advance on prior art disclosed by respondent’s patented process. Like the record disc in Leeds <fe Catlin or the stoker switch in the Mercoid cases, and unlike the dry ice in Carbice or the bituminous emulsion in Leitch, propanil is a nonstaple commodity which has no use except through practice of the patented method. Accordingly, had the present case arisen prior to Mercoid, we believe it fair to say that it would have fallen close to the wavering line between legitimate protection against contributory infringement and illegitimate patent misuse.
Ill
The Mercoid decisions left in their wake some .consternation among patent lawyers15 and a degree of confusion in the lower courts. Although some courts treated the Mercoid pronouncements as limited in effect to the specific kind of licensing arrangement at issue in those cases, others took a much more expansive view of the decision.16 Among the *200latter group, some courts held that even the filing of an action for contributory infringement, by threatening to deter competition in unpatented materials, could supply evidence of patent misuse. See, e. g., Stroco Products, Inc. v. Mullenbach, 67 USPQ 168, 170 (SD Cal. 1944). This state of affairs made it difficult for patent lawyers to advise their clients on questions of contributory infringement and to render secure opinions on the validity of proposed licensing arrangements. Certain segments of the patent bar eventually decided to ask Congress for corrective legislation that would restore some scope to the contributory infringement doctrine. With great perseverance, they advanced their proposal in three successive Congresses before it eventually was enacted in 1952 as 35 U. S. C. § 271.
A
The critical inquiry in this case is how the enactment of § 271 affected the doctrines of contributory infringement and patent misuse. Viewed against the backdrop of judicial precedent, we believe that the language and structure of the statute lend significant support to Rohm & Haas’ contention that, because § 271 (d) immunizes its conduct from the charge of patent misuse, it should not be barred from seeking relief. The approach that Congress took toward the codification of contributory infringement and patent misuse reveals a compromise between those two doctrines and their competing policies that permits patentees to exercise control over non-staple articles used in their inventions.
Section 271 (c) identifies the basic dividing line between contributory infringement and patent misuse. It adopts a restrictive definition of contributory infringement that distinguishes between staple and nonstaple articles of commerce. It also defines the class of nonstaple items narrowly. In essence, this provision places materials like the dry ice of the Carbice case outside the scope of the contributory infringement doctrine. As a result, it is no longer necessary to resort *201to the doctrine of patent misuse in order to deny patentees control over staple goods used in their inventions.
The limitations on contributory infringement written into § 271 (c) are counterbalanced by limitations on patent misuse in § 271 (d). Three species of conduct by patentees are expressly excluded from characterization as. misuse. First, the patentee may “deriv[e] revenue” from acts that “would constitute contributory infringement” if “performed by another without his consent.” This provision clearly signifies that a patentee may make and sell nonstaple goods used in connection with his invention. Second, the patentee may “licens[e] or authoriz[e] another to perform acts” which without such authorization would constitute contributory infringement. This provision’s use in the disjunctive of the term “author-iz[e]” suggests that more than explicit licensing agreements is contemplated. Finally, the patentee may “enforce his patent rights against. . . contributory infringement.” This provision plainly means that the patentee may bring suit without fear that his doing so will be regarded as an unlawful attempt to suppress competition. The statute explicitly states that a patentee may do “one or more” of these permitted acts, and it does not state that he must do any of them.
In our view, the provisions of § 271 (d) effectively confer upon the patentee, as a lawful adjunct of his patent rights, a limited power to exclude others from competition in non-staple goods. A patentee may sell a nonstaple article himself while enjoining others from marketing that same good without his authorization. By doing so, he is able to eliminate competitors and thereby to control the market for that product. Moreover, his power to demand royalties from others for the privilege of selling the nonstaple item itself implies that the patentee may control the market for the nonstaple good; otherwise, his “right” to sell licenses for the marketing of the nonstaple good would be meaningless, since no one would be willing to pay him for a superfluous authorization. See Note, 70 Yale. L. J. 649, 659 (1961).
*202Rohm & Haas’ conduct is not dissimilar in either nature or effect from the conduct that is thus clearly embraced within § 271 (d). It sells propanil; it authorizes others to use pro-panil; and it sues contributory infringers. These are all protected activities. Rohm & Haas does not license others to sell propanil, but nothing on the face of the statute requires it to do so. To be sure, the sum effect of Rohm & Haas’ actions is to suppress competition in the market for an un-patented commodity. But as we have observed, in this its conduct is no different from that which the statute expressly protects.
The one aspect of Rohm & Haas’ behavior that is not expressly covered by § 271 (d) is its linkage of two protected activities — sale of propanil and authorization to practice the patented process — together in a single transaction. Petitioners vigorously argue that this linkage, which they characterize pejoratively as “tying,” supplies the otherwise missing element of misuse. They fail, however, to identify any way in which this “tying” of two expressly protected activities results in any extension of control over unpatented materials beyond what § 271 (d) already allows. Nevertheless, the language of § 271 (d) does not explicitly resolve the question when linkage of this variety becomes patent misuse. In order to judge whether this method of exploiting the patent lies within or without the protection afforded by § 271 (d), we must turn to the legislative history.
B
Petitioners argue that the legislative materials indicate at most a modest purpose for § 271. Relying mainly on the Committee Reports that accompanied the “Act to Revise and Codify the Patent Laws” (1952 Act), 66 Stat. 792, of which § 271 was a part, petitioners assert that the principal purpose of Congress was to “clarify” the law of contributory infringement as it had been developed by the courts, rather than to effect any significant substantive change. They note that *203the 1952 Act undertook the major task of codifying all the patent laws in a single title, and they argue that substantive changes from recodifications are not lightly to be inferred. See United States v. Ryder, 110 U. S. 729, 739-740 (1884). They further argue that, whatever the impact of § 271 in other respects, there is not the kind of “clear and certain signal from Congress” that should be required, for an extension of patent privileges. See Deepsouth Packing Co. v. Laitram Corp., 406 U. S. 518, 531 (1972). We disagree with petitioners’ assessment. In our view, the relevant legislative materials abundantly demonstrate an intent both to change the law and to expand significantly the ability of patentees to protect their rights against contributory infringement.
The 1952 Act was approved with virtually no floor debate. Only one exchange is relevant to the present inquiry. In response to a question whether the Act would effect any substantive changes, Senator McCarran, a spokesman for the legislation, commented that the Act “codif[ies] the patent laws.” 98 Cong. Rec. 9323 (1952). He also submitted a statement, which explained that, although the general purpose of the Act was to clarify existing law, it also included several changes taken “[i]n view of decisions of the Supreme Court and others.” Ibid. Perhaps because of the magnitude of the recodification effort, the Committee Reports accompanying the 1952 Act also gave relatively cursory attention to its features. Nevertheless, they did identify § 271 as one of the “major changes or innovations in the title.” H. R. Rep. No. 1923, 82d Cong., 2d Sess., 5 (1952).17 In explaining the provisions of § 271, the Reports stated that they were intended “to codify in statutory form the principles of contributory infringement and at the same time [to] eliminate . . . doubt and confusion” that had resulted from “decisions of the courts *204in recent years.” Id., at 9. The Reports also commented that §§ 271 (b), (c), and (d) “have as their main purpose clarification and stabilization.” Ibid.
These materials sufficiently demonstrate that the 1952 Act did include significant substantive changes, and that § 271 was one of them.
The principal sources for edification concerning the meaning and scope of § 271, however, are the extensive hearings that were held on the legislative proposals that led up to the final enactment. In three sets of hearings over the course of four years, proponents and opponents of the legislation debated its impact and relationship with prior law. Draftsmen of the legislation contended for a restriction on the doctrine of patent misuse that would enable patentees to protect themselves against contributory infringers. Others, including representatives of the Department of Justice, vigorously opposed such a restriction.
Although the final version of the statute reflects some minor changes from earlier drafts, the essence of the legislation remained constant. References were made in the later hearings to testimony in the earlier ones.18 Accordingly, we regard each set of hearings as relevant to a full understanding of the final legislative product. Cf., e. g., Schwegmann Bros. v. Calvert Distillers Corp., 341 U. S. 384, 390 (1951); Transcontinental & Western Air, Inc. v. CAB, 336 U. S. 601, 605-606, n. 6 (1949). Together, they strongly reinforce the conclusion that § 271 (d) was designed to immunize from the charge of patent misuse behavior similar to that in which the respondent has engaged.
1. The 19I¡.8 Hearings. The first bill underlying § 271 was H. R. 5988, proposed to the 80th Congress. During the hearings on this bill its origin and purpose were carefully ex*205plained. The New York Patent Law Association, which had supervised drafting of the legislation, submitted a prepared memorandum that candidly declared that the purpose of the proposal was to reverse the trend of Supreme Court decisions that indirectly had cut back on the contributory infringement doctrine. Hearings on H. R. 5988, etc., before the Subcommittee on Patents, Trade-Marks, and Copyrights of the House Committee on the Judiciary, 80th Cong., 2d Sess., 4 (1948) (1948 Hearings). The memorandum explained the rationale behind contributory infringement, and it gave as one example of its proper application the protection of a patent for use of a chemical:
“[0]ne who supplies a hitherto unused chemical to the public for use in a new method is stealing the benefit of the discovery of the property of this chemical which made the new method possible. To enjoin him from distributing the chemical for use in the new method does not prevent him from doing anything which he could do before the new property of the chemical had been discovered.” Ibid.
It criticized several decisions, including Leitch and Carbice as well as the two Mercoids, on the ground that together they had effectively excluded such “new-use inventions” from the protections of the patent law. 1948 Hearings, at 4-5. It went on to explain that the proposed legislation was designed to counteract this effect by providing that “the mere use or enforcement of the right to be protected against contributory infringement . . . shall not be regarded as misuse of the patent.” Id., at 6. This approach, the memorandum stated, “does away with the ground on which the Supreme Court has destroyed the doctrine of contributory infringement” and “is essential to make the rights against contributory infringers which are revived by the statute practically useful and enforceable.” Ibid.
Testimony by proponents of the bill developed the same *206theme. Giles Rich, then a prominent patent lawyer, was one of the draftsmen. He highlighted the tension between the judicial doctrines of contributory infringement and patent misuse. He stated that early patent misuse decisions “seem to us now to have been just,” but that “this doctrine has been carried too far — so far that it . . . has practically eliminated from the law the doctrine of contributory infringement as a useful legal doctrine.” Id., at 9. To illustrate this point, he contrasted the Carbice and Mercoid cases, and noted that the latter had involved an item without any noninfringing use. Because it incorporated a staple-nonstaple distinction in the definition of contributory infringement, Mr. Rich argued that the bill would “correct [the] situation” left by Mercoid “without giving sanction to practices such as those in the Carbice case.” 1948 Hearings, at 11.
Rich’s testimony was followed by that of Robert W. Byerly, another draftsman. He stressed the confusion in which the Mercoid decisions had left the lower courts, and the need for Congress to define the scope of protection against contributory infringement by drawing a clear line between deliberate taking of another’s invention and legitimate trade in staple articles of commerce. Id., at 13-16. Byerly discussed the practical difficulties some patentees would encounter if suits against direct infringers were their only option to protect against infringement. Id., at 13-14. He argued that the breadth of the Court’s misuse decision in Mercoid I could be discerned from the fact that it “overruled” Leeds & Catlin. 1948 Hearings, at 14. He explained the section of the bill restricting the scope of patent misuse as intended to give the patentee recourse to either or both of two options: “A man can either say, ‘you cannot sell the part of my invention to somebody else to complete it,’ or he can say, ‘yes, you can sell the part of my invention to help others complete it provided you pay me a royalty.’ ” Id., at 16.
The bill attracted opponents as well, some of whom de*207fended the result of the Mercoid decisions.19 In addition, Roy C. Hackley, Jr., Chief of the Patent Section, Department of Justice, made an appearance on behalf of the Department. He took the position that statutory clarification of the scope of contributory infringement was desirable, but he warned Congress against using language that might “permit illegal extension of the patent monopoly.” 1948 Hearings, at 69. On this ground he opposed the portion of the proposed bill that included language substantially similar to what is now § 271 (d). Ibid.
2. The 19JfB Hearings. The 1948 bill did not come to a vote, but the patent bar resubmitted its proposal in 1949. Again, there were fairly extensive hearings, with debate, and again Rich led the list of favorable witnesses. He renewed his attempt to explain the legislation in terms of past decisions of this Court. The result in the- Carbice case, he argued, was proper because the patentee had tried to interfere with the market in an old and widely used product. On the other hand, he cited the Mercoid cases as examples of a situation where “[t]here is no practical way to enforce that patent, except through a suit for contributory infringement against the party who makes the thing which is essentially the inventive subject matter [and] which, when put into use, creates infringement.” Hearings on H. R. 3866 before Subcommittee No. 4 of the House Committee on the Judiciary, 81st Cong., 1st Sess., 11 (1949) (1949 Hearings).
To restore the doctrine of contributory infringement where it was most needed, Rich'argued, it was essential to restrict pro tanto the judicially created doctrine of patent misuse:
“I would like to recall that we are dealing with a problem which involves a conflict between two doctrines, contributory infringement and misuse.
*208“It is crystal clear, when you have thoroughly studied this subject, that the only way you can make contributory infringement operative again as a doctrine, is to make some exceptions to the misuse doctrine and say that certain acts shall not be misuse. Then contributory infringement, which is there all the time, becomes operative again.
“Contributory infringement has been destroyed by the misuse doctrine; and to revive it you do not have to do anything with contributory infringement itself. You go back along the same road until you get to the point where you have contributory infringement working for you again.” Id., at 13-14.
Rich warned against going too far. He took the position that a law designed to reinstate the broad contributory infringement reasoning of Henry v. A. B. Dick Co., 224 U. S. 1 (1912), “would kill itself in time.” 1949 Hearings, at 17. The proposed legislation, however, “stopped short of that” and “said that you can control only things like the switches in the Mercoid case, which are especially made or adapted for use in connection with such patent and which are not suitable for actual, commercial, noninfringing use.” Ibid.
In the 1949 Hearings, the Department of Justice pressed more vigorous opposition to the contributory infringement proposal than it had in 1948. Represented by John C. Sted-man, Chief, Legislation and Clearance Section, Antitrust Division, the Department argued that legislation was unnecessary because the Mercoid decisions were correct, because they had not produced as much confusion as the proponents of the new legislation claimed, and because the legislation would produce new interpretive problems. 1949 Hearings, at 50-56. Stedman defended the result of the Mercoid decisions on the ground that marketing techniques employed in those cases were indistinguishable in effect from tying schemes previously considered by the Court. He took the view that the staple-*209nonstaple distinction should be irrelevant for purposes of patent misuse. “If the owner of the patent is using his patent in a way to prevent the sale of unpatented elements, then the misuse doctrine would apply.” 1949 Hearings, at 54. Sted-man added that the effect of the legislation would be to revive the Leeds & Catlin decision, a result the Department of Justice opposed. 1949 Hearings, at 59. Later in the hearings, he offered several methods of exploiting patent rights that arguably would eliminate the need for the contributory infringement doctrine, and he stated that a suit for contributory infringement could involve patent misuse, even if there were no conditional licensing of patent rights. Id., at 76-77.
After Stedman’s opening testimony, Rich was recalled for further questioning. Rich agreed with Stedman’s assessment of the effect that the legislation would have, but argued that the Justice Department’s arguments ignored the bill’s limitation of contributory infringement to nonstaple articles. To clarify the effect of the statute, Rich declared:
“[I]t- is absolutely necessary, to get anywhere in the direction we are trying to go, to make some exception to the misuse doctrine because it is the conflict between the doctrine of contributory infringement and the doctrine of misuse that raises the problem.” Id., at 67.
He added:
“The exception which we wish to make to the misuse doctrine would reverse the result in the Mercoid case; it would not reverse the result in the Carbice case.” Ibid.
In response to questioning, Rich agreed that the bill would preserve both the contributory infringement and misuse doctrines as they had existed in this Court’s cases prior to the Mercoid decisions. 1949 Hearings, at 68. He asserted that the method by which the patentee’s invention was exploited in Mercoid was necessary given the nature of the businesses involved. 1949 Hearings, at 69. When asked whether the *210proposed legislation would allow that kind of licensing activity, Rich responded with an unqualified “Yes.” Ibid.
3. The 1951 Hearings. By the time the proposal for a statutory law of contributory infringement and patent misuse was presented to the 82d Congress, the battle lines of the earlier hearings had solidified substantially, and the representatives of the patent bar once again found themselves faced with the formidable opposition of the Department of Justice.
In his opening remarks before the 1951 Hearings, Rich reminded the congressional Subcommittee that, as a practical matter, it was necessary to deal with the contributory infringement and the misuse doctrines as a unit “if we are to tackle the problem at all.” 1951 Hearings, at 152. He urged on the Subcommittee the need to eliminate confusion in the law left by the Mercoid decisions by drawing a “sensible line” between contributory infringement and patent misuse that would be “in accordance with public policy as it seems to exist today.” 1951 Hearings, at 152. Rich also attempted to play down the controversiality of the proposal by arguing that a restrictive definition of contributory infringement had been incorporated into the bill. Id., at 153-154.
When questioned about the effect of the bill on present law, Rich replied that it would not extend the contributory infringement doctrine unless “you take the point of view that there is no such things [sic] as contributory infringement today.” Id., at 158. He rejected the suggestion that the legislation would return the law of contributory infringement to the A. B. Dick era, and he reminded the Subcommittee that the law “would not touch the result of the Carbice decision.” 1951 Hearings, at 161. Rich concluded his opening testimony with this explanation of subsection (d):
“It deals with the misuse doctrine, and the reason it is necessary is that the Supreme Court has made it abundantly clear that there exist in the law today two doc*211trines, contributory infringement on the one hand, and misuse on the other, and that, where there is a conflict, the misuse doctrine must prevail because of the public interest inherently involved in patent cases.
“Other decisions following Mercoid have made it quite clear that at least some courts are going to say that any effort whatever to enforce a patent against a contributory infringer is in itself misuse. . . . Therefore we have always felt — we who study this subject particularly — that to put any measure of contributory infringement into law you must, to that extent and to that extent only, specifically make exceptions to the misuse doctrine, and that is the purpose of paragraph (d).
“It goes with, supports, and depends upon paragraph (c).” Id., at 161-162.
The Department of Justice, now represented by Wilbur L. Fugate of the Antitrust Division, broadly objected to “writing the doctrine of contributory infringement into the law.” Id., at 165. Its most strenuous opposition was directed at what was to become § 271 (d). Fugate warned that this provision “would have the effect of wiping out a good deal of the law relating to misuse of patents, particularly with reference to tying-in clauses.” Ibid. He repeatedly asserted that the language of subsection (d) was unclear, and that it was im-possiblé to tell how far it would serve to insulate patentees from charges of misuse. See id., at 167-169. But as the Department construed it, the subsection would “seriously impair the doctrine of misuse of patents in favor of the doctrine of contributory infringements.” Id., at 168. Fugate would not say that any of the three acts protected by subsection (d) were per se illegal, but he felt that they could become evidence of misuse in some contexts. Id., at 168-169.
When Representative Crumpaeker challenged Fugate’s interpretation of the statute, Fugate replied that Rich had advanced the same construction, and he called upon Rich to *212say whether he agreed. Id., at 169. The following colloquy then took place:
“Mr. RICH: I will agree with [Mr. Fugate’s interpretation] to this extent: That as I testified it is necessary to make an exception to misuse to the extent that you revive contributory infringement in paragraph (c), and this whole section (d) is entirely dependent on (c). Where (d) refers to contributory infringement, it only refers to contributory infringement as defined in (c) and nothing more.
“Mr. CRUMPACKER: In other words, all it says is that bringing an action against someone who is guilty of contributory infringement is not a misuse of the patent.
“Mr. RICH: That is true.” Ibid.
Rich and Fugate then discussed the law in the courts before and after the Mercoid decisions. In an effort to clarify the intendment of the statute, Congressman Rogers asked Rich to identify misuse decisions exemplifying the acts specified in the three parts of subsection (d). Rich identified the Leitch and Carbice cases as examples of situations where deriving revenue from acts that would be contributory infringement was held to be evidence of misuse; he stated that the Mercoid cases exemplified misuse from licensing others; and he referred to Stroco Products, Inc. v. Mullenbach, supra, as an example of a case where the mere bringing of an action against contributory infringers was found to exemplify misuse. 1951 Hearings, at 174-175. He again reminded the Subcommittee that the scope of subsection (d) was implicitly limited by the restrictive definition of contributory infringement in subsection (c), and he assured the Subcommittee that “[i]f [a pat-entee] has gone beyond those and done other acts which could be misuse, then the misuse doctrine would be applicable.” Id., at 175. As an example of such'“other acts,” he suggested that a patentee would be guilty of misuse if he tried to license others to produce staple articles used in a patented invention. Ibid.
*213c
Other legislative materials that we have not discussed bear as well on the meaning to be assigned to § 271 (d); but the materials that we have culled are exemplary, and they amply demonstrate the intended scope of the statute. It is the consistent theme of the legislative history that the statute was designed to accomplish a good deal more than mere clarification. It significantly changed existing law, and the change moved in the direction of expanding the statutory protection enjoyed by patentees. The responsible congressional Committees were told again and again that contributory infringement would wither away if the misuse rationale of the Mer-coid decisions remained as a barrier to enforcement of the patentee’s rights. They were told that this was an undesirable result that would deprive many patent holders of effective protection for their patent rights. They were told that Congress could strike a sensible compromise between the competing doctrines of contributory infringement and patent misuse if it eliminated the result of the Mercoid decisions yet preserved the result in Carbice. And they were told that the proposed legislation would achieve this effect by restricting contributory infringement to the sphere of nonstaple goods while exempting the control of such goods from the scope of patent misuse. These signals cannot be ignored. They fully support the conclusion that, by enacting §§271 (c) and (d), Congress granted to patent holders a statutory right to control nonstaple goods that are capable only of infringing use in a patented invention, and that are essential to that invention’s advance over prior art.
We find nothing in this legislative history to support the assertion that respondent’s behavior falls outside the scope of §271 (d).20 To the contrary, respondent has done nothing *214that would extend its right of control over unpatented goods beyond the line that Congress drew. Respondent, to be sure, has licensed use of its patented process only in connection with purchases of propanil. But propanil is a nonstaple product, and its herbicidal property is the heart of respondent’s invention. Respondent’s method of doing business is thus essentially the same as the method condemned in the Mercoid decisions, and the legislative history reveals that § 271 (d) was designed to retreat from Mercoid in this regard.
There is one factual difference between this case and Mer-coid: the licensee in the Mercoid cases had offered a sub-license to the alleged contributory infringer, which offer had been refused. Mercoid II, 320 U. S., at 683. Seizing upon this difference, petitioners argue that respondent’s unwillingness to offer similar licenses to its would-be competitors in the manufacture of propanil legally distinguishes this case and sets it outside §271 (d). To this argument, there are at *215least three responses. First, as we have noted, § 271 (d) permits such licensing but does not require it. Accordingly, petitioners’ suggestion would import into the statute a requirement that simply is not there. Second, petitioners have failed to adduce any evidence from the legislative history that the offering of a license to the alleged contributory in-fringer was a critical factor in inducing Congress to retreat from the result of the Mercoid decisions. Indeed, the Leeds & Catlin decision, which did not involve such an offer to license, was placed before Congress as an example of the kind of contributory infringement action the statute would allow. Third, petitioners’ argument runs contrary to the long-settled view that the essence of a patent grant is the right to exclude others from profiting by the patented invention. 35 U. S. C. § 154; see Continental Paper Bag Co. v. Eastern Paper Bag Co., 210 U. S. 405, 424-425 (1908); Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U. S. 100, 135 (1969). If petitioners’ argument were accepted, it would force pat-entees either to grant licenses or to forfeit their statutory protection against contributory infringement. Compulsory licensing is a rarity in our patent system,21 and we decline to manufacture such a requirement out of § 271 (d).
IV
Petitioners argue, finally, that the interpretation of § 271 (d) which we have adopted is foreclosed by decisions of this *216Court following the passage of the 1952 Act. They assert that in subsequent cases the Court has continued to rely upon the Mercoid decisions, and that it has effectively construed § 271 (d) to codify the result of those decisions, rather than to return the doctrine of patent misuse to some earlier stage of development. We disagree.
The cases to which petitioners turn for this argument include some that have cited the Mercoid decisions as evidence of a general judicial “hostility to use of the statutorily granted patent monopoly to extend the patentee’s economic control to unpatented products.” United States v. Loew’s, Inc., 371 U. S. 38, 46 (1962); see also Blonder-Tongue Laboratories, Inc. v. University of Illinois Foundation, 402 U. S. 313, 343-344 (1971).. These decisions were not directly concerned with the doctrine of contributory infringement, and they did not require the Court to evaluate § 271 (d) or its impact on the holdings in Mercoid. Like other cases that do not specifically mention those decisions, see, e. g., Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U. S., at 136, they state the general thrust of the doctrine of patent misuse without attending to its specific statutory limitations.
In another case, Deepsouth Packing Co. v. Laitram Corp., 406 U. S. 518 (1972), the Court dealt only with the scope of direct infringement under §271 (a). The question under consideration was whether a patent is infringed when unpat-ented elements are assembled into the combination outside the United States. The Court held that such assembly would not have constituted direct infringement prior to the enactment of §271 (a), and it concluded that enactment of the statute effected no change in that regard. The Court cited Mercoid I for the well-established proposition that unless there has been direct infringement there can be no contributory infringement. 406 U. S., at 526. Again, the Court did not have occasion to focus on the meaning of § 271 (d).
The only two decisions that touch at all closely upon the issues of statutory construction presented here are Aro Mfg. *217Co. v. Convertible Top Co., 365 U. S. 336 (1961) (Aro I), and Aro Mfg. Co. v. Convertible Top Co., 377 U. S. 476 (1964) (Aro II). These decisions emerged from a single case involving an action for contributory infringement based on the manufacture and sale of a specially cut fabric designed for use in a patented automobile convertible top combination. In neither case, however, did the Court directly address the question of § 271 (d)’s effect on the law of patent misuse.
The controlling issue in Aro I was whether there had been any direct infringement of the patent. The Court held that purchasers of the specially cut fabric used it for “repair” rather than “reconstruction” of the patented combination; accordingly, under the patent law they were not guilty of infringement. 365 U. S., at 340, 346. Since there was no direct infringement by the purchasers, the Court held that there could be no contributory infringement by the manufacturer of the replacement tops. This conclusion rested in part on a holding that § 271 (c) “made no change in the fundamental precept that there can be no contributory infringement in the absence of a direct infringement.” Id., at 341. It in no way conflicts with our decision.
As petitioners observe, Aro I does quote certain passages from the Mercoid decisions standing for the proposition that even single elements constituting the heart of a patented combination are not within the scope of the patent grant. 365 U. S., at 345. In context, these references to Mercoid are not inconsonant with our view of § 271(d). In the course of its decision, the Court eschewed the suggestion that the legal distinction between “reconstruction” and “repair” should be affected by whether the element of the combination that has been replaced is an “essential” or “distinguishing” part of the invention. 365 U. S., at 344. The Court reasoned that such a standard would “ascrib[e] to one element of the patented combination the status of patented invention in itself,” and it drew from the Mercoid cases only to the extent that they described limitations on the scope of the patent *218grant. 365 U. S., at 344-345. In a footnote, the Court carefully avoided reliance on the misuse aspect of those decisions. Id., at 344, n. 10. Accordingly, it had no occasion to consider whether or to what degree § 271 (d) undermined the validity of the Mercoid patent misuse rationale.22
Aro II is a complicated decision in which the Court mustered different majorities in support of various aspects of its opinion. See 377 U. S., at 488, n. 8. After remand from Aro I, it became clear that the Court’s decision in that case had not eliminated all possible grounds for a charge of contributory infringement. Certain convertible top combinations had been sold without valid license from the patentee. Because use of these tops involved direct infringement of the patent, there remained a question whether fabric supplied for their repair might constitute contributory infringement notwithstanding the Court’s earlier decision.
Aro II decided several questions of statutory interpretation under § 271. First, it held that repair of an unlicensed combination was direct infringement under the law preceding enactment of § 271, and that the statute did not effect .any change in this regard. 377 U. S., at 484. Like the constructions of § 271 (a) in Aro I and Deepsouth Packing Co., this conclusion concerns a statutory provision not at issue in this case.
Second, the Court held that supplying replacement fabrics specially cut for use in the infringing repair constituted contributory infringement under § 271 (c). The Court held that the specially cut fabrics, when installed in infringing equipment, qualified as nonstaple items within .the language of § 271 (c), and that supply of similar materials for infringing repair had been treated as contributory infringement under the judicial law that § 271 (c) was designed to codify. 377 *219IT. S., at 485-488. It also held that § 271 (c) requires a showing that an alleged contributory infringer knew that the combination for which his component was especially designed was both patented and infringing. 377 U. S., at 488-491. We regard these holdings as fully consistent with our understanding of §271 (c). In any event, since petitioners have conceded contributory infringement for the purposes of this decision, the scope of that subsection is not directly before us.
Third, the Court held that the alleged contributory in-fringer could not avoid liability by reliance on the doctrine of the Mercoid decisions. Although those decisions had cast contributory infringement into some doubt, the Court held that § 271 was enacted “for the express purpose ... of overruling any blanket invalidation of the [contributory infringement] doctrine that could be found in the Mercoid opinions.” 377 U. S., at 492. Although our review of the legislative history finds a broader intent, it is not out of harmony with Aro IPs analysis. The Court explicitly noted that a defense of patent misuse had not been pressed. Id., at 491. Accordingly, its discussion of legislative history was limited to those materials supporting the observation, sufficient for purposes of the case, that any direct attack on the contributory infringement doctrine in its entirety would be contrary to the manifest purpose of §271 (c). Since the Court in Aro II was not faced with a patent misuse defense, it had no occasion to consider other evidence in the hearings relating to the scope of § 271 (d).
Finally, in a segment of the Court's opinion that commanded full adherence of only four Justices, 377 U. S., at 493-500, it was stated that an agreement in which the pat-entee had released some purchasers of infringing combinations from liability defeated liability for contributory infringement with respect to replacement of convertible tops after the agreement went into effect. The plurality rejected the patentee’s attempt to condition its release by reserving *220“rights in connection with future sales of replacement fabrics.” Id., at 496. It relied on the Carbice and Mercoid decisions, as well as United States v. Loew’s, Inc., supra, for the proposition that a patentee “cannot impose conditions concerning the unpatented supplies, ancillary materials, or components with which the use [of a patented combination] is to be effected.” 377 U. S., at 497. This statement is qualified by the circumstances to which it applied. Because the Court already had determined in Aro I that replacement of wornout convertible top fabric constituted a permissible repair of the combination, the agreement sought to control an unpatented article in the context of a noninfring-ing use. The determination that the agreement defeated liability does not reflect resort to the principles of patent misuse; rather it betokens a recognition that the patentee, once it had authorized use of the combination, could not manufacture contributory infringement by contract where under the law there was none.
Perhaps the quintessential difference between the Aro decisions and the present case is the difference between the primary-use market for a chemical process and the replacement market out of which the Aro litigation arose. The repair-reconstruction distinction and its legal consequences are determinative in the latter context, but are not controlling here. Instead, the staple-nonstaple distinction, which Aro I found irrelevant to the characterization of replacements, supplies the controlling benchmark. This distinction ensures that the patentee’s right to prevent others from contributorily infringing his patent affects only the market for the invention itself. Because of this significant difference in legal context, we believe our interpretation of § 271 (d) does not conflict with these decisions.
V
Since our present task is one of statutory construction, questions of public policy cannot be determinative of the outcome *221unless specific policy choices fairly can be attributed to Congress itself. In this instance, as we have already stated, Congress chose a compromise between competing policy interests. The policy of free competition runs deep in our law. It underlies both the doctrine of patent misuse and the general principle that the boundary of a patent monopoly is to be limited by the literal scope of the patent claims. But the policy of stimulating invention that underlies the entire patent system runs no less deep. And the doctrine of contributory infringement, which has been called “an expression both of law and morals,” Mercoid I, 320 U. S., at 677 (Frankfurter, J., dissenting), can be of crucial importance in ensuring that the endeavors and investments of the inventor do not go unrewarded.
It is, perhaps, noteworthy that holders of “new use” patents on chemical processes were among those designated to Congress as intended beneficiaries of the protection against contributory infringement that § 271 was designed to restore. See 1948 Hearings, at 4, 5, 18. We have been informed that the characteristics of practical chemical research are such that this form of patent protection is particularly important to inventors in that field. The number of chemicals either known to scientists or disclosed by existing research is vast. It grows constantly, as those engaging in “pure” research publish their discoveries.23 The number of these chemicals that have known uses of commercial or social value, in contrast, is small. Development of new uses for existing chemicals is thus a major component of practical chemical research. *222It is extraordinarily expensive.24 It may take years of unsuccessful testing before a chemical having a desired property is identified, and it may take several years of further testing before a proper and safe method for using that chemical is developed.25
Under the construction of § 271 (d) that petitioners advance, the rewards available to those willing to undergo the time, expense, and interim frustration of such practical research would provide at best a dubious incentive. Others could await the results of the testing and then jump on the profit bandwagon by demanding licenses to sell the unpat-ented, nonstaple chemical used in the newly developed process. Refusal to accede to such a demand, if accompanied by any attempt to profit from the invention through sale of the unpatented chemical, would risk forfeiture of any patent protection whatsoever on a finding of patent misuse. As a result, noninventors would be almost assured of an opportunity to share in the spoils, even though they had contributed nothing to the discovery. The incentive to await the discoveries of others might well prove sweeter than the incentive to take the initiative oneself.
*223Whether such a regime would prove workable, as petitioners urge, or would lead to dire consequences, as respondent and several amici insist, we need not predict. Nor do we need to determine whether the principles of free competition could justify such a result. Congress’ enactment of § 271 (d) resolved these issues in favor of a broader scope of patent protection. In accord with our understanding of that statute, we hold that Rohm & Haas has not engaged in patent misuse, either by its method of selling propanil, or by its refusal to license others to sell that commodity. The judgment of the Court of Appeals is therefore affirmed.
It is so ordered.