CORPORATE DISCLOSURE STATEMENT The following chart sets forth, with respect to those plaintiffs to whom Rule 26.1 of the Federal Rules of Appellate Procedure is applicable, any parent companies, subsidiaries, affiliates or successors in interest that have issued shares to the public. Parent Companies, Subsidiaries, Affiliates or Successors in Interest That Have Issued Plaintiff Shares to the Public Academic Press General Cinema Corporation; The Neiman-Marcus Group, Inc. Alan R. Liss, Inc. John Wiley & Sons, Inc. Aspen Publishers, Inc. Wolters Kluwer N.V. Birkhauser Boston Springer Verlag New York, Inc. Butterworth & Co. Elsevier N.V., Amsterdam; (Publishers) Ltd. Reed International P.L.C., London CRC Press, Inc. The Times Mirror Company The C.V. Mosby The Times Mirror Company Company Elsevier Science Elsevier N.V., Amsterdam; Publishing Co., Inc. Reed International P.L.C., London Grune & Stratton, Inc. W.B. Saunders Company; General Cinema Corporation; The Neiman-Marcus Group, Inc. Harcourt Brace General Cinema Corporation; Jovanovich Limited The Neiman-Marcus Group, Inc. Human Sciences Press Plenum Publishing Corp. J. Schweitzer Verlag H. Luchterhand Verlag GmbH & Co.; Wolters Kluwer N.V., Netherlands John Wiley & Sons John Wiley & Sons, Inc.; Limited Wiley Heyden Ltd. Little, Brown & Time, Inc.; Time Warner, Inc. Company Pergamon Press, Ltd. Elsevier N.V., Amsterdam; Reed International P.L.C., London Raven Press Wolters Kluwer N.V. Scripta Technica, Inc. John Wiley & Sons, Inc. Van Nostrand Reinhold Thomson Publishing Corporation VCH Publishers, Inc. VCH Verlagsgesellschaft GmbH, Weinheim, Germany VNU Business Pittway Corp. Publications Inc. Walter De Gruyter, Inc. Walter De Gruyter & Company, Berlin, Germany Warren Gorham & Lamont Research Institute of America, Inc.; The Thomson Publishing Corporation Williams & Wilkins Waverly, Inc. Wolters Kluwer Wolters Kluwer N.V. Academic Publishers B.V. Table of Contents Page Corporate Disclosure Statement . . . . . . . . . . . i Table of Authorities . . . . . . . . . . . . . . . . viii Guide to Abbreviations . . . . . . . . . . . . . . xviii PRELIMINARY STATEMENT. . . . . . . . . . . . . . . . 1 QUESTIONS PRESENTED AND QUESTIONS NOT PRESENTED. . . . . . . . . . . . . . . 5 STATEMENT OF THE CASE. . . . . . . . . . . . . . . . 7 A. The Litigants and Interested Third Parties 7 B. The Fair Use Proceedings and Trial Record 8 C. The Decisions Below . . . . . . . . . . . 9 STATEMENT OF FACTS . . . . . . . . . . . . . . . . . 10 A. Texaco Engages In Its Research Activities (An Integral Part of Which Is the Systematic Photocopying of Journal Articles) for the Sole Purpose of Making a Profit. . . . . . . . 10 B. The Vital Economic Importance to Texaco of Its Researchers Having Timely Access to Journal Articles. . . . . . . . . . . . . 11 C. The Methods by Which Research Personnel at Texaco and Other For-Profit Corporations Regularly Learn About and Obtain Photocopies of Entire Journal Articles. . 12 Circulation of Journals. . . . . . . 12 Literature Searches. . . . . . . . . 13 Other Methods. . . . . . . . . . . . 13 Library Copying. . . . . . . . . . . 14 D. The Importance to Research Personnel of Speedy Circulation of Journals and How Photocopying Reduces the Number of Subscriptions Needed to Maintain Speedy Circulation . . . . . . . . . . . . . . . 14 E. Dr. Chickering's Representative Photocopying. . . . . . . . . . . . . . . 16 F. The Economic Benefits to Texaco and Other For-Profit Industrial Corporations of Research Scientists Using Photocopy Duplicates in Their Work . . . . . . . . . . . . . . 17 G. There Are a Number of Readily Available Ways to Provide Researchers with Access to Journal Articles Other Than Unauthorized Photocopying. . . . . . . . . . . . . . . 18 H. Unauthorized Photocopies Supersede Original Articles and Authorized Photocopies . . . 19 I. Academic Press and Its Important Role in the Dissemination of Scientific Information . 21 J. The Economic and Other Benefits to Scientist Authors From Having Articles Published in Journals. . . . . . . . . . . . . . . . . 24 K. The Copyright Clearance Center and the Success of Its Photocopying Authorization Services. . . . . . . . . . . . . . . . . 25 1. The Successful Development of CCC's Photocopying Authorization Services. 25 2. The Substantial and Growing Photocopying Authorizations Market . 30 L. Bell Labs' Successful Program Since 1978 of Providing Journal Articles to its Research Personnel Without Unauthorized Photocopying. . . . . . . . . . . . . . . 31 M. Texaco's Unauthorized Photocopying Is Not a Reasonable and Customary Practice . . . . 32 N. The Adverse Economic Effects on Copyright Owners If the Widespread Photocopying Practices Exemplified by Texaco Were Permitted as Fair Use . . . . . . . . . . 34 ARGUMENT . . . . . . . . . . . . . . . . . . . . . . 36 I. THE PROPER STANDARD OF APPELLATE REVIEW AND TEXACO'S ERRONEOUS VERSION OF THE FACTS . . . . . . . . . . . . . 36 A. The Proper Standard of Appellate Review Is Customary Deference to the Factual Findings of the Trial Court in Accordance with the "Clearly Erroneous" Rule and De Novo Review of Its Conclusion Concerning Fair Use . 36 B. Texaco Presents an Erroneous Version of the Facts . . . . . . . . . . . . . . . . . . 38 1. Profit-Motivated Commercial Purpose. 39 2. Texaco's Superseding, Non- transformative Multiplication of Copies of Original Journal Articles . . . . 40 3. There are a Number of Reasonable and Practical Alternatives to Free Unauthorized Photocopying of CCC- Registered Journal Articles. . . . . 40 4. Harm, Including Lost Subscription Sales 41 5. Reasonable and Customary Practices . 42 II. STATUTORY FOUR FACTOR ANALYSIS AND RELEVANT EQUITABLE CONSIDERATIONS COMPEL REJECTION OF TEXACO'S FAIR USE DEFENSE. . . . . . . . . . . . . . . . . . . . 42 A. The First Factor. . . . . . . . . . . . . 45 1. A Commercial Use For a Profit- Motivated Purpose Disfavors Fair Use 45 2. A For-Profit Research Use or Purpose Does Not Favor Fair Use. . . . . . . 47 3. Commercial Use in First Factor Analysis Does Not Require The Sale or Distribution of Copies . . . . . . . 51 4. Sony and Williams & Wilkins Do Not Support Texaco's Position. . . . . . 52 5. If, As Is Doubtful, Legislative History Concerning Section 108 Has Any Bearing On Fair Use Under Section 107, It Disfavors Fair Use In This Case 53 6. Texaco's Photocopying Is A Non- Transformative, Non-Productive, Superseding Use That Very Strongly Weighs Against Fair Use. . . . . . . 54 7. Texaco's Position Is Not Supported By Any Claimed "Reasonable and Customary Practice" of For-Profit Research-Oriented Companies To Engage In Free In-House Photocopying Of Copyrighted Journal Articles. . . 56 B. The Second Factor . . . . . . . . . . . . 57 C. The Third Factor. . . . . . . . . . . . . 59 D. The Fourth Factor . . . . . . . . . . . . 62 1. Judge Leval's Determinations Are In Full Accord With Well-Established Principles Of Fourth Factor Analysis 62 a. Loss of Additional Subscription Sales. . . . . . . . . . . . . 65 b. Loss of Licensing Revenues . . 65 2. Texaco's Many Contentions Concerning The Fourth Factor And The CCC Are Entirely Without Merit . . . . . . . 68 a. Scope of the Harm. . . . . . . 68 b. Potential Loss of Additional Subscription Sales . . . . . . 68 c. Loss of Licensing Revenues . . 69 d. The CCC. . . . . . . . . . . . 71 CONCLUSION . . . . . . . . . . . . . . . . . . . . . 72 STATUTORY APPENDIX . . . . . . . . . . . . . . . . . A1 Table of Authorities Cases Page Guide to Abbreviations Abbreviation Reference A ______ Appendix, with referenced page number. LA 73-142 Judge Leval's Opinion and Order rejecting Texaco's fair use defense, as amended and supplemented, with referenced page number in Appendix. LA 143-149 Judge Leval's Order of Certification under 28 U.S.C. 1292(b), with referenced page number in Appendix. T Br. _____ Brief of defendant-appellant Texaco Inc., with referenced page number. AA Br. _____ Amicus Curiae brief on behalf of the American Automobile Manufacturers Association and the Chemical Manufacturers Association, with referenced page number. AL Br. _____ Amicus Curiae brief on behalf of the American Library Association, with referenced page number. AR Br. _____ Amicus Curiae brief on behalf of the Association of Research Libraries, American Association of Law Libraries, Special Libraries Association, Medical Library Association, American Council of Learned Societies, National Humanities Alliance and Association of Academic Health Sciences Library Directors, with referenced page number. Academic Press Academic Press, Inc., one of the plaintiffs and the publisher of the Journal of Catalysis. Catalysis Journal of Catalysis CCC Copyright Clearance Center. Texaco Defendant-Appellant Texaco Inc. Pollart Cross Cross Examination Testimony of Dale F. Pollart, included in the trial record, with referenced page number. PX ___ Plaintiffs' Trial Exhibit Number, included in the trial record. TX ___ Texaco's Trial Exhibit Number, included in the trial record. BRIEF OF PLAINTIFFS-APPELLEES PRELIMINARY STATEMENT On a massive trial record studied with great care, Judge Leval's extensive findings of fact fully support his legal conclusion that there is no basis and justification for allowing Texaco as part of its profit-motivated research activities to engage in free unauthorized photocopying of entire copyrighted articles published in scientific and technical journals registered with the Copyright Clearance Center ("CCC"). His decision sets forth with great clarity the many affirmative factual and legal reasons in this case for upholding the copyright owners' exclusive rights granted in Section 106 of the Copyright Act to reproduce or authorize the reproduction of the copyrighted works, and for not permitting those rights to be defeated by the "privilege" or "limited exception" to those rights contained in the fair use provisions of Section 107 of the Act. Judge Leval engages in sound analysis of the four statutory guidelines of Section 107 and other relevant considerations in reaching the conclusion that the limited fair use privilege should not be extended to the practices of Dr. Chickering, "a typical Texaco researcher," in photocopying entire copyrighted journal articles as part of his research work for Texaco. Judge Leval followed a fact-intensive and scholarly approach in deciding that the first fair use factor strongly favors plaintiffs because of Texaco's superseding use of the original copyrighted work for a profit-making commercial purpose; certain aspects of the second factor favor plaintiffs although that factor tips in favor of Texaco because of the factual content of the copyrighted works; the third factor clearly favors plaintiffs because entire copyrighted works are being photocopied; and the fourth factor powerfully favors plaintiffs because journal publishers would suffer great immediate and future economic harm by the loss of substantial actual and future revenues if Texaco and other similar for-profit research- oriented companies were permitted to engage in unauthorized free in-house photocopying of entire copyrighted journal articles. Judge Leval also found other relevant considerations that strongly favor plaintiffs, particularly the success and efficacy of the CCC, which was formed in 1978 by publishers, authors and users in response to Congress's recommendation that an efficient mechanism be established to license photocopying. As Judge Leval found, the CCC is in large part responsible for creating an actual photocopy licensing market producing millions of dollars of revenues, that enables for-profit research-oriented companies to obtain advance authorization to photocopy copyrighted articles in CCC-registered journals using convenient and reasonably priced procedures. Texaco criticizes many aspects of Judge Leval's decision without any legitimate basis. Stripping away Texaco's mis- statements and mischaracterizations of that decision and the trial evidentiary record, its argument really boils down to only three major points. First, the widespread practice of photocopying journal articles, an integral element of Texaco's research and essential to its efficiency and success, deserves the shelter of fair use simply because it serves the salutary purpose of advancing that research, albeit profit-oriented. Second, free unauthorized photocopying should be accorded fair use immunity because it is supposedly "reasonable and customary." Third, this Court should utterly disregard as irrelevant, and give no consideration in the fair use analysis to the indisputable loss of present and future revenues that plaintiffs will suffer if the fair use defense were sustained; plaintiffs' willingness to license photocopying of copyrighted journal articles; and the existence and success of the CCC. Judge Leval's rejection of each of these points is fully supported factually and legally. As he found, research at Texaco and similar for-profit companies would not be impeded or harmed if they were unable to obtain "free" in-house photocopies of journal articles since they can obtain and use authorized photocopies of such articles through reasonable CCC-licensing and other available means. As Judge Leval said, precious little would be left of copyright protection under Section 106 of the Copyright Act for scientific and educational material if all that fair use required were that the copying was done for research or educational purposes, and granting a fair use privilege would have the reverse effect of impeding science by diminishing the market incentives to publish such scientific material. Extending fair use to this case would mean, Judge Leval found, the loss to plaintiffs of substantial present and future revenues from several existing sources: licensing by the CCC; bilateral licensing agreements; royalties from the sale of authorized photocopies by document delivery services; and the sale of additional subscriptions by journal publishers. Such harm, predicated on the existence of actual licensing and subscription markets, is properly cognizable in fair use analysis, as Judge Leval concluded, in making the required fourth factor evaluation concerning the effect of unauthorized photocopying on "the potential market for or value of the copyrighted work." As for the claim that Texaco's photocopying should be declared fair use because it is "reasonable and customary," there is neither legal nor factual support for such a claim. No American case and only a 230 year old English case, never followed, has ever sustained fair use on that basis. There is, in any event, as Judge Leval found, no factual basis for the claim, as the past and current practices of many for-profit research- oriented companies in obtaining and using only authorized photocopies of entire copyrighted journal articles prove that free in-house photocopying of such articles by Texaco is neither reasonable nor customary. Contrary to the contentions of Texaco and the amici, there is nothing improper or "circular" in fourth factor analysis in considering the consequences of plaintiffs' willingness to license the copying of their copyrighted works, and in many prior fair use cases, the Supreme Court, this Court and other courts have properly considered the actual or potential licensing markets that such willingness to license has created or would create. On the facts of this case, the willingness of copyright owners to license, together with the willingness of corporate photocopy users to enter into licensing arrangements, have resulted in the creation of actual licensing mechanisms and markets (through the CCC, bilateral agreements and authorized document delivery services). Those mechanisms and markets would be destroyed if Texaco's typical photocopying practices were found by the courts to be fair use. Destruction of those markets would immediately diminish the value of plaintiffs' copyrights. It is therefore nonsense for Texaco to argue that plaintiffs' willingness to license, and the licensing markets that have resulted, should be disregarded in fourth factor analysis of the adverse effect of Texaco's photocopying practices on "the potential market for or value of the copyright." Plaintiffs' willingness to license, and the steps they and the CCC have taken to make such licensing efficient and effective, are also pertinent, as Judge Leval found, in refuting Texaco's argument that the dissemination and use of information in journal articles will be impeded, and therefore the progress of science will be impeded, if free unauthorized in-house photocopying were prohibited. Indeed, that argument rests on the false premise, directly contrary to the fact, of an unwillingness by the publishers to license photocopying of journal articles, and therefore an unwillingness to allow broad dissemination and use of copies of those articles. Moreover, the trial evidence concerning the research, photocopying and licensing practices of many for-profit research-oriented companies proves beyond any possible doubt that such companies can successfully conduct their research activities without impediment, and without harm to science, while paying for permission to photocopy, just as they pay for all of the other costs of that research. Texaco is no different. At bottom, what this case is really all about is the desire of Texaco, General Motors, Ford, Chrysler and the other corporate amicus curiae to take a free ride " to obtain what they acknowledge are valuable intellectual property rights without paying anything for them. What they seek is the opportunity to engage in free in-house photocopying of copyrighted journal articles on a regular and widespread basis without in any way compensating the copyright owner publishers through any of the available methods. Those available methods of copyright compliance are currently being utilized by 11 of Texaco's major oil company competitors, and many other research-oriented American companies, to obtain the important benefits of photocopying entire journal articles and, at the same time, respect the statutory rights of the copyright owners to authorize such photocopying. The record shows that in earning billions of dollars annually Texaco spends upwards of $89 million annually to cover all of the salary, equipment and other costs of engaging in its profit motivated research activities, except they seek immunity from the cost of paying the copyright owner publishers for the right to photocopy journal articles. There is, therefore, no conceivable reason for permitting Texaco to avoid paying the publishers the expense of obtaining authorized copies of journal articles that Texaco considers are integral to the conduct of its research and valuable to it in conducting that research. Indeed, the fact that such photocopying enhances the efficiency and success, and therefore the profitability, of that research, is an additional reason for requiring Texaco to pay for the privilege, just as it pays for all other necessary research expenses. Thus, there is no conceivable justification for requiring the publishers to subsidize Texaco's profit-directed research activities, by depriving those publishers of revenues they are currently receiving and would receive from the exercise of their exclusive rights to authorize copying. QUESTIONS PRESENTED AND QUESTIONS NOT PRESENTED As the parties, the court below and the amici curiae all recognize, the broad question presented in this test case is whether for-profit research-oriented corporations, as exemplified by Texaco and one of its typical researchers, can engage in free unauthorized in-house photocopying of entire copyrighted articles published in scientific and technical journals registered with the CCC (LA 148). The trial focused on the practices of that typical researcher, Dr. Chickering, in making or obtaining photocopies of the entirety of each of 8 copyrighted articles published in Catalysis in furtherance of his research activities for Texaco. In determining that fair use question, there are a number of subsidiary issues to be resolved, including whether each of the four statutory guidelines, and any other relevant considerations, favors plaintiffs or Texaco, and the weight to be given to each of those factors in reaching an overall conclusion on the fair use question. In making those fair use determinations on this appeal, the many factual findings made by the court below are to be accepted as true in all respects because, with a few unsupported exceptions that we discuss later, Texaco does not contend that those factual findings are clearly erroneous or lack evidentiary basis. Thus, no genuine issues on this appeal concern the facts as found by the court below. (As discussed later, Texaco wrongly argues otherwise, contending that Judge Leval's decision is in all respects open to de novo appellate review and that his post-trial findings of fact are not to be given the normal deference and are not subject to the clearly erroneous rule.) Nor are there any factual or fair use issues presented on this appeal concerning the activities and photocopying practices of non- profit scientific research institutions, universities, government agencies, hospitals, museums, news organizations, law firms, the judiciary, law libraries, public and court libraries, and other entities in which most of the membership of many of the amici curiae carry on their activities. There is no evidentiary record concerning those activities and practices, and Judge Leval made no finding or determinations concerning them. He confined his analysis and rulings to the trial record, which focused on the research and photocopying practices of Texaco and other profit-motivated research-oriented industrial companies, and the publication, sale, licensing and use of scientific and technical journals including those registered with the CCC and articles appearing in such journals. There is also no issue on this appeal concerning Section 108 of the Copyright Act, except to the extent that Texaco and the amici claim that the legislative history of that section sheds light on fair use (T Br.19; AR Br. 13-14, 44-47). As Texaco acknowledges (T Br. 19), its separate 108 defense was not part of the trial below and remains for future adjudication as to copies made by the Texaco library. We agree with the library amici curiae that this Court should make clear that nothing in its decision on the fair use issues decides any possible future issues under Section 108 concerning library copying. (AR Br. 7 n.1; AL Br. 2 n.3.) STATEMENT OF THE CASE A. The Litigants and Interested Third Parties Plaintiffs in this action are 83 publishers of scientific and technical journals, among which are 20 non-profit learned societies and educational institutions that publish such journals, including the American Association for the Advancement of Science, American Chemical Society, American Geophysical Union, American Institute of Physics, American Medical Association, Cornell University, Massachusetts Medical Society, MIT Press, and Princeton University Press. Texaco, the defendant in this test case, is supported on its appeal by the American Automobile Manufacturers Association (whose members are General Motors, Ford and Chrysler), the Chemical Manufacturers Association (whose members include Texaco) and two other amicus curiae groups consisting of various library and related associations. The result in this case will affect not only the plaintiff publishers, the CCC, authorized document delivery services, Texaco and its corporate supporters, but also the many industrial companies that have entered into annual CCC and other photocopying licenses. As of time of trial, there were 110 CCC corporate licensees, including Exxon, Mobil, ARCO and eight other major petroleum company competitors of Texaco, and such other well-known for- profit research-oriented corporations as Allied Signal, AT&T and its Bell Labs division, Dow Corning, DuPont, Eastman Kodak, General Electric, IBM, Monsanto, Olin, PPG, Polaroid, Texas Instruments, 3M, Union Carbide, United Technologies, and USX (LA 87; A 272-73). B. The Fair Use Proceedings and Trial Record Plaintiffs' complaint, filed in 1985, alleged copyright infringement by Texaco with respect to articles published in journals registered with the CCC (A 12-33). Texaco's answer pleaded a host of technical copyright defenses concerning transfer, ownership and registration of copyrights, copyrightability, and a 108 library copying privilege, together with a defense and counterclaim of fair use under 107 of the Act (A 34-47). The proceedings were subsequently stayed due to the filing of Texaco's Chapter 11 petition. After the stay was lifted, proceedings resumed in earnest in late 1989, and in accordance with the parties' agreement and the court's approval, were limited to the issue of fair use, the core of Texaco's defense in this case. (LA 78; A 150.) From the 400 to 500 research scientists employed by Texaco, plaintiffs examined the files of ten randomly selected scientists at Texaco's nearby Beacon, New York research center. Those ten files contained more than 1,000 photocopies of articles from journals registered with the CCC, including, from Dr. Chickering's files, 25 photocopies of articles published in Catalysis and 30 photocopies of articles from other journals. Photocopies of eight of those Catalysis articles were ultimately selected as the exemplars to be used in the fair use trial, a procedure the parties adopted in their court-ordered stipulation for convenience and to avoid untoward discovery expenses with respect to largely duplicative matters. (LA 78, 145, 147.) Plaintiffs stipulated that the complaint would be dismissed in its entirety if the fair use defense were upheld as to each of those articles. (LA 75, 78 & n.1, 79, 110 n.14; A 150.) The trial proceedings, as a result of agreement between the parties with the concurrence and support of Judge Leval, consisted of four successive and alternating evidentiary submissions. Each of these submissions included the written testimony of trial witnesses, excerpts from the testimony of a number of deposition witnesses, and scores of documentary exhibits and interrogatory answers. Cross-examination of the trial witnesses, originally scheduled for the courtroom, was conducted, to the extent each of the parties elected to do so, at their attorneys' offices, and the written transcripts of such cross-examination were submitted as part of the trial record. The trial record consists of the testimony of 14 trial witnesses, 7 of whom were cross-examined; extracts from the testimony of 20 deposition witnesses; and 527 exhibits (including 8 interrogatory answers). Those witnesses included a number of Texaco executives and research and library personnel, senior executives of Academic Press, the CCC and the Bell Labs division of AT&T, and two experts, one called by each side. The 31- volume trial record has been reduced for the purposes of this appeal to a 7-volume Appendix. C. The Decisions Below Judge Leval made extensive findings of fact throughout his 67-page opinion (LA 73-142; 802 F. Supp. at 1-28) that led to his determinations that three of the four statutory factors and the important equitable considerations weighed heavily and strongly against fair use in this case (LA 94-106, 109, 110, 111-15, 120-36, 141). Taking all of these matters into account, and applying relevant legal principles, Judge Leval concluded that there was no basis for Texaco's affirmative defense of fair use under Section 107 of the Copyright Act as to the representative eight articles photocopied by Dr. Chickering. Granting certification under 1292(b), Judge Leval perceived, as did the parties, a strong public interest in immediate appellate review of the "issue of fair use, as to the practices of a representative Texaco researcher" because "photocopying of copyrighted scientific material is extremely widespread in commercial research-oriented profit-motivated companies." (LA 147-48; 802 F. Supp. at 28-30.) STATEMENT OF FACTS A. Texaco Engages In Its Research Activities (An Integral Part of Which Is the Systematic Photocopying of Journal Articles) for the Sole Purpose of Making a Profit Texaco, whose revenues and profits run in the billions of dollars (A 2834; see also A 2730-2833), conducts profit-motivated scientific research at six facilities in the United States (including the Beacon, New York research center) where it employs about 400-500 scientists. (LA 75; A 2271-72, 2277.) Texaco spends about $89 million annually on that scientific research, not as an "exercise in philanthropy," but, as Judge Leval found, "solely for [its own] commercial advantage." (LA 75, 136 & n.24; A 2273-74, 2278-79; see 932, 906, 945-47.) Those expenditures cover every element of Texaco's research activities, including salaries of the scientists and support staff, and the expenses of equipment, chemicals, offices, laboratories, journal subscriptions, information services, photocopy machines, paper and other photocopying supplies. The sole exception is that Texaco seeks to be relieved of the expense of obtaining permission to photocopy copyrighted journal articles. As Judge Leval found, and the parties agree, photocopying such journal articles is fostered and maintained by Texaco as a regular practice by its research and library personnel, and is an integral part, and very helpful to the more efficient and successful conduct, of Texaco's research activities. (LA 76-77; A 935, 469-70; see also A 1336, 730-32.) The research conducted for Texaco by Dr. Chickering and his colleagues is aimed solely at making money for Texaco. This is a crucial fact Texaco seeks to ignore and obscure (T Br. 6-7). As Texaco's former Director of Petroleum Research put it, the purpose of the research is "to maintain our competitive position and maximize Texaco's profitability." (A 2223.) The target of its research activities, as Texaco's 1988 Annual Report said, is "on enhancing profitability" (A 2373) and as the 1987 Annual Report expressed it: Texaco's research activities are geared to a profit- center concept: developing new products responsive to the demands of the marketplace; cost-reducing manufacturing process improvements; and technology to operate oil fields more efficiently. (A 2369.) On the basis of this and an overwhelming body of other evidence in the trial record (A 932, 934, 938-45, 2223, 2239-40, 2369, 2373, 2376), it is indisputable, as Judge Leval found, that Texaco conducts its scientific research activities (of which, as has been mentioned and which we discuss later, the regular and widespread practice of photocopying entire journal articles is an integral part) solely "for commercial gain" and "to improve Texaco's commercial performance," "for the purpose of producing profits." (LA 105, 106; see also his many related findings at LA 75, 105, 122 and 136.) B. The Vital Economic Importance to Texaco of Its Researchers Having Timely Access to Journal Articles Equally indisputable are Judge Leval's findings that scientific and technical journal articles play an essential role in the research programs of companies such as Texaco by disseminating the results of scientific research being conducted elsewhere. (LA 76, 111, 114.) Journal articles are the primary and sometimes only source of new scientific information on the narrowly focused, highly specialized topics commercial research scientists deal with. (A 726- 27, 742-45, 746-50, 1036-37, 1056, 4244, 1119, 1143, 456, 905, 1333-34.) They are essential because, as the former head of Texaco's Beacon research center emphasized, 1. Without awareness of new developments, our people could continue to pursue work already conducted and reported by others. In reinventing the wheel, we would waste time, manpower and money repeating work already patented by someone else, rather than using the reported work as a springboard to novel developments that would be the property of Texaco. 2. Unless our technical people continue to educate themselves about new developments in their fields, they rapidly become obsolete and of reduced value to the Company. (A 2239-40; accord A 2223.) In such circumstances, as Judge Leval found, it is "of great importance for scientists doing research to keep abreast of the publication of such articles" in order to take advantage of new ideas and developments and to avoid duplicating work already done or demonstrated to be fruitless. (LA 76.) C. The Methods by Which Research Personnel at Texaco and Other For-Profit Corporations Regularly Learn About and Obtain Photocopies of Entire Journal Articles As Judge Leval found (LA 76, 77, 80, 130-31; see LA 111- 12 n.14), and as is undisputed (T Br. 7), it is the regular and systematic practice of research scientists at for-profit industrial corporations such as Texaco, as exemplified by Dr. Chickering, to learn about journal articles and then make or obtain photocopies of those articles and create files of photocopies for use in their work. (A 727, 739, 746-47, 753, 1334, 949-52, 457-58, 594, 809-11.) This occurs in several ways: Circulation of Journals. Researchers primarily learn of journal articles by having their companies circulate journals to them. (LA 76-77, 80, 114; A 751; see A 949, 905, 457, 372-73, 374-75, 638-40, 675-77, 823.) Typically, a researcher is on the routing list for a number of journals of special pertinence to that researcher's work " in Dr. Chickering's case eight such journals including Catalysis. (A 905, 457, 372-73, 374-75, 638-40, 675-77, 823, 949, 831-32, 848-49, 2239-40, 1515, 549; see A 366, 1505-1508, 1509- 1511.) It is the systematic practice of researchers at Texaco and other research-motivated companies to photocopy articles from the circulation issues before passing them on (LA 76-77, 80, 104, 114; A 949, 457-58, 548, 551-53, 557-58, 562-63, 587-88, 594-95, 367, 395-96, 401-02, 403-06, 652-54, 667-68, 693-94, 695-98, 700-701, 703-04, 751, 1334), and this practice is economically very beneficial to Texaco, as is more fully discussed later. Literature Searches. It is also the widespread practice for research scientists at Texaco and other similar companies to have specially trained information service professionals in the company library perform computerized literature searches and automatic updates of such searches in connection with particular projects or experiments. These literature searches and updates generate bibliographic lists of journal articles and other materials, from which the researchers regularly request photocopies of articles. (A 993-1006, 1008-10, 806-08, 815-18, 822-26, 858-59, 2111, 2260-63, 2264, 950, 445-46, 541-47, 601-05, 611-15, 641, 642-47, 670-72, 673-74, 679-81, 746-47, 751-52, 1367-68, 912-15.) Other Methods. Researchers at Texaco and other for-profit industrial corporations also learn about and then make or obtain photocopies of journal articles as the result of seeing references to them in published abstracts of journal articles and Current Contents (a compendium of journal tables of contents) circulated by their companies to alert them to new journal articles, and in other journal articles or in patents. Researchers also provide each other with references or photocopies of journal articles, and route photocopies to a number of other colleagues. (A 1841-1979, 752, 850-56, 857, 659-60, 678, 683, 688) (LA 76-77; A 1368, 949-50, 568, 595, 649- 50, 704-05, 707, 714.) Library Copying. Personnel at Texaco's Beacon library regularly make photocopies of journal articles in response to individual requests by Texaco research scientists. In addition, the Beacon library makes photocopies of articles to satisfy requests made at several other Texaco libraries, and those libraries provide photocopies to Beacon upon request. (LA 77; A 463, 469, 810-11, 812-13; see A 814, 817-19, 824, 439-41.) D. The Importance to Research Personnel of Speedy Circulation of Journals and How Photocopying Reduces the Number of Subscriptions Needed to Maintain Speedy Circulation Instead of purchasing separate subscriptions for each of its research personnel, Texaco routinely circulates an original journal issue among groups of 15 or more research scientists. The Journal of Catalysis routing list, for example, has ranged from 15 to 38 researchers including Dr. Chickering. (A 549, 1515, 905, 457, 372- 75, 638-40, 675-77, 949, 823-26, 831-32, 848-49, 2239-40; see A 1505-08, 1509-11, 366.) "It is important," as Judge Leval found, "that scientific and technical journals be promptly circulated so that research scientists are made promptly aware of new published studies in the areas of their work." (LA 114; A 456-57, 554-56, 569-70, 581-83, 584-86, 668-69, 726-27, 731-32, 742-44, 754-55, 1333-34, 1056, 4244, 924, 925-26, 927-29, 1516, 1515, 2223, 2239-40, 2265, 1530; see also A 746-48.) As Texaco's former Director of Petroleum Research wrote, when there is "very slow circulation and loss in timeliness . . . [a]n inestimable penalty is incurred because technical persons cannot effectively avail themselves of new developments in current literature." (A 1530.) Rapid circulation with a minimum number of subscriptions is made possible by in-house photocopying. As Judge Leval found, photocopying "permits the scientists to defer reading [the circulation copy], and to keep possession of an additional copy without hoarding the original issue of the journal, so that the original can circulate without delay among colleagues (each of whom may do likewise), or return to the library where all colleagues will have access to it." (LA 104; see related findings at LA 76-77, 80, 114; A 754-55, 367, 395-96, 668.) Thus, "the principal feature of the photocopying is its capacity to give numerous Texaco scientists their own copy based on Texaco's purchase of an original." (LA 102.) This practice is widespread and prevails in other for-profit industrial corporations as well. (LA 76-77, 80; A 949, 457-58, 548, 551-53, 557-58, 562-63, 587-88, 594-95, 367, 395-96, 401-02, 403- 06, 652-54, 667-68, 693-94, 695-98, 700-01, 703-04, 751, 754-55, 1334.) Even with the use and benefits of photocopying, Texaco increased its subscriptions to Catalysis from 1 to 2 in 1983, and then to 3 in 1989, as Judge Leval found, "in order to speed up the circulation process." (LA 114; A 803, 1516-28, 2222-28, 569-70.) This and the other evidence discussed clearly supports Judge Leval's factual finding that in the absence of photocopying, Texaco "would increase the number of subscriptions somewhat" to maintain the speedy circulation of journals (LA 114): The evidence shows that scientists will make a photocopy of an article in order not to slow down the circulation process. If that photocopying stopped, the circulation would slow down; scientists would hold onto an issue for a longer time before continuing its routing. To speed up the circulation, it seems likely that Texaco would add at least a modest number of subscriptions to Catalysis which would increase Academic Press' revenues. There is, therefore, no basis for Texaco's and the amici's contention (T Br. 31, 32, 34; AR Br. 34-35) that the evidence does not support Judge Leval's factual finding that Texaco's unauthorized photocopying reduces the number of subscriptions Texaco would otherwise purchase if it elected not to obtain authorized photocopies of journal articles through CCC and other available means. E. Dr. Chickering's Representative Photocopying A total of 55 photocopies of copyrighted journal articles made by or for Dr. Chickering for use in his work for Texaco were found in his files, including the eight Catalysis articles selected as a representative test sample for this fair use test case (the "Eight Articles"). (See LA 79; PX-1016-1089; A 450, 457-69, 536-81, 594-95.) These were just the photocopies he had made over the years at Texaco and had not discarded by the time his files were reviewed in 1989. (See LA 110 n.14.) Dr. Chickering's photocopying practices are illustrative of the systematic practices of researchers at Texaco and other for-profit industrial corporations. (LA 111 n.14, 147.) He regularly makes or obtains photocopies of entire journal articles in most of the ways described above, retains them in his photocopy files, and refers to them from time to time in connection with his work. (A 594-95, 536-40, 457-58.) In particular, with respect to the Eight Articles, Dr. Chickering made or had someone make copies of six of the articles when the original issues of Catalysis containing those articles were circulated to him. (A 458-62, 464-67, 1443, 1459, 1463, 1474, 1479, 1487.) He obtained the other two from the Texaco Library after seeing a reference to them elsewhere. (A 463- 64, 469, 1470, 1498.) As Judge Leval found (LA 81), Dr. Chickering made or obtained the photocopies of the Eight Articles for use in his work for Texaco. (A 449, 457-69, 594-95, 536-40.) Nothing suggests that he made or obtained the photocopies for his "personal use" in the sense that his photocopying was in some way not related to his work for Texaco. (LA 81, 122.) Moreover, as later discussed, using photocopies saved Dr. Chickering time and increased his productivity, thereby immediately saving Texaco the cost of that time, in turn improving Texaco's profitability. (A 457-58, 459-67, 581-83, 584-91.) F. The Economic Benefits to Texaco and Other For-Profit Industrial Corporations of Research Scientists Using Photocopy Duplicates in Their Work Photocopying, whether free or not, produces a number of economic benefits in the research activities of a for-profit research- oriented company such as Texaco. It produces immediate economic benefits by saving Texaco money in a number of ways that enhance Texaco's profitability. First, it saves researchers' time, and saving researchers' time saves Texaco money, thereby improving its profitability. Every hour saved is money saved by Texaco. Dr. Chickering, for example, who, at the time of his testimony, was paid a salary of about $58,000 for about 2000 hours of work per year, cost Texaco about $29 per hour. (A 2291.) Other researchers were paid salaries up to $149,000 per year. (A 2295.) Judge Leval found, and the parties agree, that making and using photocopies saves the time of Texaco's researchers in several ways: it saves the time they would otherwise have to spend locating articles or going to the library to read articles, because photocopies retained in their photocopy files are immediately at hand; it saves the time they would otherwise have to spend repeating experiments or dealing with poor results because of errors made in taking notes of articles (including complex equations, diagrams and other data); it saves the time they would otherwise have to spend taking notes of the articles, instead of more easily annotating or highlighting material in photocopies; and it saves time because it is quicker for researchers to assist colleagues by providing them with photocopies rather than just references to those articles. (See LA 77-78; A 459-60, 461, 462, 584-87, 595, 935, 954-58, 727-32, 739-42, 754-55, 757-60, 1333-35.) The time that is saved means the saving of the cost of that time, and it becomes time that researchers can otherwise spend doing more valuable work for their companies. (A 955, 757; see A 582-83.) The value to Texaco of the time of its researchers was highlighted in several internal Texaco memoranda concerning the adverse economic consequences to Texaco of researchers having to spend time going to the library to read journal articles " just that one additional cost would be "over $58,000 per year." (A 1515, 2240, 2223.) As a corollary to this, internal Texaco memoranda highlight the inestimable penalty that Texaco would suffer from its researchers becoming obsolete as the result of the reluctance of many of those researchers to spend time traveling to the Texaco Beacon library to read articles. (A 2240, 1515; see A 731-32, 759- 60.) Texaco also derives economic benefits from researchers using photocopies in the laboratory, for in that way they avoid damaging or destroying the original, and Texaco is saved the cost it would otherwise incur of replacing the damaged originals. (A 727-28, 729-30, 740-41, 1335-36; see A 463.) G. There Are a Number of Readily Available Ways to Provide Researchers with Access to Journal Articles Other Than Unauthorized Photocopying Judge Leval found that there are a number of convenient ways that research personnel can be provided with ready access to journal articles other than by making unauthorized in-house photocopies. (LA 112-15, 127-29, 131-32.) He found that these include the purchase of some additional subscriptions to maintain speedy circulation (LA 114), and that authorized photocopies of journal articles are readily available by obtaining advance permission to photocopy through the CCC or pursuant to agreements directly with publishers, or by purchasing photocopies from document delivery services that have obtained photocopying authorization either through CCC or directly from publishers. (LA 112-115, 128-29, 131-32.) These methods of providing research personnel with ready access to journal articles without engaging in unauthorized photocopying are illustrated in the practices of AT&T Bell Laboratories ("Bell Labs"), which are described later. Texaco itself has obtained permission to photocopy directly from publishers, has purchased additional subscriptions, and has purchased photocopies from document delivery services. (A 803, 820-21, 860-62, 917-18, 1514-15, 2111, TEX-032.) Texaco ignores Judge Leval's finding (LA 113) and the record evidence when it claims (T Br. 31, 36) that document delivery services are not a viable alternative. Whatever brief time lag may exist in obtaining the photocopies, there is no evidence that it interferes with research efforts. In addition to Texaco's own purchases of photocopies from authorized document delivery services, Bell Labs' purchase of photocopies from such services demonstrates that purchasing photocopies in that manner does not interfere with research activities. (A 1423H.) Texaco and the amici are also wrong in contending (T Br. 40 n.21; AA Br. 10) that there was no evidence to support Judge Leval's finding that obtaining permissions directly from publishers is a viable option. As Judge Leval found (LA 128), Bell Labs exemplifies a research-intensive corporation that has obtained photocopying permission directly from many publishers as one component of its program to provide its research personnel with ready access to journal articles without claiming a privilege of free photocopying. Bell Labs has entered into about 206 agreements with publishers covering about 350 journals that are not registered with CCC. (LA 128; A 1423E.) H. Unauthorized Photocopies Supersede Original Articles and Authorized Photocopies As part of its research activities, "Texaco simply makes mechanical photocopies of the entirety of relevant articles." (LA 99.) These copies of the original are not "employed as part of a larger whole for some new purpose." (LA 99.) As Judge Leval found: The principal purpose of Texaco's copies is to supersede the original and permit duplication, indeed, multiplication. A scientist can make a copy, to be read subsequently and kept for future reference, without preventing the circulation of the journal among co- workers. This kind of copying contributes nothing new or different to the original copyrighted work. It multiplies the number of copies. (LA 99-100; LA 104.) As Judge Leval also found, [W]here three subscriptions to Catalysis are serving the needs of hundreds of scientists, the principal feature of the photocopying is its capacity to give numerous Texaco scientists their own copy based on Texaco's purchase of an original. The most prominent feature of this copying is that the copies supersede the original and multiply its presence. Thus even if some transformative purpose was present in transferring the article from its journal into a slender photocopy, that use is overshadowed by the primary aspect of the copying, which is to multiply copies. (LA 102.) That free in-house photocopies are substitutes for originals and authorized photocopies of journal articles is conclusively demonstrated by the fact that researchers at Texaco and other for- profit industrial corporations use the unauthorized photocopies for the same purpose that originals or authorized photocopies are used " to obtain and use the information for their work. (A 584, 952- 53, 954, 753-55, 667-68; see also A 1513.) As Judge Leval found, researchers read photocopies rather than originals to obtain the information in the article for their work; they take photocopies rather than originals into the laboratory to refer to the article in the course of experiments in order to avoid damaging the original; they photocopy articles for later reading and use, rather than hoarding the original, to speed the circulation of the original journal issue; they make notes on and highlight photocopies, rather than doing so on the originals; they circulate photocopies rather than originals to their colleagues in order to share work-related information; and they retain photocopies rather than originals in their files for later use and re-use in their work. (LA 77-78, 80, 104, 114). Originals, however, can serve the very same purposes. While research personnel might prefer to take a photocopy rather than the original into the laboratory in order to avoid damaging the original, or prefer to highlight or make notes on a photocopy rather than marking up the original, the original can be used in those very same ways. Similarly, the original journal article, clipped from the issue in which it appeared, could just as easily be kept in the researchers' files as a photocopy. The only difference would be that it would cost companies such as Texaco a great deal more money to provide each researcher with originals, or to replenish originals as they were damaged or used. Texaco is thus wrong when it claims (T Br. 23- 24) that its photocopy duplicates of journal articles are a non- superseding use of the original. Furthermore, it is undisputed that an authorized photocopy is identical in all respects to an unauthorized photocopy, and therefore use of unauthorized photocopies supersedes the use of authorized photocopies. I. Academic Press and Its Important Role in the Dissemination of Scientific Information Academic Press, the publisher of Catalysis, is a major for- profit publisher of scientific, technical and medical journals. (A 1035-36.) It currently publishes 105 journals, and since 1965 has created 70 new journals in response to the needs of the scientific community, a testament to the important role publishers play in the advancement of science. (LA 81, 107, 135; A 1036, 1050, 1057.) Catalysis, published since 1962, contains highly technical Articles, Notes and Letters to the Editors (collectively "articles") on experimental studies in various types of catalysis and chemical reactions at surfaces. (LA 81-82; A 1046-48, 1061; see A 1443- 1458, 1459-1462, 1463-69, 1470-73, 1474-78, 1479-86, 1487-97, 1498-1504.) The subscription rate for institutions ranged from $112.00 in 1972 to $828.00 in 1989, while the discounted subscription rate for individuals has been about one-half of the institutional rate. (LA 82-83; A 1048, 4248-49, 1545, 1207-09.) In addition to subscription sales, Academic Press receives revenue from granting permission to photocopy copyrighted journal articles: (a) TRS and AAS photocopying permissions fees paid to CCC, and (b) royalties paid by a document-delivery service that has an agreement with Academic Press authorizing it to sell photocopies of copyrighted articles. (A 254, 263, 1051-52, 4242-43, 4244, 4300, 1796, 2317.) Revenue information for Academic Press and Catalysis is set forth in A 4242-44, 4300. Publishers such as Academic Press play an important role in creating scientific and technical journals, in ensuring the scientific importance and accuracy of the published information, and in getting the journals distributed to the research community as quickly as possible. Academic Press establishes the criteria that govern the publication of the articles, chooses the scientific editors who make publication decisions based on those criteria, and monitors the journals' performance through discussions with members of the scientific community. In this way Academic Press ensures that the journals are meeting the needs of the scientific community and are disseminating scientifically correct and important new information. (A 1036, 4244, 1056-58, 1059-60, 1061, 1062, 1063-64, 1066, 1067-68, 1073-74, 726-27, 742-45, 456-57, 1333-34.) Academic Press also undertakes the financial risk of publishing the journals. It generally takes five to ten years before a scientific or technical journal first earns an annual profit, if ever, and even then it takes many years to recoup the losses incurred during the unprofitable period. (A 4245-46, 1128-30, 1016, 3802.) This is confirmed by the experience of Catalysis. (A 4246, 3802.) Currently 20 of Academic Press's journals are unprofitable, and 15 have been published between two and twelve years. (A 4245.) There is therefore no room for Texaco's spurious contention (T Br. 5 n.2) that there is no evidentiary basis for Judge Leval's well- founded factual finding that "the publication of scientific journals requires a large investment and a long period of losses endured in the hope of reaching eventual profitability." (A 135.) Publishers such as Academic Press, by creating the vehicles that have allowed scientific authors to have their work widely disseminated, have been instrumental in causing articles to be written. (A 726-27, 744-45, 1036-37, 4244, 1056, 1060, 1061; see A 1337-38, 1374.) As Judge Leval found: [W]ithout publishers prepared to take the financial risk of publishing and disseminating such articles, there would be no reason for authors to write them; even if they did, the articles would fail to achieve distribution that promoted the progress of science. (LA 107; see LA 135.) J. The Economic and Other Benefits to Scientist Authors From Having Articles Published in Journals Virtually all articles in Academic Press journals result from unsolicited manuscripts submitted by scientists. The journals are able to obtain unsolicited manuscripts because of their reputations for quality and the benefits to researchers from having articles published in such journals. (A 726, 743-45, 766, 1061-62, 932-33, 1333, 1337-38, 669.) Judge Leval found (LA 133) that even though authors are generally not paid for articles, they "have a substantial economic motivation as well as other interests in having their studies published in prestigious journals." This is because "such publication enhances their professional reputations in a manner that translates itself into remuneration. The remuneration is achieved through growth of prestige and a consequent ability to demand greater salaries or more prestigious and powerful positions." (LA 133; A 726, 743-45, 766, 1061-62, 932-33, 1333, 1337-38, 669.) In particular, for-profit industrial corporations such as Texaco take publication into account in granting raises and promotions, and grant awards to their employees for having articles published. (A 766.) K. The Copyright Clearance Center and the Success of Its Photocopying Authorization Services As Judge Leval found (LA 83-84), CCC is a non-profit central clearing-house established in 1977 by publishers, authors and photocopy users in response to a congressional recommendation that an efficient mechanism be established to license photocopying. Its Board of Directors is comprised of representatives of publishers, authors and photocopy users. (A 257, 259-60, 261; see A 1980-90.) At the time of trial, approximately 8,000 publishers (including Academic Press and the other 82 plaintiffs) had registered approximately 1.5 million publications with CCC. (A 1052, 268-69, 254; see A 2102, 3699-3727.) Judge Leval made extensive findings of fact (LA 83-88, 113- 15, 127-29, 131-32) regarding CCC's photocopying authorization services, which he found to be "reasonable" and "efficient" (LA 115, 127), and which he found "would also satisfy the needs of Texaco's scientists for photocopies at a reasonable cost and burden to Texaco." (LA 114-15, 129). As the U.S. Register of Copyrights also stated, CCC's photocopying authorizations services are "straightforward and effective." (PX-1672 at C000467.) Judge Leval found there is an actual, existing, substantial market in which major research-oriented for-profit corporations pay millions of dollars annually for permission to photocopy (LA 87-88), which would be destroyed if Texaco's fair use claim is upheld by the courts. 1. The Successful Development of CCC's Photocopying Authorization Services. CCC currently offers two services " the TRS and the AAS " by which users obtain blanket advance permission to photocopy copyrighted material in publications registered with CCC, thus eliminating the need to obtain permission on an article-by-article basis before copying. Under both services, publishers individually set fees for the photocopying of copyrighted material in their CCC-registered publications. The photocopying permissions fees vary from publisher to publisher; Academic Press's fees for Catalysis are $2.00 or $3.00 per article, depending upon publication date. (LA 84, 87; A 261-63, 1051-52, 1306, 1319.) CCC developed the TRS in 1977 with substantial assistance from the Manager of Information Services at Exxon, who was a member of CCC's Board of Directors. (A 252-53, 4302.) The TRS gives photocopy users advance permission to photocopy, for their internal use or for the internal use of specific clients, any copyrighted material in CCC-registered publications, provided they subsequently report and pay for such photocopying. The straightforward TRS reporting methods are described in A 252-53, 261-62, 278. (LA 84-86.) The user decides for itself whether a photocopying transaction need not be reported because, for example, the copied material is in the public domain. (A 2436-37, 306-07, 320-28.) To make the TRS more attractive to users, in 1983 CCC eliminated the requirement that specific articles be identified because photocopy users were concerned that such an identification might reveal their research activities. (LA 85; A 2079-80, 2353-54; see A 2311-14, 2448-51.) Texaco's contention (T Br. 40-41, 12) that under the TRS users are charged for photocopying articles in which the publishers do not own the copyrights is thus incorrect since, under the TRS, it is the user who determines whether the copying needs to be reported. Moreover, Judge Leval properly rejected Texaco's assertion that publishers are improperly claiming ownership of copyrights in certain articles, finding there was no evidence supporting Texaco's "remote and conjectural claim" and that such claims were not properly the subject of the fair use trial. (LA 137- 138.) Indeed, the evidence submitted by plaintiffs showing that Texaco's contentions are meritless amply supports Judge Leval's finding. (A 1090-93, 1013-15, 1427-28, 866-68, 291-92, 628A- 628N, 447A-447C.) Participation in the TRS increased over its first four years of operation, but by 1982 CCC decided to create an alternative service for major corporations. Many companies had told CCC they were unwilling to set up the administrative systems necessary for full compliance with the TRS (LA 86), while others expressed a willingness to pay for their photocopying under a licensing service that would eliminate the need to report individual transactions. (A 266-67, 2031-32, 2044, 2048, 2055, 2060-62.) Judge Leval found that the unwillingness expressed by certain companies to implement the TRS's administrative requirements did not mean those requirements were unreasonable or that those companies would not have used the TRS if CCC were unwilling to offer an alternative. (LA 132 and n.23.) "Considering the features of the TRS objectively," Judge Leval found that the TRS was "a reasonable practicable solution for the industrial user community," and that users who declared their unwillingness were using their bargaining power to negotiate for something they liked better. (LA 132.) Texaco and the amici's argument that the TRS is supposedly administratively too burdensome (T Br. 11; AA Br. 32-33) was factually rejected by Judge Leval, who found "[t]he evidence shows that the administrative burdens for a user complying with TRS are modest and manageable." (LA 132 & n.23.) The relative ease with which users can report copying under the TRS is amply shown by the trial record evidence. (A 278-79, 301-02, 1053-54, 2425, 2431- 34, 2450-51, 1443, 1549.) That the TRS is neither impracticable nor unduly burdensome is further proved by the fact that Bell Labs pioneered the successful use of a TRS reporting system in 1978 and has continued to use it to the present time. (See pp. 31-32.) Other companies, such as General Electric, Exxon, DuPont, Procter & Gamble, Atlantic Richfield and IBM have also successfully used the TRS. (A 2054.) In response to the concerns expressed by large corporate users, CCC (with the assistance of two expert econometricians from MIT and Harvard and the cooperation of certain major industrial corporations) developed the AAS and began offering it in 1983. (LA 86, 127.) The AAS gives the corporate user an annual blanket license to make unlimited numbers of photocopies, for internal use, of copyrighted material in the 1.5 million publications registered with the CCC. No reporting is required, except during limited photocopying surveys, and the licensee may select one of three methods for calculating the license fee, all based on the publishers' individually set photocopying fees. (LA 86-89; A 252-53, 262, 267-68, 280-86, 294-99; see A 1991, 1995.) CCC has continued to refine the AAS over the years. (A 267-68, 280-83, 294-99, 2046- 47, 2069.) As of time of trial, an average annual corporate AAS license fee was $75,000. (A 269-70, 4298-99.) Texaco's contention (T Br. 40-41, 12) that the manner in which the AAS license fee is calculated is unfair because, in calculating the amount of the fee, articles in which publishers allegedly own no rights are included, is another of Texaco's myriad arguments below that was properly rejected by Judge Leval. (LA 138; see pp. 26-27.) Judge Leval found that the AAS provides a workable and effective mechanism for conveying permission to photocopy. (LA 127-28.) Texaco has introduced no evidence that the method by which the AAS license fee is calculated has rendered the AAS in any way ineffective or unworkable. To the contrary, Judge Leval's finding that the AAS was "reasonable" and "effective" is based on the trial evidence, including the fact that 110 major for-profit industrial corporations (including eleven of Texaco's competitors) have voluntarily taken AAS licenses, with their fees being calculated in the manner attacked by Texaco, and there is no evidence that any of them have complained that there was anything unfair or improper about the manner in which the fee was calculated. Moreover, neither Texaco nor anyone else is compelled to take an AAS license if they do not want to. They can report their photocopying under the TRS, under which they decide whether to report the photocopying of a particular article. Each of the three methods for calculating the AAS license fee (A 280-84, 294-96) attempts to calculate the fee in a practicable way on the basis of the prospective licensee's anticipated photocopying of material in CCC-registered publications. The individual articles being photocopied during the limited photocopying surveys are not identified because CCC accommodated photocopy users' objections to disclosing that information. (A 296.) Although photocopies from all copyrighted publications are reported during the survey, only the photocopying from CCC-registered publications is used in calculating the license fee. (A 281.) Because the license fee is based on survey data that does not identify individual articles, it is possible that there may be articles copied for which permission would not be needed, for example, a public domain article. Any potential over-inclusiveness is merely a necessary tradeoff by photocopy users using the AAS (rather than the TRS) in exchange for avoiding the costs associated with having to report each photocopying transaction. (See also p. 71.) Judge Leval found that the AAS and TRS services provide a reasonable, practicable, and efficient method of licensing photocopying (LA 113, 114-5, 127-29, 131-32), and that "Texaco could conveniently, and without undue administrative burden, retain the benefits of photocopying at will, simply by complying with one of the CCC's licensing systems." (LA 129.) 2. The Substantial and Growing Photocopying Authorizations Market. CCC's development of the photocopying authorizations market through its TRS and AAS services has resulted in user payments (through 1989) of almost $28 million in fees for permissions and licenses to photocopy " almost $9.7 million under the TRS and over $18.2 million under the AAS. As of March 1990 CCC had distributed over $9.3 million to publishers, and is obligated to distribute the remainder of the fees received (less an average service charge of 30%) which it has deferred distributing (with the agreement of the copyright owners) to cover start-up and certain other costs. (LA 88; A 1993, 4299-4300, 285.) CCC's revenues and distributions to publishers have increased each year. (LA 88; A 270, 4298-4300.) At the end of 1989, there were approximately 400 users reporting under the TRS. As of September 1990, there were 110 AAS licensees, including eleven of Texaco's major petroleum company competitors and many other major research-oriented corporations. (LA 87-88; A 269-70, 4298, 272-73; see p. 7.) L. Bell Labs' Successful Program Since 1978 of Providing Journal Articles to its Research Personnel Without Unauthorized Photocopying AT&T Bell Laboratories is a prime example of a major industrial research institution that, in sharp contrast to Texaco, has implemented a system enabling its research personnel to have ready access to journal articles without engaging in unauthorized photocopying. Like Texaco, Bell Labs maintains large collections of scientific and technical journals in a number of libraries and individual departments. (A 1423C.) In 1978, the year in which the 1976 Copyright Act took effect, Bell Labs implemented uniform procedures governing the photocopying of copyrighted material, whether done at library photocopying centers, at self-service machines in libraries, departments and hallways, or at special microform printers in the libraries. (A 1423D-I.) Bell Labs initially obtained advance permission to photocopy through the TRS and by entering into direct photocopying permissions agreements with publishers of non-CCC-registered publications. It currently has about 206 direct agreements with publishers covering about 350 journals. (LA 128; A 1423E.) Bell Labs became an AAS licensee in 1986, and since then has also continued to use the TRS for photocopies that are not for its internal use and therefore not covered by the AAS license. (A 1423B.) When its employees need a copy of material for which Bell Labs does not have advance permission to photocopy, it either obtains photocopying permission from the publisher or, if it cannot do so, purchases an authorized photocopy of the article from a document delivery service. Thus, Bell Labs has never had to resort to unauthorized photocopying, even when it was unable to obtain permission to photocopy. (A 1423H-I.) The unchallenged evidence concerning Bell Labs' program since 1978 proves beyond question that it is possible for a research- intensive corporation to provide its employees with complete and useful access to journal articles through licensing from CCC and a combination of other methods without any resort to free or unauthorized photocopying. M. Texaco's Unauthorized Photocopying Is Not a Reasonable and Customary Practice The trial record evidence shows that a number of readily available, reasonably priced and effective means exist, other than unauthorized in-house photocopying, for providing research personnel at companies such as Texaco with additional originals or photocopies of copyrighted journal articles (including the TRS and AAS services of the CCC, blanket license agreements with individual publishers and purchasing photocopies from authorized document delivery services). (See pp. 18-19.) It also shows that a substantial number of major for-profit research-oriented companies have been successfully employing one or more of those means for many years to provide their research personnel with ready access to journal articles (pp. 25-31). On the basis of that trial record evidence, much of which we have reviewed in the prior pages of this brief, Judge Leval rejected Texaco's contention that its unauthorized photocopying of entire copyrighted journal articles was a "reasonable and customary practice." (LA 125-32.) He found that such practices were not "reasonable" because of the readily available alternatives to unauthorized photocopying and that they were not "customary" because many large corporations (including 11 of Texaco's competitors) have been using the CCC's licensing services and one or more of the other alternatives to free photocopying to provide their research personnel with ready access to journal articles. (LA 125-32). Moreover, since the advent of modern photocopying technology in the early 1960's (see T Br. 21), publishers of scientific and technical journals have consistently opposed the notion that unauthorized photocopying of copyrighted works was an acceptable norm. During the 10 years of legislative deliberations preceding passage of the Copyright Act of 1976, publishers consistently voiced their objections about photocopying without permission and made efforts, ultimately successful in Section 106(1), to make clear that the copyright owner had the exclusive right to reproduce or authorize the reproduction of the copyrighted work (thus eliminating the ambiguity perceived and relied on by the majority in Williams & Wilkins Co. v. United States, 487 F.2d 1345, 1350-52 (Ct. Cl. 1973, aff'd by an equally divided Court, 420 U.S. 376 (1975) majority). Indeed, that Congress did not view free unauthorized photocopying as fair use is reflected in its recommendation that an efficient mechanism be established to license photocopying, which resulted in the formation of CCC. (LA 83-84.) N. The Adverse Economic Effects on Copyright Owners If the Widespread Photocopying Practices Exemplified by Texaco Were Permitted as Fair Use There is an enormous demand among researchers at for-profit industrial corporations such as Texaco for access to journal articles. (LA 111; pp. 11-15.) That demand is being fulfilled in large part by the widespread and systematic practice of making in-house photocopies of journal articles (LA 126; pp. 12-15), a practice that generates substantial economic and other benefits for the company (pp. 17-18). Judge Leval found that free in-house photocopying of copyrighted journal articles by and for researchers at industrial corporations such as Texaco deprives publishers of significant revenue (LA 111-15): It is clear that, if the making of unauthorized photocopies is found not to be fair use, Texaco will nonetheless continue to provide its scientists with copies, so long as there exists a means of doing so that is not excessively expensive or burdensome. The publishers have persuasively shown that there exist convenient and reasonably priced procedures by which Texaco could obtain the necessary additional copies for its scientists. If court rulings established that the existing practice of making photocopies violates plaintiffs' copyright, Texaco would resort to one or more of these procedures to provide its scientists with copies that are necessary for their research, and Texaco's doing so would add significantly to the plaintiffs' revenues and the value of its copyrights. (LA 111-12.) In the same vein, Judge Leval said that he finds that if Texaco stopped making "free" photocopies, it would fill this gap through some combination of the methods discussed and, in doing so, would add significant value to the publishers' copyrights. It is impossible to predict which of the possible procedures Texaco would employ to achieve speed and efficiency, to avoid administrative burden and to control expense. But it is clear that whatever combination of procedures Texaco used, the publishers' revenues would grow significantly. (LA 115.) Specifically, he found that, absent such free copying, Texaco would do any one or more of (a) purchasing some additional subscriptions to maintain needed rapid circulation of journals, (b) obtaining permission to photocopy under CCC's TRS or AAS services, (c) ordering photocopies from document delivery services that pay royalties to the publishers, and (d) entering into blanket licenses with individual publishers. (LA 113-15.) With respect to subscriptions, Judge Leval found (A 114) that rapid circulation of journals is essential to researchers' work, and that free in-house photocopying plays a major role in maintaining rapid circulation. If this photocopying stopped, circulation would slow down, and Texaco would add at least some subscriptions to speed up circulation, as it did in the past when it increased its subscriptions to Catalysis to speed up circulation. (LA 114; pp. 14- 15.) ARGUMENT I. THE PROPER STANDARD OF APPELLATE REVIEW AND TEXACO'S ERRONEOUS VERSION OF THE FACTS A. The Proper Standard of Appellate Review Is Customary Deference to the Factual Findings of the Trial Court in Accordance with the "Clearly Erroneous" Rule and De Novo Review of Its Conclusion Concerning Fair Use Texaco is wrong in arguing that fair use trial decisions are open to full or de novo appellate review, that the usual clearly erroneous rule does not apply and that this Court need not accord normal deference to the district court's trial findings (T Br. 15). It is correct, in accordance with Harper & Row Pub., Inc. v. Nation Enters., 471 U.S. 539, 560 (1985), and this Court's decisions in New Era Pub. Int'l ApS v. Carol Pub. Group, 904 F.2d 152, 155 (2d Cir.), cert. denied, 498 U.S. 921 (1990), and Arica Inst., Inc. v. Palmer, 970 F.2d 1067, 1077 (2d Cir. 1992) (see T Br. 15), that the district court's ultimate conclusion concerning whether a particular use is a fair use under 107 is a mixed question of law and fact, and therefore that conclusion is subject to full or de novo appellate review. However, as Texaco fails to point out, the trial court's findings of fact in a fair use case, just as in any other case, are subject to the "clearly erroneous" standard of review set forth in Rule 52(a) of the FRCP. As this Court has said, "[t]he four factors listed in Section 107 raise essentially factual issues" and thus the fair use defense turns on "an examination of the facts in each case". Here, as in Weissmann v. Freeman, 868 F.2d 1313, 1324-25 (2d Cir.), cert. denied, 493 U.S. 883 (1989), the clearly erroneous standard is to be applied with respect to the statutory fair use factors. Analogously, the issue of "likelihood of confusion" in Lanham Act trademark infringement cases requires the examination and balancing of eight factors, and as to that, this Court has also recently held that findings of fact of the district court "are subject to reversal only if they are clearly erroneous, while the ultimate balancing of the factors is reviewed de novo by our Court." Nikon, Inc. v. Ikon Corp., No. 92-9356, slip op. at 1603 (2d Cir. Jan. 20, 1993); Bristol-Myers Squibb Co. v. McNeil-P.P.C., Inc., 973 F.2d 1033, 1043-44 (2d Cir. 1992). In addition, under Rule 52(a) the clearly erroneous standard applies to factual findings whether based directly on the evidence or made as inferences drawn from the evidence. Anderson v. City of Bessemer City, 470 U.S. 564, 574 (1985); Hygrade Envelope Corp. v. Gibraltar Factors Corp., 366 F.2d 584, 588 (2d Cir. 1966) (Friendly, J.). That Judge Leval made his factual findings on the basis of a written record also does not alter the application of the "clearly erroneous" standard. FRCP 52(a) ("Findings of fact, whether based on oral or documentary evidence, shall not be set aside unless clearly erroneous...."); see Anderson v. City of Bessemer City, 470 U.S. at 573-74; Bristol-Myers Squibb Co. v. McNeil - P.P.C., Inc., 973 F.2d at 1043 (2d Cir.); Allied Chem. Int'l Corp. v. Companhia de Navegacao Lloyd Brasileiro, 775 F.2d 476, 481 (2d Cir. 1985), cert. denied, 475 U.S. 1099 (1986); Weissmann v. Freeman, 868 F.2d at 1322 (2d Cir.). It is the function of the district court to review and study the entire trial record, which Judge Leval did. It is not the function of this appellate court, as Texaco intimates, to duplicate that effort by reviewing the trial record de novo, or to act as a fact finder, except to resolve a contention that a finding of fact was clearly erroneous. See Harper & Row, 471 U.S. at 560; Anderson v. City of Bessemer City, 470 U.S. at 573-74; Icicle Seafoods, Inc. v. Worthington, 475 U.S. 709, 714 (1986); cf. Amadeo v. Zant, 486 U.S. 214, 228 (1988) ("no excuse for the Court of Appeals to ignore the dictates of Rule 52(a) and engage in impermissible appellate fact finding.") Texaco is also wrong when it claims (T Br. 15 n.7) that plaintiffs have the burden of proving that Texaco's use was not fair. Fair use is an affirmative defense on which the defendant bears the burden of proof. To that effect see, e.g., Harper & Row, 471 U.S. at 562; Maxtone-Graham v. Burtchaell, 803 F.2d 1253, 1255 (2d Cir. 1986), cert. denied, 481 U.S. 1059 (1987); Wainwright Sec., Inc. v. Wall St. Transcript Corp., 558 F.2d 91, 97 (2d Cir. 1977), cert. denied, 434 U.S. 1014 (1978); Rubin v. Boston Mag. Co., 645 F.2d 80, 85 (1st Cir. 1981). B. Texaco Presents an Erroneous Version of the Facts Disregarding the proper standard of appellate review, Texaco disputes, mischaracterizes or ignores many of Judge Leval's important factual findings. In essence, it presents a rehash of its version of the facts, often based on snippets of evidence, that it submitted in its post-trial briefing to Judge Leval, and which he rejected. Texaco makes no contention that any of Judge Leval's factual findings lack evidentiary support, except in those instances that we have discussed elsewhere and refuted. Moreover, Texaco's baseless contention that Judge Leval failed to analyze the circumstances surrounding the copying by Dr. Chickering of the eight sample articles, and its astonishing insinuation that Judge Leval failed to base his findings and determinations on the extensive trial record, are contradicted by the face of Judge Leval's 67- page decision and his 1292(b) certification order. Here we review several, but by no means all, of the many instances where Texaco wrongly attempts in this Court to recast and reargue the facts. 1. Profit-Motivated Commercial Purpose. Judge Leval found, and Texaco acknowledges, that it is the regular practice of Texaco's scientists, as typified by Dr. Chickering, to make and use photocopies of entire copyrighted journal articles in the conduct of their research for Texaco. (LA 76, 77, 80, 130-131.) Texaco, however, contends here, as it did below, that such photocopying "is done for the purpose of advancing science, rather than for commercial gain." (LA 104-05; T Br. 6-7, 16-21.) This flies in the face of Judge Leval's explicit rejection of that factual contention (LA 105) and his findings that such research, in which the use of photocopies of journal articles is an integral and important element that is fostered and maintained by Texaco, is solely for Texaco's "own commercial gain," and that such photocopying "was carried on in a commercial context for the purpose of producing profits. . . ." (LA 75, 105, 106, 122, 136.) Texaco and the amici also misstate those findings in saying that Judge Leval found a "commercial purpose" simply because Texaco "is a for-profit entity." (T Br. 16, 20-21; AR Br. 11-12; AA Br. 22-24.) Moreover, Judge Leval rejected (LA 105) Texaco's contention (T Br. 6-7) concerning the purported altruistic purposes of Texaco scientists' speaking and writing activities, a contention that was refuted by Texaco's own General Manager of Research and Development, who testified that such activities were sanctioned and motivated purely by Texaco's self-interest (A 932-34). 2. Texaco's Superseding, Non-transformative Multiplication of Copies of Original Journal Articles. Although Texaco acknowledges (T Br. 22-23), as Judge Leval found (LA 99- 102, 104, 106), that photocopying by Dr. Chickering and his colleagues merely produces exact duplicates of original journal articles and is not a transformative use, Texaco nevertheless contends on this appeal that those copies do not supersede or perform the same functions as the originals. (T Br. 7-8, 24) Those same contentions were advanced by Texaco in the court below and rejected by Judge Leval. (LA 99-100, 102; see pp. 19-21.) He found that "the principal purpose of Texaco's copies is to supersede the original, and permit duplication"; "the major purpose of such photocopying has been multiplication of copies . . . so that the original can circulate without delay"; "the principal feature of the photocopying is its capacity to give numerous Texaco scientists their own copy based on Texaco's purchase of an original"; and "the most prominent feature of this copying is that the copies supersede the original and multiply its presence." (LA 99-100, 102) 3. There are a Number of Reasonable and Practical Alternatives to Free Unauthorized Photocopying of CCC- Registered Journal Articles. Texaco renews (T Br. 11-12, 31, 36, 40-41) the many contentions that it made, and which were rejected in the court below (LA 111-15, 127-32), that there are no realistic alternatives to free unauthorized photocopying. It makes a number of contentions, as it did below, that the CCC permissions and licensing services (the TRS and AAS) are ineffective, unfair and unworkable (T Br. 11-12, 40-43), entirely disregarding Judge Leval's rejection of those contentions and his many detailed findings that the services administered by the CCC are reasonably priced, convenient and effective. (LA 113-115, 127-129, 131-132.) Texaco contends here, as it did below, that document delivery services are an ineffective method of obtaining authorized photocopies of journal articles, again disregarding the evidence concerning its own use of such services and Judge Leval's findings that such services are effective (LA 112-14, 128). It also contends here, as it did below, that bilateral agreements or blanket licenses with individual publishers of journals are not an effective means of obtaining permission to photocopy journal articles, when Judge Leval rejected those contentions, and instead found that such agreements or licenses, like document delivery services, are a relatively inexpensive and prompt means for obtaining working copies of journal articles. (LA 112-13, 128.) 4. Harm, Including Lost Subscription Sales. Judge Leval found that if Texaco's practice of making unauthorized photocopies were not permitted, it would nonetheless continue to provide its scientists with copies so long as there are available means to obtain such copies that are not excessively expensive or burdensome. He found that a number of those available means actually exist, including CCC licensing systems, bilateral or blanket licensing agreements and the purchase of copies from authorized document delivery services. Also, as Judge Leval found, if photocopying stopped or were not permitted, Texaco would purchase a modest number of additional subscriptions to assure the needed prompt circulation of journals. Although, as Judge Leval said, it would be impossible to predict which one or more of those possible procedures Texaco would employ as the means of providing copies of journal articles to its scientists, he found that Texaco would use some combination of those methods and "it is clear that whatever combination of procedures Texaco used, the publishers' revenues would grow significantly." (LA 115; 111-15, 127-32) (emphasis added) Here, Texaco does not contest Judge Leval's factual findings concerning the great harm plaintiffs would suffer from loss of CCC and other licensing revenues, and instead contends as a legal matter, wrongly as we discuss later, that such harm should not be considered in fourth factor analysis. As to subscription sales, however, Texaco contends here, as it did below, that from a factual standpoint plaintiffs would not suffer cognizable harm from the loss of potential sales of additional subscriptions if photocopying stopped or were not permitted, a contention that Judge Leval rejected based on that evidence. (LA 114; see pp. 14-15.) Similarly, Texaco renews its contention here (T Br. 31, 36), which Judge Leval rejected, that plaintiffs would not suffer harm from the loss of revenues resulting from the activities of authorized document delivery services. (LA 113.) 5. Reasonable and Customary Practices. Texaco again contends, as it did below, (T Br. 2, 7-9, 26-28; see AR Br. 5, 6, 36- 40), that free unauthorized photocopying of entire copyrighted journal articles by Texaco and other for-profit research-oriented companies is a "reasonable and customary" practice. Judge Leval rejected those contentions. He found that, whatever may have been the case prior to 1978, the development and use since that time of CCC and other licensing procedures undercuts any basis for finding that such free unauthorized in-house photocopying by Texaco and other for-profit research-oriented companies is either "reasonable" or "customary." He based those findings on the existence of readily available, reasonably priced, efficient means of obtaining authorized photocopies or additional originals and the use of such means by a substantial number of research-oriented profit-motivated companies. (LA 125-32.) II. STATUTORY FOUR FACTOR ANALYSIS AND RELEVANT EQUITABLE CONSIDERATIONS COMPEL REJECTION OF TEXACO'S FAIR USE DEFENSE Section 106(1) of the Copyright Act of 1976 grants the copyright owner, subject to fair use under 107, the exclusive rights "to reproduce the copyrighted work in copies" and to authorize others to reproduce the copyrighted work (thus clearing up an ambiguity that existed under the 1909 Act (see LA 123)). By granting such exclusive rights, and thus establishing marketable rights in accordance with the constitutional concept (see LA 89), the copyright protection offered under 106 "supplies the economic incentive to create and disseminate ideas." Harper & Row, 471 U.S. at 558; Mazer v. Stein, 347 U.S. 201, 219 (1954). In contrast, fair use under 107, as the Supreme Court and this Court have held, is a "privilege" allowing the abridgement or copying of copyrighted material in special circumstances without obtaining permission of or compensating the copyright owner that deprives that owner of the usual benefits of copyright protection under 106. As such, fair use is a "limited exception" to the property rights conferred by the Copyright Act. Maxtone-Graham v. Burtchaell, 803 F.2d at 1255 (2d Cir.). The fair use doctrine is simply an effort "to prevent rigid application of the Copyright Act" when it would unreasonably prevent the dissemination of information or stifle creativity. Nothing in the record remotely suggests that dissemination of information would be impaired, or creativity in research would be stifled, if Texaco's claim of fair use were disallowed. To the contrary, access to the information is readily available through use of originals and authorized photocopies. In addition, the fair use doctrine, as this Court has observed, is "not a license for corporate theft, empowering a court to ignore a copyright whenever it determines the underlying work contains material of possible public importance." Iowa State Univ. Research Found., Inc. v. ABC, Inc., 621 F.2d 57, 61 (2d Cir. 1980). In essence, this is Texaco's principal argument " that research is of public importance and therefore for that reason alone fair use should be allowed. (This subject is also later discussed.) Moreover, fair use is not a haven for copiers who can afford to pay for permission to copy. It would be "fundamentally at odds with the scheme of copyright" to permit fair use as a means of "depriving copyright owners of their right in the property precisely when they encounter those users who could afford to pay for it." Harper & Row, 471 U.S. at 559. Texaco, General Motors, Chrysler, Ford and the members of the Chemical Manufacturers Association are precisely the type of companies that can afford to pay for their use of others' intellectual property, just as, at the time of trial, 110 AAS licensees were paying for that use. In the end, fair use is an equitable doctrine " an equitable rule of reason developed by the courts that was codified in the 1976 Act (LA 120). And, as this Court has instructed, even before undertaking the traditional four factor analysis, "the equitable considerations that exist in the case at bar" must be considered. Weissmann v. Freeman, 868 F.2d at 1324 (2d Cir.). Here, as Judge Leval found, there are no equitable or other reasons why Texaco should be allowed to engage in systematic in- house photocopying of copyrighted articles in CCC-registered journals, without paying for permission to make those photocopies. (LA 120-21.) There is no reason why publishers of scientific and technical journals should be compelled to subsidize the research activities that Texaco engages in solely for its own "commercial gain", and this is doubly true since Texaco pays for all of the other costs necessary to carry on those research activities. Rejecting Texaco's claim of fair use will not harm scientific research because plaintiffs are not seeking to prohibit or limit the photocopying of journal articles, or to deny Texaco and other companies access to or the use of original journal articles. On the other hand, if the courts were to accept Texaco's claim of fair use, it would mean the destruction of the multi-million-dollar photocopying licensing market that has been built up through great effort over the last 15 years at the recommendation of Congress. Texaco retains every option and may proceed in any way it chooses. It can pay for photocopying CCC-registered journal articles through the TRS (see LA 129 n.22), or obtain an AAS blanket annual CCC license; it can buy authorized photocopies from authorized document delivery services as it has done in the past (A 860-62, 917-18); it can, as it has done, purchase additional subscriptions to maintain speedy circulation of journals (LA 110- 111); or, it can decide to do none of these things, and revert to practices it and others followed before the modern age of low-cost photocopying. (A 957.) The issue is solely one of economics. Judge Leval's straightforward statutory four factor analysis confirms the lack of any justification for Texaco's defense of fair use and clearly supports his rejection of Texaco's contentions. A. The First Factor The many factual findings made and discussed by Judge Leval that Texaco's copying is a commercial use for a profit- motivated purpose, and that it is a superseding use which is neither transformative nor productive, led inexorably to his conclusion that the first factor strongly disfavors fair use. (LA 96-106, 136.) 1. A Commercial Use For a Profit-Motivated Purpose Disfavors Fair Use. The authorities are unanimous that a commercial use for a profit-motivated purpose is presumptively unfair. Sony, 464 U.S. at 451; Harper & Row, 471 U.S. at 562 (quoting Sony); Stewart v. Abend, 495 U.S. 207, 237 (1990); Rogers v. Koons, 960 F.2d 301, 309 (2d Cir.), cert. denied, 113 S. Ct. 365 (1992); Association of Am. Med. Coll. v. Cuomo, 928 F.2d at 525 (2d Cir.). This principle simply executes the explicit first factor statutory command that in each particular case the court must consider the "purpose and character of the use, including whether such use is of a commercial character or is for non-profit educational purposes." Thus, as this Court stated last year, the first factor "asks whether the original was copied in good faith to benefit the public or primarily for the commercial interests of the infringer." Rogers v. Koons, 960 F.2d at 309 (2d Cir.); accord MCA, Inc. v. Wilson, 677 F.2d 180, 182 (2d Cir. 1981); Pacific & Southern Co. v. Duncan, 744 F.2d 1490, 1496 (11th Cir. 1984), cert. denied, 471 U.S. 1004 (1985). As the Supreme Court expressed the point in authoritative fashion, "the crux of the profit/non-profit distinction [made by Section 107(1)] is . . . whether the user stands to profit from exploitation of the copyrighted material without paying the customary price." Harper & Row, 471 U.S. at 562; Rogers v. Koons, 960 F.2d at 309 (2d Cir.) (same). In whichever way the test is articulated, Texaco engages in unauthorized copying for its own commercial interests, and stands to profit from such copying without paying the customary price. Applying these principles, Texaco's commercial use of copies of the copyrighted work for a profit-motivated purpose strongly disfavors the claim of fair use, and its contentions otherwise (T Br. 17-21) are without merit. This is doubly true here because the copying itself adds to Texaco's profitability by increasing research efficiency and saving time and effort, and thus saving the financial cost of such additional time and effort. Of course, a commercial use for a profit-motivated purpose is not, ipso facto, disqualified from fair use protection. A court " in reaching its ultimate conclusion whether a particular use is or is not a fair use " must consider and weigh all of the facts pertaining to this and the other interrelated statutory factors and to equitable considerations. (LA 93.) Thus, there may be easy illustrations of a commercial for-profit use of an insubstantial portion of a copyrighted work where the copyright owner will not suffer any cognizable actual or potential harm, that would qualify as fair use. Here, however, there is no semblance of any such circumstances. Accordingly, Texaco's commercial for-profit copying strikes decisively against fair use in first factor analysis. 2. A For-Profit Research Use or Purpose Does Not Favor Fair Use. Texaco stretches fair use doctrine out of shape in contending that it is entitled to engage in free unauthorized in-house photocopying of entire copyrighted journal articles simply because such copying is done for the salutary purpose of conducting scientific research. Judge Leval properly rejected any such contention, since precious little copyright protection would remain for any scientific or educational material if that were sufficient to warrant fair use. (LA 100-01.) Similarly, the Supreme Court rejected Texaco's argument, in a different context, when it held in Harper & Row that the court below had erred in concluding that the unauthorized use of the copyrighted material "was excused by the public's interest" in the subject matter of that material. 471 U.S. at 569; id. at 555-59. As the Supreme Court emphasized in Harper & Row, It is fundamentally at odds with the scheme of copyright to accord lesser rights in those works that are of greatest importance to the public. 471 U.S. at 559. There is additionally a significant difference for fair use purposes between non-profit research and commercial profit- motivated research, as pointed out by the U.S. Register of Copyrights in his 1983 Report: [R]esearch in applied physics, for example, performed by an employee of an aerospace firm, and similar research by a graduate student in a university, may both involve the photocopying of the same scholarly articles but the copyright consequences are different: the former copying is of a clearly commercial nature, and less likely to be fair use. Report of the Register of Copyrights - Library Reproduction of Copyrighted Works (17 U.S.C. 108) 85 (1983). A leading commentator has made the same point, as applied to first factor analysis, and the statutory dichotomy between for-profit and non- profit purposes, when he observed: It is sometimes alleged that commercial entities such as drug, chemical, or manufacturing corporations are engaged in "research" within the meaning of Section 107. . . . Such entities are engaged in for-profit activities; and . . . under normal circumstances, for- profit "research" is not eleemosynary or altruistic and thus should be considered a presumptively unfair commercial use. W. Patry, The Fair Use Privilege in Copyright Law 416-17 (1985) ("Patry, Fair Use"). The first sentence of Section 107 non-inclusively identifies criticism, comment, news reporting, teaching, scholarship or research as possible purposes qualifying for fair use treatment. The text of that sentence does not say, and does not mean, that any unauthorized copying of copyrighted material for any of those purposes is fair use. In that respect, the Supreme Court has declared authoritatively that this enumeration merely provides several examples of the kind of uses that may, in a given case, be found to be fair use after a full four factor analysis, and none of which is presumptively a fair use: "This [statutory] listing was not intended to be exhaustive, or to single out any particular use as presumptively a "fair" use." Harper & Row, 471 U.S. at 561 (citations omitted). Moreover, the text of the first factor requires consideration and weighing of "the commercial or nonprofit character" of the activity Sony, 464 U.S. at 449; Harper & Row, 471 U.S. at 562 ("the profit/nonprofit distinction"). Accordingly, and observing the statutory distinction between non-profit and for-profit uses, the Supreme Court and appellate and district courts on many occasions have determined, based on the particular facts involved, that the unauthorized copying of copyrighted material for "news reporting" or for educational purposes (two of the purposes explicitly mentioned in the first sentence of Section 107) were "commercial uses" weighing against fair use in first factor analysis. Indeed, this Court in Weissmann gave no weight to the fact that the purpose of the use was for "teaching", "scholarship" and "research" and awarded the first factor to the copyright owner because that use was superseding, non- productive and motivated by professional advancement. 868 F.2d at 1324 (see LA 141). Thus, the outcome of first factor analysis does not rest on whether the unauthorized copying was done for one of the purposes enumerated in the first sentence of Section 107. Although this Court has indicated that if the unauthorized copying falls into one of those categories there is a strong, but not irrebuttable presumption that the first factor favors fair use, that presumption, even assuming it applies to for-profit research, has been rebutted here by the clearly commercial, superseding nature of Texaco's use. Furthermore, in each of the three recent cases in which that suggestion was made, the factual circumstances were altogether different from those involved here. None of those cases involved for-profit scientific purposes or the superseding copying of the entirety of the copyrighted work. Each of those cases involved non- superseding copying of a small portion of a copyrighted work. Also, unlike the present case, the allegedly infringing work in each of those cases was a new work into which the plaintiff's work had been incorporated, and the kind of use in question had historically been held by courts to be a fair use. See Arica Inst., Inc. v. Palmer, 970 F.2d at 1077-78 (2d Cir.) (a minor portion of plaintiff's work included in defendant's book " a work of criticism, comment, scholarship and research); Wright v. Warner Books, Inc., 953 F.2d 731, 736, 738 (2d Cir. 1991) (quotations and paraphrases of a very small portion of the original work included in defendant's biography); New Era v. Carol Pub., 904 F.2d at 156 (2d Cir.) (a small number of selected quotations included in defendant's biography). See also Judge Leval's discussion of these decisions, and their possible impact here. He concludes that this Court did not intend to adopt a categorical test for the first factor, and that plaintiffs should in any event prevail in this case as the result of the overall inquiry, because of the decisive effect in their favor of the third and fourth factors and equitable considerations. (LA 140-41). Also, neither Sega Enters. Ltd. v. Accolade, Inc., 977 F.2d 1510 (9th Cir. 1992), amended, 1993 U.S. App. LEXIS 78 (9th Cir. 1993) nor Atari Games Corp. v. Nintendo of Am., Inc., 975 F.2d 832 (Fed. Cir. 1992), supports Texaco's position that the research purpose of Dr. Chickering's copying favors fair use for first factor purposes. (See T Br. 17; AR Br. 11-12, 16; AA Br. 22-23.) Those cases involved entirely different circumstances. In both, the courts were being asked to prohibit reverse engineering of unreadable computer object codes where access to unprotected ideas and functional elements was physically impossible in the absence of the copying inherent in the reverse engineering. Sega, 977 F.2d at 1514, 1518, 1520, 1526; Atari, 975 F.2d at 843-44. Thus, in Sega the court stressed repeatedly that it was finding fair use only because "disassembly provides the only means of access to those elements of the code that are not protected by copyright. . . ." 977 F.2d at 1518; id. at 1514, 1520, 1526; accord Atari, 975 F.2d at 843-44. In stark contrast to those circumstances, Texaco cannot make any claim that unauthorized in-house photocopying of entire journal articles is the only possible means of access to unprotected ideas and information contained in journal articles. Texaco's research personnel have ready access to all of the ideas and information contained in the journal articles simply by reading original journal issues or reading authorized photocopies obtained by any of the readily available means. For these and the other reasons discussed by Judge Leval, he was correct in rejecting Texaco's contentions that the research purpose of its unauthorized photocopying favored fair use in first factor analysis or that such research purpose ipso facto rendered that copying fair use. 3. Commercial Use in First Factor Analysis Does Not Require The Sale or Distribution of Copies. There is no basis for Texaco's contention that a commercial use for a profit-motivated purpose requires the sale or distribution of unauthorized copies of the copyrighted work by the alleged infringer. This ignores the ordinary meaning of the words used in the statutory first factor, the "profit/non-profit" distinction made in that text, the absence of any such purported requirement in the statute, and prior judicial authority establishing for first factor purposes that an internal use can be "commercial." E.g., Sega Enters. Ltd. v. Accolade, 977 F.2d at 1522-23; Allen-Myland, Inc. v. IBM Corp., 746 F. Supp. at 534; Telerate Sys., Inc. v. Caro, 689 F. Supp. at 229; Schuchart & Assocs. Pro. Eng., Inc. v. Solo Serve Corp., 220 U.S.P.Q. 170, 181- 82 (W.D. Tex. 1983); Aitken, Hazen, Hoffman, Miller, P.C. v. Empire Constr. Co., 542 F. Supp. 252, 260 (D. Neb. 1982). See also Herbert v. Shanley Co., 242 U.S. 591, 594-95 (1917). Texaco repeatedly contends that its unauthorized in-house photocopying does not compete with plaintiffs' copyrighted work. Certainly the unauthorized in-house copying of journal articles by Texaco and other for-profit research-oriented companies supplants the great existing and future demand to obtain authorized photocopies or additional originals of those articles through any one or more of the readily available means, and in that important sense free photocopying competes with plaintiffs' efforts to sell journal subscriptions and license photocopying of journal articles. (LA 113-14.) See Weissmann, 868 F.2d at 1326 (2nd Cir.). 4. Sony and Williams & Wilkins Do Not Support Texaco's Position. In his decision, Judge Leval at some length discusses and correctly analyzes the Sony and Williams & Wilkins decisions and the pertinent distinctions between those cases and the facts of this one (LA 121-25, 127, 129), including among many other things, in both those cases, the non-profit non-commercial purposes of the copying and the absence of harm to the copyright owners. In light of that, we need not discuss those matters further, except to address one new contention Texaco has advanced here, that it did not make below, concerning Williams & Wilkins. Texaco is wrong (T Br. 20) in now attempting to equate the facts here to the facts in Williams & Wilkins Co. v. United States, 487 F.2d 1345 (Ct. Cl. 1973), aff'd by an equally divided Court, 420 U.S. 376 (1975). In Williams & Wilkins, as the majority emphasized, the hallmark of the entire duplicating enterprise was "untainted" by commercial or profit-making purpose on the part of the two government agencies making the photocopies, the National Institutes of Health ("NIH") and the National Library of Medicine ("NLM") (487 F.2d at 1354; LA 99). Nevertheless, purportedly showing that the copying was being done for a commercial profit- motivated purpose, Texaco refers to the distribution of photocopies by the NLM and NIH of medical journal articles to scientists at drug companies and to practicing doctors. First, this is wrong because the copying (the alleged infringing acts) was not being done by the drug companies or doctors, but by non-profit government agencies solely, as the majority in Williams & Wilkins concluded, for non-profit/non-commercial purposes. See Los Angeles News Serv. v. Tullo, 973 F.2d 791, 797 (9th Cir. 1992). Second, only 12% of the copies were distributed to drug company scientists or doctors, 487 F.2d at 1349, 1374, a fact referred to by the Williams & Wilkins minority as one of many facts they considered as support for their view that the copying was not fair use. 487 F.2d at 1366- 67, 1374. Here, of course, Texaco's photocopying is entirely for a commercial profit-motivated purpose. 5. If, As Is Doubtful, Legislative History Concerning Section 108 Has Any Bearing On Fair Use Under Section 107, It Disfavors Fair Use In This Case. As we have pointed out, and as Texaco acknowledges (T Br. 19), as to library copying, its 108 defense has not been tried or adjudicated and therefore no Section 108 issues are involved in this appeal. However, Texaco argues wrongly that the legislative history concerning Section 108 helps show that Texaco's unauthorized photocopying of journal articles is not a "commercial use" under the first fair use factor of 107. (T Br. 18-19; see AR Br. 13-14, 44-47.) As Judge Leval correctly pointed out, Section 108 explicitly provides that none of its provisions "in any way affects the right of fair use". (LA 137; 108(f)(4).) Also, the language in 108(a)(1) concerning "direct or indirect commercial advantage" does not appear in 107; it does appear in 110(4) and the legislative history of that section makes clear that such phrase covers activities "in connection with any commercial or profit making enterprise." S. Rep. No. 94-473 at 77; H.R. Rep. No. 94-1476 at 85. In advancing its contentions on this subject, Texaco relies on selected quotations in the House Report on 108 (T Br. 19), but fails to refer to a very important point made in the same House Report, and completely fails to discuss relevant portions of the Senate and Conference reports. See H.R. Report No. 94-1476 at 74- 75 (it is only "[i]solated, spontaneous making of single photocopies by a library in a for-profit organization, without any systematic effort to substitute photocopying for subscriptions or purchases. . . ." that should not be considered copying for "indirect or direct commercial advantage"); S. Rep. No. 94-473 at 67 (indicating that a "commercial organization should purchase the number of copies of a work that it requires [for its employees], or obtain the consent of the copyright owner to the making of photocopies"); id. at 70 (the systematic reproduction of copies of "articles or other contributions to copyrighted collections or periodicals . . . whether or not multiple copies are reproduced" is not authorized); H.R. Conf. Rep. No. 94-1733, 94th Cong., 2d Sess. at 73-74 (1976) ("the isolated, spontaneous making of single photocopies by a library . . . in a for-profit organization without any commercial motivation . . . would come within the scope of Section 108"). Clearly, the evidence establishes that Texaco's systematic, profit motivated, unauthorized photocopying practices that have an immediate commercial motivation and economic benefit are not the "isolated," "spontaneous" and "without any commercial motivation" practices referred to in the legislative history (pp. 17-18). Accordingly, if and to the extent the legislative history concerning Section 108 is relevant, it adds to the reasons for disfavoring fair use in this case. 6. Texaco's Photocopying Is A Non-Transformative, Non-Productive, Superseding Use That Very Strongly Weighs Against Fair Use. Judge Leval's discussion of this subject, on the facts of this case, leads inexorably to the conclusion that the first statutory factor concerning purpose and use very strongly disfavors fair use. (LA 93-96, 99-104.) Indeed, where copies duplicate and supersede the original, as a general rule fair use is precluded. Harper & Row, 471 U.S. at 550 ("the fair use doctrine has always precluded a use that "supersede[s] the use of the original""); United Tel. Co. v. Johnson Pub. Co., 855 F.2d 604, 610 (8th Cir. 1988) (same); Folsom v. Marsh, 9 F. Cas. 342, 345 (C.C. Mass. 1841); Basic Books, Inc. v. Kinko's Graphics Corp., 758 F. Supp. at 1530 (copying "that merely repackages or republishes the original is unlikely to pass the test" under the first factor); 3 M. Nimmer & D. Nimmer, Nimmer on Copyright 13.05[B], at 13-102.28(12) (1992) ("[i]f both plaintiff's and defendant's works are used for the same purpose, then . . . the defense of fair use should not be available. . . .") (emphasis omitted); id. at 13-102.28(21); see also Wainwright Sec., Inc. v. Wall St. Transcript Corp., 558 F.2d 91, 96 (2d Cir. 1977), cert. denied, 434 U.S. 1014 (1978); Marcus v. Rowley, 695 F.2d 1171, 1175 (9th Cir. 1983). Texaco's contention (T Br. 25-26; AL Br. 11-12), made in reliance on Consumers Union of U.S., Inc. v. General Signal Corp., 724 F.2d 1044 (2d Cir. 1983), cert. denied, 469 U.S. 823 (1984), that Dr. Chickering made the photocopies supposedly to duplicate accurately certain material in the articles and this favors fair use, is entirely misplaced. On this point Consumers Union is not good law, as Texaco fails to point out. See New Era Pub. Int'l ApS v. Henry Holt Co., 884 F.2d 659, 661 (2d Cir. 1989), cert. denied, 493 U.S. 1094 (1990) (the language in Consumers Union indicating that verbatim copying for accurate reporting favors fair use is "an incorrect statement of the law. . . .") Moreover, in Consumers Union, the court found the copying in question " 29 words out of an article of 2100 words " to be "insubstantial", 724 F.2d at 1050, and its discussion of concern for accuracy was not a part of its first factor analysis, 724 F.2d at 1049. In addition, in Consumers Union and the other cases cited in the ALA brief (p. 12), the courts were concerned with the copier's need to report facts accurately to the public, and emphasized the modest, minimal or limited amount copied. (See LA 103-04.) As to Texaco's contention that its use of unauthorized photocopies is ipso facto "productive" because it is part of scientific research, we need not reiterate what we have said earlier in commenting on Judge Leval's determination that such use is not "productive", as that term is understood in fair use doctrine. Nor need we reiterate what was said before (pp. 19-21, 40) concerning the lack of any basis for Texaco's contention (T Br. 23-24) that its use of photocopies is not a superseding use because researchers prefer to work with photocopies rather than originals. As to Texaco's contention (T Br. 16) based on the reference in 107 "to fair use by reproduction in copies," the comments made in the House and Senate Reports concerning 107 provide a complete refutation. That reference is "not intended to give those kinds of reproduction any special status under the fair use provision or to sanction any reproduction beyond the normal and reasonable limits of fair use." H. R. Rep. No. 94-1476 at 66; S. Rep. No. 94-473 at 62. Copying of the entirety of a copyrighted work, as occurred here, is ordinarily not fair use. See authorities cited in n.32. There is, in the end, no genuine argument that can be advanced by Texaco that its use of photocopies is anything other than superseding, non- productive and non-transformative. 7. Texaco's Position Is Not Supported By Any Claimed "Reasonable and Customary Practice" of For-Profit Research- Oriented Companies To Engage In Free In-House Photocopying Of Copyrighted Journal Articles. Judge Leval's factual finding that there is no such "reasonable and customary" practice ends the matter. (LA 129-30.) In any event, there is no basis for any claim that the concept of "reasonable and customary practices" plays a role in fair use adjudication. No American case has ever found fair use based on the existence of a customary practice of quotation, abridgment, or any other form of unauthorized copying. The only decision to that effect, and on which Texaco relies (T Br. 26-27), is one rendered more than two centuries ago in England, Dodsley v. Kinnersley, 27 Eng. Rep. 270, 271 (Ch. 1761). Its approach has not been followed by American courts. In fact, American courts that have considered the question have rejected any contention, similar to the one Texaco advances, that custom or practice plays any role in fair use adjudication. See Bellsouth Adv. & Pub. Corp. v. Donnelley Info. Pub., Inc., 719 F. Supp. 1551, 1561 (S.D. Fla. 1988), aff'd, 933 F.2d 952 (11th Cir. 1991), vac'd & reh'g en banc granted, 977 F.2d 1435 (11th Cir. 1992) (industry practice "is not relevant to the fair use defense"); Maxtone-Graham v. Burtchaell, 631 F. Supp. 1432, 1436 (S.D.N.Y.), aff'd, 803 F.2d 1253 (2d Cir. 1986), cert. denied, 481 U.S. 1059 (1987) (whether or not a custom existed was irrelevant); Meeropol v. Nizer, 417 F. Supp. 1201, 1210 (S.D.N.Y. 1976), rev'd on other grounds, 560 F.2d 1061 (2d Cir.), cert. denied, 434 U.S. 1013 (1977) ("Fair use is a legal question to be determined by the court not by alleged industry practices"); see also, Walter v. Steinkopff, 3 Ch. D. 489 (1892) (in which the custom of newspapers to copy from each other was no justification for infringement). Moreover, none of the courts in the American cases cited by Texaco (T Br. 26 n.11) actually considered evidence of alleged industry custom in determining whether the use was a fair use, or indicated that any such evidence should be considered. * * * In every respect, first factor analysis disfavors fair use, in our judgment, conclusively. B. The Second Factor In considering the "the nature of the copyrighted work," Judge Leval determined that "the aspect" that favors plaintiffs is that the articles are created and published with the purpose and intention of benefitting from copyright protection and that the copyright law in implementing the constitutional objective was intended to provide such protection in factual circumstances such as those presented here. (LA 106.) Moreover, when the copyrighted work "represents an investment of time in anticipation of a financial return," that fact weighs against fair use in second factor analysis. Rogers v. Koons, 960 F.2d 301, 310 (2d Cir.), cert. denied, 113 S. Ct. 365 (1992); accord MCA, Inc. v. Wilson, 677 F.2d 180, 182 (2d Cir. 1981). And as Judge Leval found specifically with respect to Catalysis, and comparable journals: [c]opyright protection is vitally necessary to the dissemination of scientific articles . . . . Copyright protection is essential to finance the publications that distribute them. . . . If cheap photoduplications could be freely made and sold at a fraction of the subscription price, Catalysis would not sell many subscriptions; it could not sustain itself, and articles of this sort would simply not be published. And without publishers prepared to take the financial risk of publishing and disseminating such articles, there would be no reason for authors to write them; even if they did, the articles would fail to achieve distribution that promoted the progress of science. (LA 107). See Harper & Row, 471 U.S. at 558 ("By establishing a marketable right . . . copyright supplies the economic incentive to create and disseminate ideas."). Nevertheless, Judge Leval concluded that the second factor favors Texaco because of the factual content of such articles and prior decisions indicating that ""the scope of fair use is greater with respect to factual rather than nonfactual works"." (LA 107-08; New Era Pub. v. Carol Pub., 904 F.2d at 157 (2d Cir.).) However, notwithstanding their factual content, the journal articles are expressions of highly original, creative and imaginative thinking, containing new theories, new hypotheses, and new solutions to difficult problems. They are as much or more creative than other works containing factual information that have been found by this and other courts to be creative. See Association of Am. Med. Coll. v. Cuomo, 928 F.2d at 524 (2d Cir.) (Medical Colleges Admission Test is a "creative, imaginative, and original" work); College Entrance Exam. Bd. v. Cuomo, 788 F. Supp. 134, 141 (N.D.N.Y. 1992) (various graduate school entrance tests); Allen- Myland, Inc. v. IBM Corp., 746 F. Supp. at 534 (computer microcode was a creative work because it was the product of substantial creative effort); Hi-Tech Video Prod., Inc. v. Capital Cities/ABC, Inc., 804 F. Supp. 950, 956 (W.D. Mich. 1992) (video travelogue). Moreover, the public interest in not restraining the free flow of factual information, see Harper & Row, 471 U.S. at 563, is not implicated here because journal articles remain freely available. Although we believe that consideration of all these matters should have led Judge Leval to award the second factor to plaintiffs, in our judgment the rejection of the fair use defense based on consideration and weighing of all the four statutory factors and other equitable considerations is not affected even if second factor analysis favors Texaco. C. The Third Factor Because copying an entire copyrighted work ordinarily militates against a finding of fair use, Judge Leval correctly concluded that "[t]his factor clearly favors the plaintiffs, as Chickering has copied the entirety of the [eight] copyrighted articles in question." (LA 109.) The copying here is substantial not only from a quantitative standpoint, but from a qualitative one as well. "[T]he fact that a substantial portion of the infringing work was copied verbatim is evidence of the qualitative value of the copied material. . . ." Harper & Row, 471 U.S. at 656. Accord Cable/Home Comm. Corp. v. Network Prods., Inc., 902 F.2d 829, 845 (11th Cir. 1990). Here, Dr. Chickering and the other Texaco research personnel make or obtain copies of entire articles, thus conclusively showing that these articles are qualitatively significant. Only in an unusual exceptional case like Sony or Williams & Wilkins, neither of which are appropriate to the facts presented here (see LA 121; those cases are at "the remote extremities" of fair use), has copying an entire work been held to be fair use. Texaco does not dispute the fact that entire articles were copied and it ignores the explicit text of the third statutory factor that commands consideration of "the amount and substantiality of the portion used in relation to the copyrighted work as a whole" (emphasis supplied). Instead, Texaco advances the same baseless argument that Judge Leval rejected below (LA 109-10), that copying the entirety of a copyrighted journal article is not a quantitatively or qualitatively substantial taking because each article is published in a larger copyrighted work " the journal issue. (T Br. 28-29.) There can be no doubt that each article is a separate copyrighted work. As Judge Leval found, an individual article becomes a copyrighted work long before it is published in a particular journal, since copyright inheres in all original, fixed "works of authorship" from the moment of creation (LA 110). 17 U.S.C.A. 102(a), 302(a) (West 1977 & Supp. 1992). Moreover, the provisions of the Copyright Act covering collective works make it abundantly clear that a journal article is a copyrighted work in and of itself, apart from the other articles and the journal collectively. See 17 U.S.C.A. 101 (West 1977) (defining "collective work" as an assemblage of "separate and independent works"); id. 201(c) ("Copyright in each separate contribution to a collective work is distinct from copyright in the collective work as a whole. . . ."). It makes no difference that Academic Press, for reasons of convenience and economy (LA 110), simultaneously registers both the individual articles and the issue of Catalysis in a manner prescribed by the Copyright Office. The manner of registration simply has nothing to do with whether each article is a separate copyrighted work. A number of decisions confirm Judge Leval's conclusion that the "copyrighted work" against which the copying is measured is the article, not the larger publication in which it may appear. See Consumers Union of U.S., Inc. v. General Signal Corp., 724 F.2d at 1050 (2d Cir.); Weissmann v. Freeman, 868 F.2d at 1325 (2d Cir.); Hustler Mag., Inc. v. Moral Majority, Inc., 796 F.2d at 1154- 55; Pacific & Southern Co. v. Duncan, 744 F.2d at 1497; Quinto v. Legal Times of Wash., Inc., 506 F. Supp. 554, 560 (D.D.C. 1981). Even the Williams & Wilkins majority recognized that photocopying an entire article was copying an entire copyrighted work, for it felt obliged to go to some effort to reject the established legal principle that copying the entire work generally cannot be fair use. 487 F.2d at 1353; see id. at 1366 (Cowen, C.J., dissenting). Texaco's contentions concerning the third factor are baseless and misleading. Judge Leval correctly awarded the third factor to plaintiffs. D. The Fourth Factor The fourth factor requires an examination of the facts to determine "the effect of the use on the potential market for or value of the copyrighted work," an inquiry which Judge Leval conducted. As a result he made a series of factual findings that the actual and potential markets for the copyrighted work would be greatly harmed, and the value of those copyrighted works would be very substantially diminished by the unauthorized copying of those works, as exemplified by Dr. Chickering. (LA 111, 120, 125.) On this basis, Judge Leval concluded that the evidence "powerfully demonstrated" plaintiffs' entitlement to prevail as to the fourth factor (LA 111), giving "strong support to the conclusion that this copying is not a fair use." (LA 120.) 1. Judge Leval's Determinations Are In Full Accord With Well-Established Principles Of Fourth Factor Analysis. The fourth factor "is undoubtedly the single most important element of fair use." Harper & Row, 471 U.S. at 566; Rogers v. Koons, 960 F.2d at 311 (2d Cir.); Weissmann v. Freeman, 868 F.2d at 1325 (2d Cir.). This is true because the underlying concept of constitutional and statutory copyright protection is to create marketable rights that supply economic incentives to create and disseminate ideas in published works. Thus, in the overall weighing process involved in making a fair use determination, the greater the economic harm caused by the unauthorized use, the more a fair use will be disfavored, because of the greater adverse effect on those economic incentives. Conversely, the absence of economic harm caused by the unauthorized use strongly favors a fair use, because no economic incentive would be adversely affected. Rogers v. Koons, 960 F.2d at 312 (2d Cir.). Nevertheless, in traditional fourth factor analysis relatively little need be shown by the copyright owner concerning harm to prevail on the fourth factor. "Actual present harm need not be shown" and it is not "necessary to show with certainty that future harm will result" from the use. Sony, 464 U.S. at 451. See, e.g., Salinger v. Random House, Inc., 811 F.2d 90, 99 (2d Cir. 1984), cert. denied, 484 U.S. 890 (1987); Meeropol v. Nizer, 560 F.2d at 1070 (2d Cir.). Sony went on to formulate an additional concept, particularly favorable to copyright owners in commercial settings, when it declared that although it is necessary in the case of non-commercial unauthorized use to show "by a preponderance of the evidence that some meaningful likelihood of future harm exists," in the case of a commercial unauthorized use, the likelihood of meaningful future harm should be "presumed." Sony, 464 U.S. at 451. Accord Rogers v. Koons, 960 F.2d at 312 (2d Cir.); Association of Am. Med. Coll. v. Cuomo, 928 F. 2d at 525 (2d Cir.); Financial Info., Inc. v. Moody's Investors Serv., Inc., 751 F. 2d 501, 510 (2d Cir. 1984), cert. denied, 484 U.S. 820 (1987). This "commercial use" presumption of harm disfavoring a fair use under the fourth factor is thus interrelated with the "commercial use" presumption disfavoring a fair use under the first factor. Sony, 464 U.S. at 451; Association of Am. Med. Coll. v. Cuomo, 928 F.2d at 525 (2d Cir.); Financial Info., Inc. v. Moody's Investors Serv., Inc., 751 F.2d at 510 (2d Cir.); Narell v. Freeman, 872 F.2d 907, 914 (9th Cir. 1989). It can be presumed that if the copier is making a commercial use of the copyrighted work (first factor), the copier would be willing to pay an economically reasonable amount for that use, and therefore it can be presumed that failure to pay causes harm to the copyright owner (fourth factor). As a matter of economics, and as actual experience with the CCC and other available licensing means has proved, the profit-motivated copier will be willing to pay to the copyright owner an amount equal to or less than the economic value to that copier of making that use. Fourth factor actual or potential harm is not simply the harm caused by the particular unauthorized use being considered (here, Dr. Chickering's unauthorized copying of eight Catalysis articles), which itself may be sufficient, but also the harm, often more important, that would flow from widespread conduct of the kind engaged in by defendant (here, the unauthorized photocopying of journal articles by personnel at Texaco and other profit-motivated companies). See Sony 464 U.S. at 451; Harper & Row, 471 U.S. at 568 (quoting Sony); Rogers v. Koons, 960 F.2d at 312 (2d Cir.); H.R. Rep. No. 90-83, 90th Cong., 1st Sess. 35 (1967) ("Isolated instances of minor infringements, when multiplied many times, become in the aggregate a major inroad on copyright that must be prevented."); 3 Nimmer on Copyright 13.05[A] at 13-102.26 (the question is not "the extent of damages to plaintiff caused by the particular activities of the defendant," but rather "whether unrestricted and widespread conduct of the sort engaged in by the defendant would result in a substantially adverse impact on the potential market" for plaintiff's work.). Judge Leval, therefore, correctly concluded that fourth factor harm is not determined solely on the basis of what Academic Press lost as the result of Dr. Chickering's copying of the Eight Articles, but "[i]n appraising the effect of Texaco's photocopying on the values of the publishers' copyrights, Chickering's eight copies are considered as representative, and not as the universe of alleged infringement." (LA 110 n.14.) On the basis of what has been said, Judge Leval's findings and the trial evidence make clear that the widespread unauthorized photocopying of the kind typified by Dr. Chickering, if permitted as fair use, would cause great actual economic harm to the existing and potential markets for the copyrighted works, and thereby greatly diminish the value of those copyrights. The factual and evidentiary showing of such harm here is so real and great " including the loss and dismantling of existing CCC and other journal licensing mechanisms " there is no need, as has often occurred in other cases where fair use defenses have been rejected, to theorize about potential future harm, or to invoke evidentiary or substantive presumptions disfavoring fair use. Indeed, no prior fair use case has ever involved a greater showing of such clear and irrefutable actual and potential harm. Fourth factor considerations, therefore, powerfully disfavor Texaco's fair use defense. As Judge Leval found, there is a great need in conducting profit-motivated research to obtain and use photocopies of copyrighted journal articles and if Texaco stopped making ""free" photocopies" it would fulfill that need through some combination of the readily available methods (LA 111-12). In doing that, the publishers' revenues would grow significantly and the value of their copyrights would be enhanced (LA 112, 113, 115). Thus, as Judge Leval found, if Texaco's free unauthorized in-house photocopying were permitted, the markets for and the values of the copyrights would be substantially diminished (LA 120). As Judge Leval found, there are two readily demonstrable types of harm that plaintiffs have and will suffer. a. Loss of Additional Subscription Sales. If unautho- rized photocopying were not permitted and a company such as Texaco elected not to make or obtain authorized photocopies of journal articles, it is clear, as Judge Leval found (LA 114), that a company such as Texaco would purchase a modest number of addi- tional subscriptions to assure the necessary prompt circulation of journal issues to researchers. That is needed to avoid the many penalties of decreased efficiency, difficulties in research and adverse effects on profitability if circulation of those journals were delayed. Since, for example, the cost of an annual subscription to Catalysis is about $825 (A 1545), even one or two additional subscriptions purchased by each of the many petroleum and chemical companies interested in that field would add many thousands of dollars of revenue. b. Loss of Licensing Revenues. Massive actual harm would be suffered by the copyright owners from the unauthorized photocopying of the type typified by Dr. Chickering in the loss of its existing and future revenues resulting from the permissions and licensing programs administered by the CCC. Additional actual harm would also result from the loss of existing and future licensing revenues from bilateral or blanket license agreements and from the loss of royalties from authorized document delivery services that sell authorized copies of journal articles. None of this is theorizing. The truth is established beyond dispute by Judge Leval's findings based on the evidence concerning the CCC and the other available means of obtaining authorized copies or additional originals, and the evidence concerning the practices of many companies which avail themselves of such means to conduct their research successfully by using authorized copies and compensating the copyright owners for that use. The loss of actual or potential revenues from licensing the copyrighted work is, of course, cognizable under conventional fourth factor analysis, and Texaco's contentions otherwise, which we discuss later, are entirely unfounded. The exclusive rights to reproduce or authorize the reproduction of the copyrighted work are marketable rights that can be exploited in different ways. A copyrighted book may be sold in as many copies as are published and purchased; the copyright owner may also exploit the copyright by licensing paperback rights, or by licensing a newspaper or magazine to publish a chapter or serialize the entire book. A copyrighted film may be rented to theater owners or may be licensed to home video distributors and licensed for pay-per-view, cable and broadcast television presentation or other use. Theater plays or musicals may be licensed for performance and print. In the music field, records or tapes may be sold, and performances are regularly licensed through ASCAP, BMI or other means. In the computer field, multiple copies of programs are sold or a single copy is licensed for multiple reproduction and use. Thus among the means by which marketable rights afforded by copyright protection are exploited by copyright owners, licensing often figures most prominently. For example, in both Harper & Row and Stewart v. Abend, the loss of potential licensing revenues due to the unauthorized use adversely affected the value of the copyrighted work and therefore, for purposes of fourth factor analysis, fair use was disfavored. 471 U.S. at 568 (quoting 3 Nimmer on Copyright 13.05[B], at [13- 102.28(4) (1992)]; 495 U.S. 207 (1990). As the Supreme Court explained, there is harm under the fourth factor if there is an adverse effect upon "the value of any of the rights in the copyrighted work. . . ." Harper & Row, 471 U.S. at 568. In Harper & Row, the harm was the loss of a portion of the fee the copyright owner was to receive for granting a pre-publication license to Time Magazine for permitting the publication of a small portion of President Ford's memoirs. In Stewart v. Abend the loss which would have occurred if the unauthorized use were permitted was the opportunity to license rights in the story to other film makers. There are many other cases in this and other courts where fourth factor harm was recognized to include the loss or potential loss of revenues from licensing or the sale of rights that would occur if the unauthorized use were permitted as fair use. Financial Info., Inc. v. Moody's Investor Serv., Inc., 751 F.2d at 509-10 (2d Cir.) ("[I]f Moody's were unable to copy from [plaintiff's works, plaintiff] might be in a position to license that use for a fee"); DC Comics, Inc. v. Reel Fantasy, Inc., 696 F.2d at 28 (2d Cir.) ("one of the benefits of ownership of copyrighted material is the right to license its use for a fee."); Meeropol v. Nizer, 560 F.2d at 1070 (2d Cir.) (harm to letters because the "market . . . for the sale of motion picture rights might be affected"); United Tel. Co. v. Johnson Pub. Co., 855 F.2d at 610 (permitting defendant to use plaintiff's copyrighted data would defeat plaintiff's "market in licensing" that data); Hi-Tech Video Prod., Inc. v. Capital Cities/ABC Inc., 804 F. Supp. at 955 ("deprivation of a would-be licensing fee . . . constitutes sufficient harm"); Basic Books, Inc. v. Kinko's Graphics Corp., 758 F. Supp. at 1534 (loss of permissions fees for granting permission to photocopy portions of plaintiff's copyrighted works weighed heavily against fair use); Richard Anderson Photography v. Brown, 1990 U.S. Dist. LEXIS 19846 at 5 (W.D. Va. 1990) (loss of potential licensing fees "clearly affects the value of the copyrighted work"); Telerate Sys., Inc. v. Caro, 689 F. Supp. at 229-30 (defendant's use deprived plaintiff of the "opportunity to charge for the privilege of copying"); Craft v. Kobler, 667 F. Supp. 120, 129 (S.D.N.Y. 1987) (possibility of future lost license fees); D.C. Comics, Inc. v. Unlimited Monkey Bus., Inc., 598 F. Supp. 110, 113, 118 (N.D. Ga. 1984) (defendant's use caused harm to plaintiff's business of licensing use of "SUPERMAN" name, character, costumes and images). Accordingly, in statutory fourth factor analysis, the courts have consistently and correctly considered the adverse effects of the unauthorized use of actual and potential licensing markets for the copyrighted work and thus on the value of that work. 2. Texaco's Many Contentions Concerning The Fourth Factor And The CCC Are Entirely Without Merit. Texaco advances a multitude of contentions in a vain attempt to show that no actual or potential harm will occur under fourth factor analysis if Dr. Chickering's unauthorized use of copyrighted works were permitted as fair use. a. Scope of the Harm. For the reasons previously discussed (pp. 63-64), Texaco is wrong in contending that consideration of harm must be limited to what Academic Press lost by reason of Dr. Chickering's photocopying of the eight articles (T Br. 35-36), and the two cases it cites do not support its contention. b. Potential Loss of Additional Subscription Sales. Texaco does not challenge consideration of lost subscription sales under fourth factor analysis, but instead contends wrongly (T Br. 32-35), as has been previously shown (pp. 14-15), that Judge Leval's "factual inferences" concerning lost additional subscription sales lack basis in and are contrary to the record. c. Loss of Licensing Revenues. Texaco does not dispute that if the unauthorized copying typified by Dr. Chickering were permitted as fair use, substantial harm would be suffered by loss of CCC and other licensing revenues, thus diminishing the value of the copyrights. Instead, Texaco argues, as it did unsuccessfully below, that this harm should be ignored and disregarded in fourth factor analysis. That argument, in whatever form it takes, is groundless in view of the authorities and reasons previously discussed, establishing that the loss of licensing revenues is harm to be considered under fourth factor analysis. Texaco ignores all of that well grounded authority and wrongly relies on the unsound reasoning of Williams & Wilkins footnote 19, which has never been followed by any court. Fourth factor analysis does not, in and of itself, decide the ultimate issue of fair use as that footnote presumes. Rather, as the footnote overlooks, fourth factor analysis requires evaluation of the "effect of the use upon the potential market for or value of the copyrighted work" as one of the elements, albeit an important one, to be weighed in the over-all determination of fair use. It is therefore necessary to determine what would happen economically if Dr. Chickering's unauthorized use, and widespread unauthorized use of the same type, were permitted, and it is clear that the effect would be that the copyright owners would lose the substantial licensing revenues they would otherwise earn if the use were not permitted. Under the flawed reasoning of footnote 19, the defendant would always claim that whatever harm plaintiff has suffered, whether lost sales or lost licensing revenues, such harm should not be considered because if the defendant's use is a fair use, the plaintiff would not be entitled to the lost revenues. That would read the statutory fourth factor out of fair use analysis. Moreover, in Williams & Wilkins there was only claimed theoretical lost future royalty income and no demonstrable lost licensing revenues, as here. Texaco's contention based on Sony (T Br. 38-39) is groundless because in that case there was no evidence, as here, that users actually paid or were prepared to pay license fees. Texaco's further contention that fair use analysis cannot consider the willingness of a copyright owner to grant a license is also wrong. The need for photocopies of journal articles, plaintiffs' willingness to license photocopying through CCC and other means, and users' willingness to pay for the right to make or obtain authorized photocopies, have together created an actual licensing market that would be adversely affected, indeed destroyed, if unauthorized photocopying was permitted. Also, as a result of plaintiffs' willingness to license, there is no basis for any claim that the progress of science will be impeded, since for-profit research- oriented companies, such as Texaco, are afforded the opportunity to obtain authorized photocopies of journal articles through reasonably available means. Texaco's contention (T Br. 41) that consideration of licensing income for fourth factor analysis involves a "quintessentially legislative judgment" is inexplicable. In the 1976 Act, codifying the fair use doctrine, Congress has commanded the courts in the fourth factor to continue to consider the economic effects of the unauthorized use in determining whether that use is a fair use. d. The CCC. Texaco makes a number of factual contentions, which Judge Leval rejected and which have been previously discussed, concerning the TRS and AAS permissions and licensing services administered by the CCC. In addition, Texaco contends that because the AAS license fee is calculated in an allegedly unfair manner (a factual contention also rejected by Judge Leval), the AAS constitutes copyright misuse. That argument was advanced below by Texaco, together with a number of other arguments, that Judge Leval rejected in saying at the conclusion of his decision that "Texaco's other points are without merit." (LA 138.) In any event, it is questionable whether a defense of copyright misuse exists, and even if it did, it would constitute a separate defense for future adjudication, since a claim of misuse has never been considered as an element of the fair use inquiry. Moreover, the Supreme Court and this Court have made unmistakably clear that a blanket annual license, such as the AAS, is an appropriate method of granting and obtaining permission to copy from specified copyrighted works; is a practical means of accommodating the needs of users for rapid access to the repertory of copyrighted works and the copyright owners' right to be compensated; and eliminates the burdens and costs of individual reporting by users. Broadcast Music, Inc. v. Columbia Broadcasting Sys., Inc., 441 U.S. 1, 20-22 (1979); Buffalo Broadcasting Co. v. American Soc'y of Composers, Authors & Pub., 744 F.2d 919, 927 (2d Cir. 1984), cert. denied, 469 U.S. 1211 (1985). As Judge Leval found, the AAS annual license service accomplishes these three goals for copyrighted articles published in CCC-registered journals (LA 113-15, 127-29). Texaco's various other contentions concerning the purported inadequacy or inefficiency of the CCC were rejected by Judge Leval who concluded, based on all of the evidence that it was an efficient and workable mechanism. (LA 115, 127.) In the end, there is nothing to Texaco's many contentions concerning the fourth factor. On the facts and on the law, there cannot be the slightest doubt that great harm would be inflicted on the copyright owners if the unauthorized use were permitted and there is no justification for permitting Texaco and others like it to inflict that harm. CONCLUSION There is not a single good reason to sustain Texaco's defense of fair use and there are many good reasons to reject it. Judge Leval's decision should therefore be affirmed, with costs pursuant to 28 U.S.C. 1912. Dated: New York, New York April 19, 1993 Respectfully submitted, PROSKAUER ROSE GOETZ & MENDELSOHN Of Counsel: 1585 Broadway New York, New York 10036 Stephen Rackow Kaye (212) 969-3000 Jon A. Baumgarten James F. Parver Attorneys for Appellees Christopher A. Meyer Karen E. Clarke Susan L. Hochman STATUTORY APPENDIX The full text of Section 106 of the Copyright Act, 17 U.S.C. 106, is as follows: 106.Exclusive rights in copyrighted works Subject to sections 107 through 120, the owner of copyright under this title has the exclusive rights to do and to authorize any of the following: (1) to reproduce the copyrighted work in copies or phonorecords; (2) to prepare derivative works based upon the copyrighted work; (3) to distribute copies or phonorecords of the copyrighted work to the public by sale or other transfer of ownership, or by rental, lease, or lending; (4) in the case of literary, musical, dramatic, and choreographic works, pantomimes, and motion pictures and other audiovisual works, to perform the copyrighted work publicly; and (5) in the case of literary, musical, dramatic, and choreographic works, pantomimes, and pictorial, graphic, or sculptural works, including the individual images of a motion picture or other audiovisual work, to display the copyrighted work publicly. The full text of Section 107 of the Copyright Act, 17 U.S.C. 107, as amended October 24, 1992, is as follows: 107.Limitations on exclusive rights: Fair use Notwithstanding the provisions of sections 106 and 106A, the fair use of a copyrighted work, including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright. In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include" (1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and (4) the effect of the use upon the potential market for or value of the copyrighted work. The fact that a work is unpublished shall not itself bar a finding of fair use if such finding is made upon consideration of all the above factors. FOOTNOTES 1. The texts of Sections 106 and 107 are set forth in the Statutory Appendix to this brief. See pages 42-43 for the authorities establishing that fair use under 107 is a privilege and limited exception to the exclusive rights granted copyright owners in 106. 2. Texaco's reference to the fact that as of the time of trial only 19% of the Fortune 500 companies were CCC licensees (T Br. 12) is misleading because a large number of those companies do not engage in significant scientific and technical research. Currently, as we have been advised, there are more than 225 CCC industrial corporate licensees, who include about 28% of the Fortune 500 companies and 64% of the Fortune 100 companies. 3. Subject matter jurisdiction of this copyright infringement action rests on 28 U.S.C. 1331 and 1338(a), and, as to damages, also on 28 U.S.C. 1334, since in Texaco's Chapter 11 proceedings the 83 plaintiffs filed proofs of claim for damages, subsequently transferred to the court below for determination (A 53, 160-72). 4. As to Texaco's commercial exploitation of the patents resulting from this research, see A 933-34, 2369, Pollart Cross pp. 19-21; see also A 916. 5. The five other Texaco researchers deposed in this action all engaged in the various photocopying practices described above. (A 368-69, 376-77, 381-96, 398-404, 407-414, 417-434, 608-09, 614-15, 617-21, 626-28, 646-50, 652-54, 658, 662, 667-68, 674, 682-83, 685-89, 692-95, 699-701, 703-05, 707-713, 715-18; see also A 751-53, 1367-68.) 6. The amici are therefore wrong when they suggest that "personal use" copying is what is involved in this case. (AR Br. 5, 6, 7, 10.) Judge Leval made a specific factual finding rejecting that contention as made by Texaco in the proceedings below. (LA 81, 122.) 7. A 377-82, 385-86, 388-89, 395-96, 402, 415-17, 431-33, 459, 461-63, 465, 467, 536-38, 550, 564-66, 578, 609-10, 651, 661, 663- 66, 668, 689-91, 695, 697-98, 703-06, 727-30, 740-41, 746-48, 752- 55, 935-36, 952-54. 8. There is no evidentiary or other basis for the assertions that the institutional rate is double the individual rate because journal issues sent to institutional subscribers are used by and circulated to many people (T Br. 4-5) or because publishers expect copies to be made (AL Br. 13 n.6). The institutional rate is the "proper subscription rate for the journal" and individual subscriptions are sold at "a steeply discounted rate" as a "service to the community." Often the personal subscription rate either provides no profit or may not even cover the cost of providing the subscription. (A 1207-09.) 9. Texaco mischaracterizes Academic Press's profits as having been "extraordinary." (T Br. 5, see A 3802, 4320-21, 4363.) In doing that, Texaco also irresponsibly repeats what it did below in making what Judge Leval called a "demagogic effort to undermine publishers' rights by tarring them as wealthy profiteers24 [that] carries no force in copyright analysis, which does not begrudge copyright profits." (LA 135-36.) In note 24, Judge Leval added that this was "an odd argument . . . to be made by an oil company that reported over $2.4 billion net income for fiscal 1989 on revenues of over $34 billion." (LA 136.) 10. Users learn that a publication is registered with CCC in a number of ways: (i) from masthead statements in the front of the journals, (ii) from catalogs and supplements distributed to TRS users, and (iii) from Schedule A to the AAS license. The information needed to report photocopying under the TRS, in the case of Academic Press, is found in standardized fee codes at the bottom of the first page of each copyrighted article and on the masthead page. (LA 85 n.5; A 1443, 1459, 1463, 1471, 1474, 1479, 1487, 1499, 1547-1794, 1053-54, 4242, 1069, 278-79.) 11. AAMA's suggestion (Br. 33) that under the TRS a company may report and pay for copying an article written by the company's employee when it should not be required to do so, simply ignores the fact that it is the company itself that decides whether to report a particular photocopying transaction. Moreover, the author's affiliation is readily apparent from the face of the article. See, e.g., A 1443, 1463, 1479, 1487. 12. Under all three methods, the annual license fee is determined on the basis of data derived from limited photocopying surveys, factored by the licensee's employee population and the copying fees individually set by the publishers. (LA 87; A 281-85, 294-95.) 13. If the licensee photocopies more than what is anticipated or photocopies more articles with higher photocopy fees than anticipated, there is no surcharge. 14. The license fee is calculated on the basis of limited photocopying surveys designed by outside experts, during which each employee who makes a photocopy from a copyrighted publication is required to make an extra copy of the front cover or title page, and indicate on it the number of pages copied and the number of copies made. (A 295-96.) 15. The automobile manufacturers make a new and unfounded contention (AAMA Br. 13, 33, 34-35) that, apart from reporting titles of articles, reporting to the CCC the names of the journals that were photocopied raises problems of research confidentiality. There is no evidence that any company ever had such a problem or ever made such a claim, or that the CCC conveys that information to any competitor or anyone else. Moreover, companies reporting photocopying under the TRS or as part of the AAS license surveys plainly do not find that any such risk exists. 16. Texaco misstates the record when it claims that only $4.2 million has been distributed to publishers (T Br. 12). 17. Publishers' photocopying has no bearing on the existence of the custom Texaco contends exists. They have little occasion to photocopy copyrighted materials of other publishers. A 1251, 1281- 82, 1288-89, 868, 899-901. Texaco's evidence of publishers' photocopying consists of only a few isolated instances (T Br. 26) when a photocopy may have been made. Texaco has introduced no evidence that publishers systematically photocopy entire copyrighted articles without permission, and there is no evidence whatsoever concerning the nature, purpose and frequency of any such copying. 18. See Justice Blackmun's reference to the "recent advent of inexpensive and readily available copying machines" as having "changed the dimensions of the problem" of harm to the copyright owner. Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 467 n.16 (1984) (dissent). A description of the history and development of modern photocopying is provided in Van Dyk Research Corp. v. Xerox Corp., 478 F. Supp. 1268, 1272-1279, 1285, 1288-1302 (D.N.J. 1979), aff'd, 631 F.2d 251 (3d Cir. 1980), cert. denied, 452 U.S. 905 (1981). 19. See, e.g., 1983 Report of the Register of Copyrights at 160, 161-62, App. IV at 162 (comments by attorney for Academic Press's parent, HBJ), App. VI at 37, 99-100 (comments regarding negative impact of photocopying on small publishers of scholarly works and decline in sales of offprints since 1969); Copyright Law Revision, 1975: Hearings Before the House Subcomm. on Courts, Civil Liberties, and the Administration of Justice, 94th Cong., 1st Sess. 239, 311 (1975) (Association of American Publishers testimony regarding attempts to establish a photocopying permissions clearinghouse and HBJ comments concerning the need for compensation for photocopying); Copyright Law Revision, 1973: Hearings on S. 1361 Before the Senate Subcomm. on Patents, Trademarks, and Copyrights, 93d Cong., 1st Sess. 127, 129, 131, 139, 149 (1973) (testimony by several publishers in opposition to a proposed statutory provision permitting photocopying of single journal articles); Copyright Law Revision, 1967: Hearings on S. 597 Before the Senate Subcomm. on Patents, Trademarks, and Copyrights, 90th Cong., 1st Sess. 975 (1967) (testimony on economic harm to publishers from photocopying); Copyright Law Revision, 1965: Hearings Before Subcomm. No. 3 of the House Comm. on the Judiciary, 1431, 1460, 1513 (1965) (same). 20. D.C. Comics, Inc. v. Reel Fantasy, Inc., 696 F.2d 24, 28 (2d Cir. 1982); Meeropol v. Nizer, 560 F.2d 1061, 1068 (2d Cir.), cert. denied, 434 U.S. 1013 (1977); see Weissmann v. Freeman, 868 F.2d at 1326 (2d Cir.). 21. As to these matters, Texaco makes the following statement (T Br. 12) as its introduction to Judge Leval's decision: Without analyzing the circumstances surrounding the copying of any of the eight articles and rarely referring to any part of the `voluminous submitted record' (A 146), Judge Leval held that Dr. Chickering's copying was not fair use. 22. Harper & Row, 471 U.S. at 547. Association of Am. Med. Coll. v. Cuomo, 928 F.2d 519, 523 (2d Cir.), cert. denied, 112 S. Ct. 184 (1991); Rosemont Enters. v. Random House, Inc., 366 F.2d 303, 306 (2d Cir. 1966), cert. denied, 385 U.S. 1009 (1967). 23. Consumers Union of U.S., Inc. v. General Signal Corp., 724 F.2d 1044, 1048 (2d Cir. 1983), cert. denied, 469 U.S. 823 (1984); Association of Am. Med. Coll. v. Cuomo, 928 F.2d at 523 (2d Cir.). 24. See Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 448 (1984); Harper v. Row, 471 U.S. at 560; LA 120-21. 25. The ordinary meaning of the word "commercial" is an activity relating to commerce or trade, looking toward financial profit or having profit as the primary aim. Webster's New International Dictionary (2d ed. 1934), (3d ed. 1964); Webster's New Twentieth Century Dictionary (2d. ed. 1979); Webster's Ninth New Collegiate Dictionary (1988); Oxford English Dictionary Vol. II (1978). 26. See Telerate Sys., Inc. v. Caro, 689 F. Supp. 221, 229 (S.D.N.Y. 1988) ("customary price" was the price normally charged by the plaintiff); New Line Cinema Corp. v. Bertlesman Music Group, Inc., 693 F. Supp. 1517, 1526 (S.D.N.Y. 1988) (finding that defendant "stands to profit financially by using Freddy [the copyrighted work] without making the usual licensing arrangements. . . ."). Here, the customary price for obtaining additional original journal articles or copies of such articles depends on the available means selected by the user to obtain those originals or copies. Thus, the customary price may be the licensing or permission fees charged through CCC-administered services, the fee under bilateral or blanket licensing agreements, the cost of purchasing photocopies from authorized document delivery services, or the cost of additional subscriptions. 27. Contrary to the contention made at AR Br. 17-18 & n.7, the fact that two scientists may carry on the same research, one for a profit-motivated purpose and one for a non-profit purpose, does not mean that the fair use result must be the same or that a different result would be anomalous. 28. News reporting: E.g., Harper & Row, 471 U.S. at 562; Iowa State Univ. Research Found., Inc. v. ABC, Inc., 621 F.2d at 61 (2d Cir.); Hi-Tech Video Prod., Inc. v. Capital Cities/ABC, Inc., 804 F. Supp. 950, 955 (W.D. Mich. 1992); Radji v. Khakbaz, 607 F. Supp. 1296, 1300 (D.D.C. 1985); Quinto v. Legal Times of Wash., Inc., 506 F. Supp. 554, 560 (D.D.C. 1981); Pacific & Southern Co. v. Duncan, 774 F.2d at 1496. Educational purposes: E.g., Basic Books, Inc. v. Kinko's Graphic Corp., 758 F. Supp. 1522, 1531-32 (S.D.N.Y. 1991); Association of Am. Med. Coll. v. Mikaelian, 571 F. Supp. 144, 152- 53 (E.D. Pa. 1983), aff'd, 734 F.2d 3 (3d Cir. 1984); Educational Testing Servs. v. Katzman, 793 F.2d 533, 543 (3d Cir. 1986); Allen- Myland, Inc. v. IBM Corp., 746 F. Supp. 520, 534 (E.D. Pa. 1990); Weissmann v. Freeman, 868 F.2d at 1324 (2d Cir.). In particular, the verbatim copying of copyrighted works for use by students has not been a justification for finding fair use. Basic Books, Inc. v. Kinko's Graphics Corp., 758 F. Supp. at 1531; Marcus v. Rowley, 695 F.2d 1171 (9th Cir. 1983); Wihtol v. Crow, 309 F.2d 777 (8th Cir. 1962); Encyclopaedia Britannica Educ. Corp. v. Crooks, 542 F. Supp. 1156 (W.D.N.Y. 1982); Macmillan Co. v. King, 223 F. 862 (D. Mass. 1914). 29. These are the kind of cases Congress contemplated would result in fair use protection. In legislative reports concerning 107, it is indicated that "the sort of activities the Courts might regard as fair use" involve either the taking of a small portion of a copyrighted work (for example, the "quotation of short passages in a scholarly or technical work") or the creation of a new work. H.R. Rep. No. 94-1476, 94th Cong., 2d Sess. 65 (1976); S. Rep. No. 94- 473, 94th Cong., 2d Sess. 62 (1975). 30. Wright v. Warner Books, Inc., 953 F.2d at 740 (2d Cir.); Lish v. Harper's Mag. Found., 807 F. Supp. 1090, 1100 (S.D.N.Y. 1992). 31. That a work is creative weighs against a finding of fair use under the second factor. See Harper & Row, 471 U.S. at 563. 32. Sony, 464 U.S. at 450: Rogers v. Koons, 960 F.2d at 311 (2d Cir.); Weissmann v. Freeman, 868 F.2d at 1325 (2d Cir.) (quoting 3 Nimmer on Copyright, 13.05[A] at [13-102.24-.25 (1992)]; Hustler Mag., Inc. v. Moral Majority, Inc., 796 F.2d 1148 (9th Cir. 1986); Pacific & Southern Co. v. Duncan, 744 F.2d 1490, 1497 (11th Cir. 1984), cert. denied, 471 U.S. 1004 (1985); Strauss v. Hearst Corp., 1988 Copyright L. Dec. (CCH) 26,244 at p. 21,721 (S.D.N.Y. 1988). It is only in an exceptional case like Sony or Williams & Wilkins that copying an entire work has been found to be fair use, and in each case the court found that the copying was done for a non-profit, non-commercial purpose and there was no harm to the copyright owner. 33. Registration of a collective work includes registration of the copyright in each individual article for which copyright ownership has been transferred to the registrant, as is made clear in publications of the Copyright Office, copies of which (Circular 62, Circular R62, Instructions for Short Form SE) were submitted to Judge Leval as annexes, with relevant portions marked, to Plaintiffs' Post-Trial Sur-Reply Memorandum Concerning Fair Use, which is included in the record on appeal. See Gordon & Breach, Science Pubs., Inc. v. Information on Demand, Inc., No. C-88-1695 EFL, slip. op. at 2-3 (N.D. Cal. Sept. 8, 1989). Because the eight articles were properly registered, the jurisdiction and standing arguments raised by the ALA (AL Br. 15-15 & n.7) are also meritless. 34. It is undisputed, as Judge Leval found (LA 110 n.14, 148), that photocopying of entire journal articles is widespread at for- profit, research-oriented companies. Furthermore, the "evidence submitted . . . supports the inference of voluminous photocopying in the aggregate by Texaco's scientists, that would justify very substantial license fees if such copying is not fair use." (LA 110 n.14) 35. The inference that Judge Leval drew from the facts in evidence that Texaco would purchase a modest number of additional subscriptions if it chose not to pay for permission to photocopy is a correct and permissible inference. See Schulz v. Pennsylvania R.R., 350 U.S. 523, 526 (1956) (function of the factfinder is to reach conclusions based on "proof of circumstances from which inferences can fairly be drawn"); Sidney Blumenthal & Co. v. Rossie Velvet Co., 89 F.2d 29, 31 (2d Cir. 1937) (proper for trial judge's conclusion to be based on inference drawn from proven facts); United States Lighterage Corp. v. Petterson Lighterage & Towing Corp., 142 F.2d 197, 199 (2d Cir. 1944) (same). 36. The decision in Williams & Wilkins has been severely criticized by copyright scholars, one of whom agreed with a dissenting judge in that case that it is the "Dred Scott decision of copyright law" (see LA 123 n.19). 37. Neither the Supreme Court nor the Second Circuit has ruled on the question and "[t]he copyright misuse doctrine has been ill received in the lower courts." Broadcast Music, Inc. v. Hearst/ABC Viacom Entertainment Servs., 746 F. Supp. 320, 328 (S.D.N.Y. 1990) (citing cases). The copyright misuse defense has been rejected repeatedly; courts have either doubted its validity, or assumed its validity but found its demanding requirements not met. See, e.g., United Tel Co. v. Johnson Pub. Co., 855 F.2d at 611-612 (citing cases); 3 Nimmer on Copyright 13.09[A] at 13-143-44. It has been upheld on the merits only rarely, in cases that involved obvious restraints on competition. PRC Realty Sys., Inc. v. National Ass'n of Realtors, 1992 U.S. App. LEXIS 18017 (4th Cir. Aug. 4, 1992); Lasercomb Am., Inc. v. Reynolds, 911 F.2d 970, 972-73 (4th Cir. 1990); QAD. Inc. v. ALN Assocs., Inc., 770 F. Supp. 1261 (N.D. Ill. 1991); M. Whitmark & Sons v. Jensen, 80 F. Supp. 843 (D. Minn. 1948). Here, calculation of the license fee (which is based on individually set publishers' fees) is not being used to restrain competition.