In the words of the district court: “This case presents the next chapter in the continuing narrative of this Circuit’s treatment *734of the fair use defense to a charge of copyright infringement.” Wright v. Warner Books, Inc., 748 F.Supp. 105, 107 (S.D.N.Y.1990). The principal question presented is whether defendants’ sparing use of creative expression from the unpublished letters and journals of the late author Richard Wright constitutes fair use as a matter of law. This question and others arise from plaintiff Ellen Wright’s appeal from an order of the United States District Court for the Southern District of New York, by Circuit Judge John M. Walker, sitting by designation, granting summary judgment, dismissing plaintiff’s copyright infringement, breach of contract, and libel claims. Although we disagree with portions of the district court’s analysis, we agree with its conclusions and therefore affirm.
BACKGROUND
This action stems from a dispute over the publication of a biography of the late African-American author Richard Wright, best known for his works Native Son and Black Boy. Plaintiff holds the copyrights in the published and unpublished works of her husband, who died in 1960. The biography, entitled Richard Wright Daemonic Genius, was written by an acquaintance of Wright, defendant Dr. Margaret Walker, and published by defendant Warner Books, Inc. in 1988.
This protracted dispute has taken three turns, each of which has narrowed the parties’ disagreements. Dr. Walker first completed a draft of her biography of Richard Wright in the early to mid-1980s. Her publisher at the time, Howard University Press, sought plaintiff’s permission in 1984 to use large portions of Wright’s unpublished and published works in the biography. Plaintiff refused. Whether due to its inability to obtain plaintiff’s consent or to other factors unrelated to this dispute, Howard University Press in 1986 decided not to publish Dr. Walker’s book. A second publisher, Dodd, Mead, then agreed to publish the book. It, too, however, later withdrew its commitment to publish the biography, for reasons irrelevant to this dispute. That version of the biography was never published.
Unable to obtain plaintiff’s consent, Dr. Walker rewrote portions of the earlier manuscript using less of Wright’s published and unpublished works. Over plaintiff’s objections, this new, expurgated version was published by Warner Books in November 1988. Plaintiff responded by bringing this lawsuit in May 1989. Her complaint challenged the biography’s use of portions of a wide range of Wright’s works: letters to Dr. Walker written in the 1930s, letters to Wright’s translator Margrit de Sabloni-ere, journal entries, the essay “I Choose Exile,” and his published works including Black Boy, Native Son, and Pagan Spain. Plaintiff claimed she was entitled to damages for copyright infringement, false designation of origin, breach of a manuscript access agreement between Yale University and Dr. Walker of which agreement plaintiff claimed to be a third-party beneficiary, and libel. She also sought a permanent injunction prohibiting further publication and distribution of the biography.
After discovery was completed, plaintiff moved for summary judgment on the copyright claims. Defendants thereafter cross-moved for summary judgment on all counts in the complaint. Finding no material factual disputes, the district court held that the four fair use factors enumerated in 17 U.S.C. § 107 all favored defendants and granted summary judgment in their favor on the copyright claim. The court dismissed plaintiff’s request for a permanent injunction and her claim that the biography’s use of Wright’s journals constituted a breach of a research agreement between Dr. Walker and Yale University’s Beinecke Library. Plaintiff voluntarily withdrew her claim of false designation of origin. The court dismissed without prejudice the state law libel claim for lack of jurisdiction.
On appeal this dispute has taken one final turn. Plaintiff has abandoned most of her original claims. She no longer challenges the biography’s use of Wright’s published works. Nor does she challenge the use of the letters written to Margrit de Sabloniere or of the essay “I Choose Ex*735ile.” She also does not challenge the district court’s decision to dismiss her libel claim. Two of plaintiffs original claims remain: (1) the biography’s use of the unpublished Wright/Walker letters and allegedly unpublished journal entries constitutes infringement, and (2) the biography’s use of the journal entries violates Dr. Walker’s research agreement with Yale University.
DISCUSSION
Our review of a district court’s summary judgment decision is de novo. Herbert Const. Co. v. Continental Ins. Co., 931 F.2d 989, 993 (2d Cir.1991). We resolve all ambiguities and draw all reasonable inferences in favor of the non-moving party. Id.
These principles apply with equal force to the question of fair use. See Maxtone-Graham v. Burtchaell, 803 F.2d 1253, 1257-58 (2d Cir.1986), cert. denied, 481 U.S. 1059, 107 S.Ct. 2201, 95 L.Ed.2d 856 (1987). Although “[f]air use is a mixed question of law and fact,” Harper & Row Publishers v. Nation Enterprises, 471 U.S. 539, 560, 105 S.Ct. 2218, 2230, 85 L.Ed.2d 588 (1985), on more than one occasion courts in this Circuit have resolved fair use determinations at the summary judgment stage. See, e.g., Maxtone-Graham, 803 F.2d 1253; Berlin v. E.C. Publications, Inc., 329 F.2d 541 (2d Cir.), cert. denied, 379 U.S. 822, 85 S.Ct. 46, 13 L.Ed.2d 33 (1964); Time Inc. v. Bernard Geis Assocs., 293 F.Supp. 130 (S.D.N.Y.1968). “‘[T]he mere fact that a determination of the fair use question requires an examination of the specific facts of each case does not necessarily mean that in each case involving fair use there are factual issues to be tried. ’ ” Maxtone-Graham, 803 F.2d at 1258 (quoting Meeropol v. Nizer, 417 F.Supp. 1201, 1208 (S.D.N.Y.1976), rev’d in part on other grounds, 560 F.2d 1061 (2d Cir.1977), cert. denied, 434 U.S. 1013, 98 S.Ct. 727, 54 L.Ed.2d 756 (1978)) (brackets inserted by Maxtone-Graham). The fact-driven nature of the fair use determination suggests that a district court should be cautious in granting Rule 56 motions in this area; however, it does not protect the copyright holder from summary disposition of her claims where there are no material factual disputes.
I. Copyright Infringement
Section 106 of the Copyright Revision Act of 1976 (“Copyright Act” or “Act”), 17 U.S.C. § 106, “confers a bundle of exclusive rights to the owner of the copyright,” Harper & Row, 471 U.S. at 546, 105 S.Ct. at 2223, among them “the right of first publication.” Salinger v. Random House, Inc., 811 F.2d 90, 95 (2d Cir.), cert. denied, 484 U.S. 890, 108 S.Ct. 213, 98 L.Ed.2d 177 (1987); see 17 U.S.C. § 106(3). Letters, like other literary works, aré entitled to copyright protection, Salinger, 811 F.2d at 94; so also are journal entries.
Two statutory exceptions circumscribe the rights of a copyright holder. First, as a threshold matter, section 102 of the Copyright Act does not extend copyright protection to ideas or facts. Harper & Row, 471 U.S. at 547, 105 S.Ct. at 2224. Only “original works of authorship fixed in any tangible medium of expression, ” 17 U.S.C. § 102(a) (emphasis added), bearing “the stamp of the author’s originality” are copyrightable. Harper & Row, 471 U.S. at 547, 105 S.Ct. at 2224. Second, the Act permits even original expression to be appropriated in certain circumstances. Section 107 states that the “fair use” of a copyrighted work does not infringe the copyright in that work. Whether use of copyrightable expression is fair is determined on a case-by-case basis within the context of the four non-exclusive factors enumerated in section 107. Harper & Row, 471 U.S. at 549, 105 S.Ct. at 2225; New Era Publications Int'l, ApS v. Henry Holt & Co. (New Era I), 873 F.2d 576, 588 (2d Cir.) (Oakes, C.J., concurring), reh’g denied, 884 F.2d 659 (2d Cir.1989), cert. denied, 493 U.S. 1094, 110 S.Ct. 1168, 107 L.Ed.2d 1071 (1990). Those' factors are: (1) “the purpose and character of the use;” (2) “the nature of the copyrighted work;” (3) “the amount and sub-stantiality of the portion used in relation to the copyrighted work as a whole;” and (4) “the effect of the use upon the potential *736market for or value of the copyrighted work.” 17 U.S.C. § 107. Before reaching the question of fair use, however, we should determine whether the biography has appropriated fact or expression.
A. Fact or Expression?
Most of the passages on which plaintiff bases her allegations of copyright infringement convey facts or ideas. The copyrighted works are ten journal entries— eight from January 1945, one from February 1945, one from September 1947—and six letters from Wright to Dr. Walker. Dr. Walker paraphrases fourteen portions of the ten journal entries. These portions are short. All but two of them are one to three sentences long. Most importantly, of the fourteen sections taken from the journal entries, only three, under a generous reading of expression, adopt Wright’s creative style. Two are quoted in the district court's opinion. See 748 F.Supp. at 110 n. 3. The remaining one involves Wright’s discussion of the literary techniques he employed in his works. Even these, however, do not involve the kind of figurative appropriation that was condemned in Salinger. See 811 F.2d at 96. Yet, for purposes of this appeal, we will treat the biography’s use of three portions of the paraphrased journal entries as borderline expression.
The biography’s use of the six Wright/Walker letters merits similar treatment. The biography copies ten brief passages from the letters and paraphrases five equally short portions of them. The paraphrasing solely communicates facts (including some ideas) relating to events in Wright’s life and mutual interests of the correspondents. Of the ten quoted sections, four bear Wright’s stamp of creativity and meet the threshold test of copyright protection. The other six tersely convey mundane details of Wright’s life and serve only to illustrate Dr. Walker's friendship with Wright. One, for example, explains Wright’s regrets at not having written earlier; another requests Dr. Walker’s permission to use one of her poems; still another requests clippings regarding a murder story; and a fourth discusses how much money Dr. Walker will need to visit New York. These examples are emblematic of the quoted passages from the letters and paraphrased portions of the letters and journals that are not entitled to copyright protection. Because we conclude that some portions of the journal entries and letters contain at least borderline expression, we disagree with those portions of the district court’s opinion that intimate that the biography takes only facts from Wright’s letters and journals. See 748 F.Supp. at 110, 114 (journal entries), 111 (letters).
Under a liberal interpretation of the fact/expression dichotomy, three paraphrased sections of Wright’s journals and four quoted portions of the Wright/Walker letters are protected by copyright. Unless the doctrine of fair use applies to these appropriations, they constitute infringement.
B. Fair Use?
1. Purpose and Character of the Use
We agree with the district court’s analysis of the first fair use factor: The purpose and character of the biography’s use of the Wright/Walker letters and of Wright’s journal entries clearly favors defendants. Id. at 108. Dr. Walker’s book is a scholarly biography. As such, it “fits comfortably within several of the statutory categories of uses” that Congress has indicated may be fair—“ ‘criticism,’ ‘scholarship,’ and ‘research.’ ” Salinger, 811 F.2d at 96; 17 U.S.C. § 107. The analytic research contained in Dr. Walker’s biography furthers the goals of the copyright laws by adding value to prior intellectual labor. Moreover, there is a strong presumption that factor one favors the defendant if the allegedly infringing work fits the description of uses described in section 107. “ ‘[I]f a book falls into one of these categories [i.e., criticism, scholarship or research], assessment of the first fair use factor should be at an end,’ ” New Era Publications Int'l, ApS v. Carol Publishing Group (New Era II), 904 F.2d 152, 156 (2d Cir.) (quoting New Era I, 884 F.2d at 661 (Miner, J., concurring in denial of rehearing in banc)), cert. denied, — U.S. -, 111 S.Ct. 297, 112 L.Ed.2d 251 (1990), even *737though, as will often be the ease, the biographer and publisher “ ‘anticipate profits,’ ” New Era II, 904 F.2d at 156 (quoting Salinger, 811 F.2d at 96).
Plaintiff raises only one challenge to the district court’s analysis.. She argues that Dr. Walker’s “bad faith” in refusing to give the letters to plaintiff or to obtain her permission to use them “indicates that the character and purpose of the Book is outside the realm of ‘criticism’ and ‘scholarship.’ ” Plaintiff correctly refers us to the Supreme Court’s observation that “'the propriety of the defendant’s conduct’ ” is relevant to the “character” of the use. Harper & Row, 471 U.S. at 562, 105 S.Ct. at 2231 (quoting 3 M. Nimmer, Copyright § 13.05[A], at 13-72 (1984)). She errs, however, in concluding that Dr. Walker acted in bad faith. Dr. Walker’s conduct bore no similarity to the sharp practices condemned in Harper & Row where “[t]he trial court found that The Nation knowingly exploited a purloined manuscript.” See Harper & Row, 471 U.S. at 563, 105 S.Ct. at 2232. Here the letters were sent to Dr. Walker, not stolen by her. As the recipient of the letters, Dr. Walker was entitled to keep them. See Salinger, 811 F.2d at 94-95. In any event, Dr. Walker actually gave plaintiff the six letters well before the biography was published. Equally meritless is plaintiff’s claim that Dr. Walker’s failure to get plaintiff’s permission to use the letters affects this analysis. As the district court aptly observed, the lack of permission is “beside the point” as long as Dr. Walker’s use meets the standards of fair use, 748 F.Supp. at 108, a question we now examine further with regard to the remaining factors.
2. Nature of the Copyrighted Work
The district court held that factor two favored the defendants with respect to the Wright/Walker letters, even though the letters were unpublished. In support of its conclusion, the court noted that (1) Dr. Walker paraphrased the letters, and (2) Dr. Walker “used the letters not to recreate Wright’s creative expression, but simply to establish facts necessary to her biography.” 748 F.Supp. at 111. The court also ruled for the defendants on the journal entries. Although recognizing that most of the journal entries had “not previously appeared in print,” id. at 110, the court determined that factor two “favors neither side overwhelmingly, but on balance, I conclude that it favors defendants.” Id. at 111. The court rested its conclusion principally on three grounds: (1) Dr. Walker “paraphrased, rather than directly quoted, Wright’s work;” (2) the paraphrasing involved “straightforward factual reportage;” and (3) no privacy interests were implicated in view of Wright’s death in 1960. Id. at 110-11. We disagree with this analysis.
Unpublished works are the favorite sons of factor two. “The fact that a work is unpublished is a critical element of its ‘nature.’ ” Harper & Row, 471 U.S. at 564, 105 S.Ct. at 2233. The “ ‘scope of fair use is narrower with respect to unpublished works’ because ‘the author’s right to control the first public appearance of his expression weighs against such use of the work before its release.’ ” New Era I, 873 F.2d at 592 (Oakes, C.J., concurring) (quoting Harper & Row, 471 U.S. at 564, 105 S.Ct. at 2232). In the context of biographers’ use of unpublished letters, we have narrowed the inquiry further. We concluded a one paragraph analysis of the issue in New Era I in this fashion: “Where use is 'made of materials of an ‘unpublished nature,’ the second fair use factor has yet to be applied in favor of an infringer, and we do not do so here. ‘Since the copyrighted letters are unpublished, the second factor weighs heavily in favor of [New Era].’ ” New Era I, 873 F.2d at 583 (quoting Salinger, 811 F.2d at 97). Our precedents, then, leave little room for discussion of this factor once it has been determined that the copyrighted work is unpublished.
Against this backdrop, there are three problems with the district court’s analysis. First and foremost, the court gave insufficient weight to the unpublished status of the letters and journal entries. Second, as we indicated earlier, some of the appropriated passages conveyed Wright’s expres*738sive language. Third, the court’s rationales are not relevant to factor two. To be sure, whether the infringer paraphrased or copied, whether he borrowed fact or expression, or whether his use implicates the author’s privacy interests or not, all may enter into the infringement equation. They just have no bearing on factor two. Factor two focuses solely on the nature of the copyrighted work. The court’s explanations apply to other aspects of the analysis and cannot be used in piggyback fashion to hold together a weak link in the fair use calculation. Thus, while these aforementioned three concerns may, and do in this case, help to overcome the burden placed on defendants who seek to justify use of unpublished materials, they do not figure into the factor two inquiry. See New Era I, 873 F.2d at 593 (Oakes, C.J., concurring). We believe that factor two favors the plaintiff.
3. Amount and Substantiality of the Portion Used
The district court determined that factor three favored the defendants. We agree. This factor addresses “the amount and substantiality of the portion used in relation to the copyrighted work as a whole.” 17 U.S.C. § 107(3). We examine the volume and substantiality of the work used with reference to the copyrighted work, not to the allegedly infringing work. Harper & Row, 471 U.S. at 564-65, 105 S.Ct. at 2232-33; New Era II, 904 F.2d at 158. But see Harper & Row, 471 U.S. at 565-66, 105 S.Ct. at 2232-33; Salinger, 811 F.2d at 99 (considered infra). Both direct quotes and close paraphrases count as being “used.” Salinger, 811 F.2d at 97-98. (This factor also has both a quantitative and a qualitative element to it. New Era II, 904 F.2d at 158. As a result, the factor has favored copyright holders where the portion used formed a significant percentage of the copyrighted work, see, e.g., Salinger, 811 F.2d at 98 (one-third of seventeen letters and ten percent of forty-two letters used) or where the portion used was “ ‘essentially the heart of’ ” the copyrighted work, Harper & Row, 471 U.S. at 565, 105 S.Ct. at 2233 (quoting the district court below, 557 F.Supp. 1067, 1072 (S.D.N.Y.1983)). See New Era II, 904 F.2d at 158.
Quantitatively, the district court found that Dr. Walker used no more than one percent of the Wright/Walker letters or of the journal entries. 748 F.Supp. at 112. While this percentage may be higher with respect to the letters alone, it is clear that Dr. Walker utilized a very small portion of those letters. Plaintiff does not address the district court’s analysis on this score or offer any rationale why, from a quantitative perspective, the minimal percentage of the works taken should favor her.
Qualitatively, plaintiff places undue emphasis on a quoted passage from one letter and a paraphrase from a journal entry. True, the fifty-five word passage from the letter gives a reflective account of Wright’s views on the art of writing, and the paraphrase conveys some expression regarding Wright’s views of his development as a writer. The quoted passage is indeed stylistic. However, it is the only quoted piece of expression that represents anything close to the central point communicated in any of the letters. The journal entry, although it admittedly deals with a fascinating topic, is carefully paraphrased and in no way preempts the other journal entries. These passages, along with the other five that merit copyright protection, have no plausible parallel in the critical passages taken from President Ford’s memoir discussing his decision to pardon President Nixon, which indeed were the “heart” of that copyrighted work. See Harper & Row, 471 U.S. at 564-65, 105 S.Ct. at 2232-33. Moreover, in contrast to Harper & Row, it is unclear whether the letters and journal entries in this case even have an “identifiable core that could be appropriated.” Maxtone-Graham, 803 F.2d at 1263.
There is some confusion whether factor three also should be examined in relation to the allegedly infringing work. New Era II flatly rejected the view that factor three should be considered in relation to the work accused of infringement, yet recognized that prior cases have analyzed the issue from this perspective. See, e.g., Harper & Row, 471 U.S. at 565-66, 105 S.Ct. at 2233-*73934 (quotes played a central role in infringing work and made up thirteen percent of it); Salinger, 811 F.2d at 99 (quoted sections “ma[de] the book worth reading”). It therefore briefly discussed factor three in this context. See 904 F.2d at 159. We agree with New Era II that the language of section 107 does not direct us to examine factor three in relation to the infringing work. However, because our precedents have applied this gloss to factor three, and because this perspective gives an added dimension to the fair use inquiry, we too briefly consider the amount and substantiality of the protected passages in relation to the work accused of infringement. From a quantitative perspective, the expressive portions comprise a slight fraction of the biography — at most, two pages of a 428 page book. The qualitative picture is no brighter for plaintiff. The borrowed expression enhances Dr. Walker’s analysis and serves to establish her credibility as one who, having known Wright, has a unique insight into his career. Yet it does not “make the book worth reading.” See Salinger, 811 F.2d at 99. Factor three favors defendants.
4. Effect on the Market
Finally, we agree with the district court’s conclusion that factor four also favors defendants. The fourth factor focuses on “the effect of the use upon the potential market for or value of the copyrighted work.” 17 U.S.C. § 107(4). This, the Supreme Court has said, is “undoubtedly the single most important element of fair use.” Harper & Row, 471 U.S. at 566, 105 S.Ct. at 2233. Analysis of this factor requires us to balance “ ‘the benefit the public will derive if the use is permitted and the personal gain the copyright owner will receive if the use is denied.’” 748 F.Supp. at 112 (quoting MCA, Inc. v. Wilson, 677 F.2d 180, 183 (2d Cir.1981)). The effect on the market must be attributable to the seven instances in which the biography takes Wright’s expression. Disclosure of factual content, as we have shown, is not proscribed by the copyright monopoly.
Dr. Walker’s biography does not pose a significant threat to the potential market for Wright’s letters or journals. In contrast to Salinger and New Era I, marginal amounts of expressive content were taken from Wright’s works. And what was taken represents only a small, unfeatured portion of the biography. The biography in no way supplants Wright’s letters and journals. Impairment of the market for these works is unlikely.
Plaintiff insists that she has shown potential market impairment on the grounds that she and two co-editors agreed with Harper & Row in 1969 to publish a collection of Wright’s letters, which plaintiff claims would include the Wright/Walker letters. The existence of an agreement to publish the letters, however, does not compel a finding that this factor favors plaintiff. While the agreement apparently remains in effect, little has been done on the project in two decades, even though copies of the Wright/Walker letters were given to plaintiff more than five years ago. Plaintiff has offered no evidence that the project will go forward. What evidence there is suggests the contrary. In a 1979 letter to Harper & Row, plaintiff wrote that the project should wait until she obtained Wright’s letters to Walker, Ralph Ellison, and George Padmore. In the letter she placed particular emphasis on the importance of obtaining the Ellison and Padmore letters. Ellison, however, has refused to give plaintiff the letters in his possession, and the Wright/Padmore letters have been lost. Publication of the collection thus seems highly improbable. However, even if the project were to go forward, we can see no likelihood of harm. The six Walker letters presumably would represent a minor component of the collection and, in any event, Dr. Walker’s sparing use of them has done little, if anything, to supplant them. In fact, “it is not beyond the realm of possibility that [Dr. Walker’s] book might stimulate further interest in [Wright’s letters].” Maxtone-Graham, 803 F.2d at 1264.
Plaintiff’s final point under this category merits brief attention. She argues that the district court incorrectly categorized the letters as published and concluded, as a *740result, that the letters “have already secured that section, or at least some part, of the market plaintiff might hope to target.” 748 F.Supp. at 113. The alleged publication occurred when Dr. Walker printed portions of some of the letters in the journal New Letters in 1971. The court thus should not have designated all of the letters as published. Moreover, it is doubtful that those letters, having been disseminated by the alleged infringer without plaintiffs permission, should be given “published” status. “[T]he consent of the copyright owner is essential to a finding of ‘publication.’ ” Testa v. Janssen, 492 F.Supp. 198, 202 (W.D.Pa.1980) (citing Bartok v. Boosey & Hawkes, Inc., 523 F.2d 941, 945 (2d Cir.1975)). This error, however, does not undercut our earlier conclusions. Factor four favors the defendants.
The district court correctly held that defendants were entitled to summary judgment. Three of the four fair use factors clearly favor the defendants. The one that does not — the nature of the copyrighted work — raises an obstacle to this conclusion, but not an insurmountable one. In Salinger, we held that unpublished works “normally enjoy complete protection against copying any protected expression.” 811 F.2d at 97 (emphasis added); see Harper & Row, 471 U.S. at 555, 105 S.Ct. at 2228 (“Under ordinary circumstances, the author’s right to control the first public appearance of his undisseminated expression will outweigh a claim of fair use.”) (emphasis added). Neither Salinger, Harper & Row, nor any other case, however, erected a per se rule regarding unpublished works. The fair use test remains a totality inquiry, tailored to the particular facts of each case. Because this is not a mechanical determination, a party need not “shut-out” her opponent on the four factor tally to prevail.
Weighing the amalgam of relevant factors, we are convinced that defendants’ use of Wright’s works is fair. Dr. Walker’s biography of Richard Wright is a scholarly work, one that surely will contribute to the public’s understanding of this important Twentieth Century novelist. The book does not exploit the literary value of Wright’s letters or journals. Nor does it diminish the marketability of Wright’s letters or journals for future publication. While the biography draws on works that we have characterized as unpublished for the purposes of this appeal, it takes only seven protected segments from Wright’s letters and journals. These portions are short and insignificant, with the possible exception of a fifty-five word description of the art of writing. This use is de minimis and beyond the protection of the Copyright Act. Cf. Salinger, 811 F.2d at 96 (“If he copies more than minimal amounts of (unpublished) expressive content, he deserves to be enjoined.”). In short, this is not a reprise of Salinger and New Era I. The biography’s use of Wright’s expressive works is modest and serves either to illustrate factual points or to establish Dr. Walker’s relationship with the author, not to “enliven” her prose. Because there are no material factual disputes that impede our determination of fair use, we conclude that it was appropriate to grant defendants’ motion for summary judgment and dismiss plaintiff’s claim of copyright infringement.
II. Third-Party Beneficiary Contract Claim
Plaintiff also contends that Dr. Walker’s use of Wright’s journals violates an agreement between Dr. Walker and Yale University’s Beinecke Library. Plaintiff brings this claim as a third-party beneficiary of the contract — as owner of the copyrights in the materials covered by the contract. Yale obtained the materials in the Wright Archives, including Wright’s journals, from plaintiff for $175,000. The record reveals no direct evidence that Dr. Walker signed the library agreement at issue. For purposes of considering defendants’ motion for summary judgment on this claim, however, we will assume that Dr. Walker signed the agreement.
The agreement requires the scholar to abide by, among other rules, this one: “Yale University Library manuscripts may not be published in whole or in part unless such publication is specifically authorized.” *741Dr. Walker’s use of Wright’s journals did not breach this agreement. The agreement restricts wholesale or partial “publication” of manuscripts. It does not speak to paraphrasing of facts or ideas, nor even of expression, contained in the manuscripts. As we indicated earlier, Dr. Walker did not quote from the Wright journals. Her careful paraphrasing, moreover, was almost exclusively factual, with the exception of less than a handful of insignificant instances where she adopted borderline creative expression. It defies common sense to construe this agreement as giving scholars access to manuscripts with one hand but then prohibiting them from using the manuscripts in any meaningful way with the other. Plaintiff has offered no evidence that Yale intended the agreement to be so pointless.
Plaintiff also has failed to present any evidence indicating that Yale intended this agreement to preclude fair use. Judge Walker found that “[t]he agreement does not appear to have contemplated the exclusion of fair use by biographers.” 748 F.Supp. at 114. Construing similar agreements, Judge Leval reached a like conclusion in Salinger:
[T]he intention and scope of [the] agreements are to protect the literary property interests of the owner and not to give the copyright owner an arbitrary power to block legitimate non-infringing use.
_ [Tjhis restriction should be understood as applying only to quotations and excerpts that infringe copyright. To read them as absolutely forbidding any quotation, no matter how limited or appropriate, would severely inhibit proper, lawful scholarly use and place an arbitrary power in the hands of the copyright owner going far beyond the protection provided by law.
650 F.Supp. 413, 427 (S.D.N.Y.1986), rev’d on other grounds, 811 F.2d 90 (2d Cir.1987). Here, where the material was paraphrased — not even closely in most instances — it seems particularly paradoxical to construe the agreement as conferring greater protection on the copyright holder than that provided by the Copyright Act. As Judge Walker accurately stated: “[A]ny more restrictive interpretation of the agreement ... would amount to a finding that Yale University sought to prevent the airing of historical facts rather than the unfair exploitation of another’s creative imagination or style — a finding at odds with the very purpose of a great university.” Wright, 748 F.Supp. at 114. In sum, because the agreement’s reference to “publication” does not on its face refer to paraphrasing, and because plaintiff has offered no evidence or convincing rationale why the agreement would prohibit fair use, we affirm the district court’s decision to dismiss plaintiff’s third-party beneficiary claim.
CONCLUSION
The judgment of the district court is affirmed.