No. 97-7444.United States Court of Appeals, Second Circuit.Argued October 22, 1997.
Decided April 3, 1998.
 [EDITORS’ NOTE: THIS PAGE CONTAINS HEADNOTES. HEADNOTES ARE NOT AN OFFICIAL PRODUCT OF THE COURT, THEREFORE THEY ARE NOT DISPLAYED.]
Al J. Daniel, Jr., New York City, pro se.
Michael Newcity, Durham, NC, pro se.
Julian H. Lowenfeld, New York City, pro se and for Plaintiffs-Appellees.
William J.T. Brown, New York City (Donovan Leisure Newton
Irvine), for Plaintiffs-Appellees Moskovsky Komsomolets and AR Publishing Co., Inc.
Appeal from United States District Court for the Southern District of New York, John G. Koeltl, J.
Before: WINTER, Chief Judge, MESKILL, Circuit Judge, and POLLACK, District Judge.[*]
POLLACK, Senior District Judge:
 This Circuit’s opinion is that the district court has a “responsibility to protect its own officers in such matters as fee disputes.” Cluett, Peabody Co. v. CPC Acquisition Co., 863 F.2d 251, 256 (2d Cir. 1988).
 During the final phases of a non-jury copyright litigation, after the case was tried but while it was sub judice, appellant Al J. Daniel (“Daniel”), an attorney for plaintiffs and counsel to the principal attorney for plaintiffs, Julian H. Lowenfeld (“Lowenfeld”), moved for leave to withdraw from further representation of the plaintiffs and Lowenfeld in the suit. Appellant, Michael Newcity (“Newcity”), are expert witness for plaintiffs, moved for his fees and expenses. Specifically, appellants claimed non-payment of alleged contractual compensation due them from the plaintiffs and Lowenfeld and invoked the supplemental jurisdiction of the district court pursuant to 28 U.S.C. § 1367 to fix their fees and expenses pursuant to retainer agreements, and to fix the amount of Daniel’s charging lien under New York Judiciary Law § 475.
 The motions to withdraw as counsel and for fees, respectively, were unopposed and were granted. However, the district judge declined to exercise his supplemental jurisdiction over the motions for fees by Newcity and Daniel and dismissed them without prejudice. The district judge added that the plaintiffs and their attorneys are in no way prejudiced thereby in their claims for statutory fees under the copyright law, by his failure to resolve their internecine (sic) fee disputes. On the same day, the district judge announced that the plaintiffs were entitled to judgment against the defendants on the merits for deliberate copyright infringement and directed the submission of an appropriate judgment. The district judge further directed the plaintiffs to submit a separate application for “modest” statutory attorneys’ fees.
 Accordingly, the Judgment subsequently submitted and signed provided that “the plaintiffs may submit any appropriate application for the costs of the litigation and reasonable attorneys’ fees, as allowed by law, to be determined on the basis of applications submitted separately in accordance with Federal
Rule of Civil Procedure 54(d).” All attorneys complied.
 Appellees, and plaintiffs’ other counsel entitled to a statutory award under 17 U.S.C. § 505, filed the information required by the district court. Appellees sought a judgment requiring defendants to pay 100% of the time and expenses for which Newcity had billed plaintiffs, and for 80% of the time and expenses billed by Daniel. The statutory claim for Daniel’s fees and expenses against defendants was based upon the identical statements upon which his contractual claims against plaintiffs and Lowenfeld (except as to quantum) were based. Daniel’s statutory fee claim was voluntarily reduced by 20% in light o Hensley v. Eckerhart 461 U.S. 424, 430 and n. 3, 103 S.Ct. 1933, 1937-38 and n. 3, 76 L.Ed.2d 40 (1983), and other factors.
 On June 10, 1997, the district court issued an Opinion and Order awarding plaintiffs a total of $360,398.52 against defendants for statutory attorneys’ fees and costs. Included in said total was a statutory award for Lowenfeld of $243,948.86 which was for fees and costs including Newcity’s total unpaid claim for services as an expert witness; $65,656.58 for Daniel, consisting of all of his expenses and 50% of his statutory fee claim; $28,802.58 of legal fees to another attorney, one Trope; and $21,990.50 to yet another attorney, one Berman. Judgment was entered accordingly on July 3, 1997.
 The defendants have appealed the Judgment of liability for copyright infringement against them. That appeal is pending. The defendants did not appeal the judgment for attorneys’ fees and costs.
 Appellants Daniel and Newcity have appealed herein from the dismissal of the motion invoking supplemental jurisdiction to enforce their alleged contractual fee arrangements with the Appellees, subject of course to any offset by way of a collection of the statutory award which depends on the outcome of the Judgment against the defendants.
 The district court had obtained total familiarity with the subject matter of the suit and the professional services of the moving parties thereon and of the virtual totality of all the compensation arrangements contended for and disputed, all of which were fully disclosed on the record of the proceedings before the court. All of these would have to be considered anew and relitigated in possibly more than one state court unfamiliar with the proceedings and the trial services if supplemental jurisdiction is not exercised.
A. Supplemental Jurisdiction
 The district court declined to exercise supplemental jurisdiction over appellants’ motion for attorneys’ fees because, “the disputes over fees are entirely separate from the underlying action,” the “adjudication of [appellants’] claims would not serve the interests of judicial economy and would require considerable judicial resources,” and rendering a decision over the supplemental claim “might cast aspersions . . . and compromise some of [the parties] arguments for statutory attorney’s fees.”
 In rendering its decision, the district court made no reference to 28 U.S.C. § 1367, which provides federal judges with both the power to exercise supplemental jurisdiction and the discretion, in specified circumstances, to decline to entertain such claims. The district court’s decision to decline supplemental jurisdiction was based loosely on the discretionary factors set forth in United Mine Workers of America v. Gibbs, 383 U.S. 715, 726, 86 S.Ct. 1130, 1139, 16 L.Ed.2d 218 (1966), which emphasized that “pendent jurisdiction is a doctrine of discretion, not of plaintiff s right.” Gibbs directed federal courts, in determining whether to exercise jurisdiction over a case involving state-law claims, to “consider and weigh in each case, at every stage of the litigation, the values of judicial economy, convenience, fairness, and comity.” Carnegie-Mellon Univ. v. Cohill, 484 U.S. 343, 350, 108 S.Ct. 614, 619, 98 L.Ed.2d 720 (1988). See generally Executive Software N. Am., Inc. v. United States Dist. Court, 24 F.3d 1545, 1552-55 (9th Cir. 1994) (discussing the “flexibility” of the Gibbs doctrine and the criticism resulting from that flexibility).
 The primary issue for this court is whether section 1367 altered the Gibbs analysis. Section 1367 reads, in pertinent part, as follows:
(a) Except as provided in subsection . . .
(c) . . ., in any civil action of which the district courts have original jurisdiction, the district courts shall have supplemental jurisdiction over all other claims that are so related to claims in the action within such original jurisdiction that they form part of the same case or controversy under Article III of the United States Constitution. . . .
(c) The district courts may decline to exercise supplemental jurisdiction over a claim in subsection (a) if —
(1) the claim raises a novel or complex issue of State law,
(2) the claim substantially predominates over the claim or claims over which the district court has original jurisdiction,
(3) the district court has dismissed all claims over which it has original jurisdiction, or
(4) in exceptional circumstances, there are other compelling reasons for declining jurisdiction.
 Facially, the statute appears to have transformed the Gibbs
analysis in important respects. See Executive Software, 24 F.3d at 1555-57 (discussing differences and similarities between Gibbs
and section 1367). Nowhere within subsection 1367(c), for example, are the underlying criteria of judicial economy, convenience, fairness, and comity mentioned. We have previously stated that “the discretion implicit in the word `may’ in subdivision (c) of § 1367 permits the district court to weigh and balance several factors, including considerations of judicial economy, convenience, and fairness to litigants.” Purgess v. Sharrock, 33 F.3d 134, 138 (2nd Cir. 1994) (citations omitted). In other words, this Circuit has recognized that the Gibbs
factors are merged into the subsection 1367(c) analysis, but it has not stated exactly how they are merged.
 Other circuits have answered this question to varying degrees and with varying results. The Seventh Circuit has taken the approach that subsection 1367(c) merely incorporates the discretionary factors set forth in Gibbs and its progeny. See Brazinski v. Amoco Petroleum Additives Co., 6 F.3d 1176, 1182 (7th Cir. 1993). The First, Third, and D.C. Circuits have taken a similar approach. See Rodriguez v. Doral Mortgage Corp., 57 F.3d 1168, 1177 (1st Cir. 1995) (“the district court, in reaching its discretionary determination on the jurisdictional question, will have to assess the totality of the attendant circumstances”) Borough of West Mifflin v. Lancaster, 45 F.3d 780, 788 (3d Cir. 1995) (“Section 1367(c) . . . was intended simply to codify the preexisting pendent jurisdiction law, enunciated in Gibbs and its progeny. . . .”); Diven v. Amalgamated Transit Union Local 689, 38 F.3d 598, 601 (D.C. Cir. 1994) (“Despite Congress’ use of `shall’ [in subsection 1367(a)], the statute fairly exudes deference to judicial discretion — at least once the threshold determinations have been met and the court moves on to consider the exceptions.”).
 The Ninth Circuit has adopted a different approach. The Ninth Circuit has held that by listing the four circumstances in subsection 1367(c), Congress intended to cabin the previous common law analysis. See Executive Software N. Am., Inc. v. United States Dist. Court, 24 F.3d 1545, 1556-57 (9th Cir. 1994). Once a court identifies one of the factual predicates which corresponds to one of the subsection 1367(c) categories, the exercise of discretion “is informed by whether remanding the pendent state claims comports with the underlying objective of `most sensibly accommodat[ing]’ the values of `economy, convenience, fairness, and comity.'” Id. at 1557 (citation omitted). In addition, the Court found that the catch-all created by subsection 1367(c)(4) should be employed when circumstances are “quite unusual.” 24 F.3d at 1558. This would require a district court to “articulate why the circumstances of the case are exceptional in addition to inquiring whether the balance of the Gibbs values provide compelling reasons for declining jurisdiction in such circumstances.” Id.
 The Ninth Circuit’s model has been adopted by the Eighth and Eleventh Circuits. See McLaurin v. Prater, 30 F.3d 982, 985 (8th Cir. 1994) (“[Subsection 1367(c) I plainly allows the district court to reject jurisdiction over supplemental claims only in the four instances described therein.”); See Palmer v. Hospital Auth., 22 F.3d 1559, 1569 (11th Cir. 1994) (“The breadth of discretion afforded federal courts in these cases has been codified by section 1367(c). . . . [W]hile supplemental jurisdiction must be exercised in the absence of any of the four factors of subsection 1367(c), when one or more of these factors is present, the additional Gibbs considerations may, by their presence or absence, influence the court in its decision concerning the exercise of such discretion.”).
 We now adopt the interpretation of Section 1367 set forth i Executive Software, and hold that section 1367 has indeed altere Gibbs’ discretionary analysis. Initially, we note that the Ninth Circuit’s approach is consistent with the terms and structure of section 1367. On the one hand, subsection 1367(a) uses the term “shall” which implies that once it is determined that a supplemental claim is related to the claim within the court’s original jurisdiction such that they form the same case or controversy, supplemental jurisdiction over the related claim is mandatory. On the other hand, the use of “may” in subsection 1367(c) seems to confer on federal courts at least some discretion to not hear claims over which there is supplemental jurisdiction in the enumerated circumstances. If subsection 1367(c) were merely to incorporate Gibbs, subsection 1367(c) would be rendered superfluous. See United States v. Menasche 348 U.S. 528, 538-39, 75 S.Ct. 513, 519-20, 99 L.Ed. 615 (1955) (“[it is] our duty to give effect, if possible, to every clause and word of a statute”); Ohio Power Co. v. Federal Energy Regulatory Comm’n, 880 F.2d 1400, 1406 (D.C. Cir. 1989), rev’d on other grounds, 498 U.S. 73, 111 S.Ct. 415, 112 L.Ed.2d 374 (1990).
 This interpretation not only comports with the text and structure of the statute, but is also consistent with the legislative history. On July 26, 1990, Representative Kastenmeier and Representative Moorhead introduced House Bill 5381, which contained a supplemental jurisdiction proposal. The proposal sought to circumscribe judicial discretion to dismiss supplemental claims which they have the power to entertain by codifying the three examples mentioned in Gibbs. See Federal Courts Study Committee Implementation Act and Civil Justice Reform Act: Hearings on H.R. 5381 and H.R. 3898 Before the Subcomm. on Courts, Intellectual Property, and the Administration of Justice of the House Comm. on the Judiciary, 101st Cong., 2d Sess. (1990) at 30 [hereinafter “Hearings”].
 At the hearings on the bill, Judge Joseph Weis, Jr., former chair of the Federal Courts Study Committee, proposed a substitute which contained four grounds for discretionary dismissal: when the supplemental claim “raises a complex or novel issue of State law;” when the non-federal claim “predominates” over the federal claim; when the judge has dismissed the federal claim; and when “there are other appropriate reasons, such as judicial economy, convenience, and fairness to litigants, for declining jurisdiction.” Hearings at 98. By using language derived directly from Gibbs, it appears that Judge Weis was, in effect, attempting to codify the Gibbs discretionary prong. A proposal introduced on September 11, 1990 by Professors Rowe, Burbank, and Mengler also
contained these four bases for dismissal. Hearings at 722.
 On September 13, 1990 the subcommittee met and favorably reported House Bill 5381, which adopted Judge Weis’ proposal as to subsection 1367(c), but contained two changes. First, they reinserted “substantially” back into the phrase “substantially predominates.” Next, they narrowed the catchall exception by restricting it to “exceptional circumstances [when] there are other compelling reasons for declining jurisdiction.” H.R.Rep. No. 734, 101st. Cong., 2d Sess. at 11. Thus, what remained is not simply a codification of Gibbs.
 In his Opinion and Order dated March 10, 1997, the district judge gave as reasons for declining supplemental jurisdiction concerns of judicial economy, convenience, fairness,
and the possibility of the casting of “aspersions” against the parties. However, the discretion to decline supplemental jurisdiction is available only if founded upon an enumerated category of subsection 1367(c).
 None of the supplemental claims raises “a novel or complex issue of State law. . . .” 28 U.S.C. § 1367(c)(1). Appellants’ claims are straightforward and the relevant facts are easily determined. Appellees do not deny that there were agreements with both Daniel and Newcity, and that Daniel and Newcity performed legal services in the action. The only issue which remains is which of the Appellees are liable for the undisputed fees and expenses owed to the Appellants.
 Subsection 1367(c)(2) is not implicated because Appellants’ supplemental claims do not “substantially predominate” over the copyright claims. In contrast to the extensive and complex copyright litigation involved in the action, Appellants merely seek payment for services rendered and fees incurred throughout the course of the litigation.
 Subsection 1367(c)(3), which empowers a federal court to dismiss motions seeking supplemental jurisdiction in a case if it has “dismissed all claims over which it has original jurisdiction” has no bearing in this instance. The federal claims in this action were not dismissed; they were sustained five months after the motions involving supplemental jurisdiction were served.
 Subsection 1367(c)(4) authorizes federal courts, upon the recognition of “exceptional circumstances,” to decline supplemental jurisdiction. The use of “exceptional circumstances” indicates that “Congress has sounded a note of caution that the bases for declining jurisdiction should be extended beyond the circumstances identified in subsections (c)(1)-(3) only if the circumstances are quite unusual.” Executive Software, 24 F.3d at 1558. In other words, declining jurisdiction outside the ambit of 1367(c)(1)(3) appears as the exception rather than the rule. Thus, federal courts “must ensure that the reasons identified as `compelling’ are not deployed in circumstances that threaten this principle.” Id. None of the reasons cited by the district court qualify as adequate grounds for declining supplemental jurisdiction under 1367(c)(4).
B. Daniel’s Charging Lien Under New York Judiciary Law § 475
 New York Judiciary Law § 475 (“Section 475”) governs attorneys’ charging liens in federal courts sitting in New York See Markakis v, S.S. Mparmpa Christos, 267 F.2d 926, 927 (2d Cir. 1959) (Section 475 “creates an equitable right and remedy cognizable in the federal courts”); Brooks v. Mandel-Witte Co., 54 F.2d 992, 994 (2d Cir. 1932), cert. denied Mandel-Witte Co. v. Brooks, 286 U.S. 559, 52 S.Ct. 641, 76 L.Ed. 1292 (1932) (“federal courts sitting in a state have enforced
statutes of that state creating attorney’s liens whether the suit for services in which the lien was claimed was originally brought in a state court or in a federal court”).
 Moreover, the Second Circuit has “long recognized that the lien created by section 475 . . . is enforceable in federal courts in accordance with its interpretation by New York courts.” In re Chesley v. Union Carbide Corp., 927 F.2d 60, 67 (2d Cir. 1991) see also In re McCrory Stores Corp., 19 F. Supp. 691, 693 (S.D.N.Y. 1937) (“The lien being a matter of substantive law created by a statute of New York, the bounds placed on it by authoritative decisions of the New York courts are bounds on it here.”)
1. History of Section 475
 Judicial remedies “for the protection of attorneys against the knavery of their clients” took root at common law. Goodrich v. McDonald, 112 N.Y. 157, 163, 19 N.E. 649 (1889). Lord Kenyon observed that it had been “settled long ago, that a party should not run away with the fruits of the cause without satisfying the legal demands of his attorney, by whose industry, and in many instances at whose expense, those fruits are obtained.” Read v. Dupper, 101 Eng. Rep. 595, 596 (1795). The theory underlying the attorney’s lien is “in analogy to the lien which a mechanic has upon any article which he manufactures.” Williams v. Ingersoll, 89 N.Y. 508, 517 (1882).
 In New York, the beginnings of the doctrine of an attorney’s charging lien are found in 1841. At that time, the New York Court for the Correction of Errors declared that there was no doubt “that by the law of this state, as it has always existed from the time of its first settlement, the lawyer, as well as the physician, was entitled to recover a compensation for his services; and that such services were never considered here as gratuitous or honorary merely.” Adams v. Stevens Cagger, 26 Wend. 451, 455 (1841). In 1858, the Court of Appeals noted: “As in other cases of lien, the attorney is protected, because it is by his labor and skill that the judgment has been recovered. The judgment being under the control of the court, and the parties within its jurisdiction, it will see that no injustice is done to its own officers.” Rooney v. Second Ave. R. Co., 18 N.Y. 368, 369 (1858).
 The common law doctrine of an attorney’s charging lien was codified in New York in section 66 of the Code of Civil Procedure in 1876. As amended in 1899, section 66 of the Code of Civil Procedure was endowed with the same statutory language that we apply today. Section 475, enacted in 1909, was lifted bodily from section 66 of the Code of Civil Procedure after that statute’s demise:
From the commencement of an action, special or other proceeding in any court or before any state, municipal or federal department, except a department of labor, or the service of an answer containing a counterclaim, the attorney who appears for a party has a lien upon his client’s cause of action, claim or counterclaim, which attaches to a verdict, report, determination, decision, judgment or final order in his client’s favor, and the proceeds thereof in whatever hands they may come; and the lien cannot be affected by any settlement between the parties before or after judgment, final order or determination. The court upon the petition of the client or attorney may determine and enforce the lien.
 N.Y. Judiciary Law § 475 (McKinney’s 1997). Writing in 1903, the New York Court of Appeals observed that, with the 1899 amendment of section 66 of the Code of Civil Procedure, “[t]he law has made great progress in protecting members of the bar. . . .” Fischer-Hansen v. Brooklyn Heights R. Co., 173 N.Y. 492, 495, 66 N.E. 395 (1903). There, the court stated:
[T]he statute created a lien in favor of the attorney on his client’s cause of action, in whatever form it may assume in the course of the litigation, and enables him to follow
the proceeds into the hands of third parties, without regard to any settlement before or after judgment; that all the world must take notice of the lien; and that it was unnecessary for the attorney to give notice of his claim to the other party.
 Id. at 498, 66 N.E. 395 (citing Peri v. N.Y.C. H.R.R. Co., 152 N.Y. 521, 46 N.E. 849 (1897) (quotations omitted)). Moreover, the court held: “The statute is remedial in character, and hence should be construed liberally in aid of the object sought by the legislature, which was to furnish security to attorneys by giving them a lien upon the subject of the action.” Fischer-Hansen, 173 N.Y. at 499, 66 N.E. 395.
2. Attorney of Record
 Despite the liberal interpretation afforded Section 475, the New York cases make clear that the charging lien provided for by Section 475 is for the benefit of an “attorney of record” only Weinstein v. Seidmann, 173 A.D. 219, 220, 159 N.Y.S. 371 (1st Dept. 1916). See also Cataldo v. Budget Rent A Car Corp., 226 A.D.2d 574, 574, 641 N.Y.S.2d 122 (2d Dept. 1996) (“[t]he Court of Appeals has clearly stated that the emphasized language grants a lien to the `attorney of record'”) (citations omitted); In re Sebring, 238 A.D. 281, 285, 264 N.Y.S. 379 (4th Dept. 1933) (Section 475 gives a lien to “the attorney who appears for a party. It is not broad enough to include counsel; it is confined to the attorney of record”); Gary v. Cohen, 34 Misc.2d 971, 973, 231 N.Y.S.2d 394 (N.Y. Sup. Ct. 1962) (“[a]n attorney’s statutory or charging lien is confined to the attorney of record in the cause; trial counsel is not thus protected”).
 Appellees contend that Daniel was not an attorney of record in the matter and is therefore not entitled to a charging lien under Section 475. They contend that Daniel served merely “of counsel” to Lowenfeld according to Daniel’s own characterization of himself sometimes as “of counsel” in various pleadings and at oral argument on this appeal. We find that Daniel was not, in fact, merely “of counsel,” in the sense advocated by Appellees, but that he was an “attorney of record” for the purposes of Section 475.
 Rodriguez v. City of New York, 66 N.Y.2d 825, 498 N.Y.S.2d 351, 489 N.E.2d 238 (1985) offers the most intelligible statement regarding who may be called an “attorney of record” for the purposes of Section 475. There, an attorney seeking a charging lien was originally retained by the plaintiff to be her attorney in the action. Because of his advanced age and limited practice, the attorney retained a second attorney to act “of counsel” and the two agreed to split the fee evenly. However, the attorney hired to act “of counsel” in fact handled all of the pleadings in the case as well as the entire trial.
 The court held that the initially retained attorney was not entitled to a charging lien because “it is undisputed that [his] name never appeared on any of the pleadings, motion papers, affidavits, briefs or record in plaintiffs action.” Id. at 827, 498 N.Y.S.2d at 353, 489 N.E.2d at 240. The “of counsel” attorney was, in fact, the sole attorney of record in the case because he was the only one who ever appeared in the case. The fact that the initially retained attorney’s name appeared on a retainer statement in a negligence case, filed pursuant to the rules of the Appellate Division, was insufficient to qualify him as an attorney of record. The absence of his name anywhere in the record showed that he was not, as he claimed, an attorney of record.
 The restriction provided by Rodriguez on who may be an `attorney of record’ for the purposes of Section 475 has been implemented on several occasions by the New York courts. See, eg., Cataldo, 226 A.D.2d at 574, 641 N.Y.S.2d 122 (“before an attorney can be granted a lien pursuant to [Section 475] he or she must have appeared for the client by participating in a legal proceeding on the client’s behalf or by having his [or her] name affixed to the pleadings, motions, records, briefs, or other papers submitted in the matter” (quotations omitted)); Ebert v. New York City Health Hosp. Corp., 210 A.D.2d 292, 292, 619 N.Y.S.2d 756 (2d Dept. 1994). (“[T]he mere existence of a retainer agreement is insufficient to create a charging lien pursuant to [Section 475]. That statute provides that an attorney `appear’ for the client, in the sense of participating in a legal proceeding
on the client’s behalf or by having his name affixed to the pleadings, motions, records, briefs, or other papers submitted in the matter.”); Lai, Ling Cheng v. Modansky Leasing Co., 137 A.D.2d 781, 783, 525 N.Y.S.2d 328 (2d Dept. 1988) (“an attorney whose name nowhere appears in the pleadings, motion papers, affidavits, briefs or record in a plaintiffs action is not entitled to seek a . . . charging lien under [Section 475]”).
 The court in Rodriguez was not troubled, as we are not, by the parties’ use of the term “of counsel” as a further description of their relationship. The record plainly demonstrated, as it does here, that the “of counsel” attorney was, in fact, an attorney of record. The record is replete with evidence that Daniel was an “attorney of record” in this case. Daniel signed and submitted numerous papers and letters to the court, and appeared at trial, as “attorney for the plaintiffs.” There is no indication anywhere in the record that Daniel was treated by the Clerk of the Court, the Magistrate Judge or the District Court Judge as anything other than an attorney of record for plaintiffs. The JointPretrial Order, signed on the very same page by Judge Koeltl, and Messrs. Lowenfeld and Daniel, respectively, is executed as follows:
“Dated: June 20, 1996
“Julian H. Lowenfeld
Julian H. Lowenfeld
Attorney for Plaintiffs
(except Fromer Associates, Inc.
Isaac Fromer, and Unlimited Brokerage, Inc.)
“John G. Koeltl
John G. Koeltl
United States District Judge”
“Al J. Daniel, Jr.
Al J. Daniel
Attorney for Plaintiffs
(except Fromer Associates, Inc. Isaac Fromer and Unlimited Brokerage, Inc.)”
 The references to Daniel as “of counsel” made by Daniel and Lowenfeld in their submissions to the court offer little to the analysis. Nothing in the record suggests that Daniel acted simply as an advisor or as a law clerk to Lowenfeld, or that he merely assisted Lowenfeld in the case’s preparation. It is true that he appeared in the case, along with attorney Trope, some time after it had been initiated by Lowenfeld, but there is no doubt that he entered the case, allegedly with the plaintiffs’ consent, to serve as co-counsel with Lowenfeld and Trope.
 Lowenfeld’s own statements are telling. In his Memorandum in Support of Motion for Award of Costs and Reasonable Attorney’s Fees seeking statutory compensation under 17 U.S.C. § 505, Lowenfeld states: “Daniel was retained on behalf of plaintiffs a month before trial in the case, when the discovery posture of the case, the number of witnesses, exhibits, and issues involved, required that Daniel work full-time on this case, and prevented his pursuing other matters.” (emphasis added). This does not sound like a description of work by someone who is merely a legal advisor or assistant. Lowenfeld concedes in the same memorandum that he was “less experienced in litigation” than his co-counsel, Daniel and Trope. This statement begs the question of who might have in fact been assisting whom. Thus, despite what he may have called himself, Daniel was clearly not acting merely “of counsel” in this matter in the constricted sense put forth by appellees. Daniel was an attorney of record for the purposes of Section 475.
 It should be noted that Daniel’s permitted withdrawal as attorney of record does not affect his entitlement to the statutory lien under Section 475. The New York Court of Appeals has stated that “[i]t has long been held that attorneys who terminate their representation for just cause continue to be entitled to enforce their liens.” Klein v. Eubank, 87 N.Y.2d 459, 462, 640 N.Y.S.2d 443, 663 N.E.2d 599 (1996). Moreover, “an attorney’s participation in the proceeding at one point as counsel of record is a sufficient predicate for invoking the statute’s protection.” Id. In this case, Daniel properly terminated
his relationship with his clients when he was not paid as agreed.
3. More than one Attorney of Record
 Appellees also contend that there may be only one “attorney of record” in a case. They point to Kitsch v. Riker Oil Co., 23 A.D.2d 502, 503, 256 N.Y.S.2d 536 (2d Dept. 1965) which found that there “is no authority for a party to be represented by more than one attorney in an action” and Stinnett v. Sears Roebuck Co., 201 A.D.2d 362, 364, 607 N.Y.S.2d 646 (1st Dept. 1994) which held that “[w]here a case . . . does not involve special circumstances or highly complex litigation, a party may not be represented by more than one attorney of record.”
 These cases are incongruous with statements made by the New York Court of Appeals in Rodriguez, which indicate that, in New York, there may indeed be more than one “attorney of record” for the purposes of Section 475. In Rodriguez, the attorney moving to have a charging lien declared in his favor asserted that the “attorney of record” should be estopped from denying that the movant attorney was also an attorney of record. He claimed that it was the “attorney of record” who had control of all the pleadings and papers in the case and failed to place the movant attorney’s name on them. The court disagreed, stating: “There is evidence in the record . . . that respondent sent movant copies of motions and pleadings for his review and that movant apparently approved and returned them without requesting that his name be added to them as an attorney of record.” Id. at 828, 498 N.Y.S.2d at 353, 489 N.E.2d at 240. The clear implication of this statement is that the court would not have had any objection to an additional attorney of record.
 Furthermore, we have found no cases which have limited the charging lien to only one attorney in an action where both attorneys appeared together on behalf of plaintiffs in the case. Indeed, as far back as 1873, the New York Court of Appeals held that a lien could exist in favor of more than one attorney in a case. In Zogbaum v. Parker, 55 N.Y. 120 (1873), an attorney was hired by plaintiff as “attorney and counsel” and a second attorney, referred to as a “counselor of the court,” was also hired by plaintiff to assist the first attorney in the trial. It was held that the attorneys “had a valid lien upon the recovery, by virtue of their agreement with [plaintiff]” and because “[a]ttorneys have a lien for their compensation for services upon claims which they prosecute and judgments they recover, whether arising out of tort or contract.” Id. at 122. Nowhere did the court consider that one of them might not be entitled to a charging lien by virtue of the other’s like claim.
4. Equitable Assignment
 Although we find Daniel to have been an attorney of record, and thereby entitled to have had his charging lien determined by the district court under Section 475, it is worth noting that even if we were to conclude that Daniel was not an attorney of record, there are cases to suggest that he would nonetheless have been an equitable assignee of the cause of action by virtue of his agreement with Lowenfeld and plaintiffs.
 In Woodbury v. Andrew Jergens Co., 69 F.2d 49 (2d Cir. 1934) (L. Hand), the court held that, under a plaintiffs agreement to pay one third of the amount recovered from the defendant to his attorneys, one of whom was not an attorney of record, the attorney who was not of record “became by the law of New York an equitable assignee of the cause of action pro tanto,” though he had no charging lien for fees. Id. at 50. Moreover, the court held that “it can scarcely be that an equitable assignee is in a weaker position than a statutory lienor, who is often referred to as such an assignee.” Id. Thus, the distinction between an “attorney of record” and one who is “of counsel” may be of little practical significance in cases where attorneys have agreed among themselves to share in the fruits of their combined labor. As noted in In re Coleman, 87 F.2d 753 (2d Cir. 1937), “[w]hile the statute refers to the `attorney who appears for a party,’ and it has therefore been said that the attorney of record alone is entitled to a lien, it has been held by the lower state courts that the lien may be transferred or assigned.” Id. at 754.
 Similarly, in Harwood v. La Grange, 137 N.Y. 538, 540, 32 N.E. 1000 (1893), plaintiff, who had been employed as an attorney to assist as counsel in the prosecution of several cases on a contingency basis, was held to be the holder of an “equitable lien” of the amount received for such services. The court there stated that “[i]t was not important for the plaintiff to show that the agreement was made directly with the plaintiffs in those actions. It was sufficient that he was employed under the agreement made with [co-counsel], who acted, in making it, with the authority of [the clients], and on their behalf.” Id. The court held that it was “not open to doubt” that the “agreement gave the plaintiff an equitable lien on or ownership as equitable assignee in the proceeds of the action” and that it had “been so determined in many cases.” Id.
5. Failure of district court to administer Daniel’s charging lien under Section 475
 Because Daniel possessed a charging lien under Section 475 by operation of law, the district court’s task would have been simply to fix the amount of the lien under the retainers as found by the court. The fact that Daniel may have been accorded a measure of justice in this case under the Copyright Act under 17 U.S.C. § 505 does not change the result here. Section 475 “has provided a lien in all cases, and not merely where the client falls to provide some other form of security or protection, and the courts cannot themselves substitute another form of protection for that provided in the statute.” Robinson v. Rogers, 237 N.Y. 467, 472, 143 N.E. 647 (1924). If the amount of the charging lien has been fixed by agreement, as here, execution is appropriate on the judgment for the amount agreed to by the parties. Machcinski v. Lehigh Valley R. Co., 272 F. 920, 922 (2d Cir. 1921).
 Appellants were entitled to have their contract fee claims considered and decided under the district court’s supplemental jurisdiction and the appellant attorney was entitled to the district court’s determination of a statutory lien under New York Judiciary Law § 475.
 Reversed for proceedings consistent with the foregoing under 28 U.S.C. § 1367 and remanded for proceedings pursuant to New York Judiciary Law § 475.
(c) Dismissal or Remand. — If a non-federal claim in an action is asserted under subsection (a), the district court shall, within 90 days after the commencement of the action or, if later, within 90 days after the assertion of the non-Federal claim, determine whether the non-Federal claim should be dismissed or remanded. The court shall dismiss or remand the non-Federal claim if it is not a permissible claim under subsection section (a). The court may dismiss or remand the non-Federal claim if —
(1) the Federal claim is dismissed;
(2) the non-Federal claim substantially predominates over the Federal Claim; or
(3) the non-Federal claim should be tried separately.
Upon entry of an order dismissing or remanding the non-Federal claim, the district court shall file with the order a written statement of the reasons for dismissal or remand.