SELYA, Circuit Judge. This appeal requires us to determine whether the Indian Gaming Regulatory Act, 25 U.S.C. §§ 2701-2721, 18 U.S.C. §§ 1166-1168 (1988) (the Gaming Act), applies to lands now held in trust by the United States for the benefit of the Narragansett Indian Tribe (the Tribe). This determination is tinged with more than the usual quotient of public interest, because the Tribe's ability to import casino gambling into Rhode Island likely hangs in the balance. After careful reconnaissance of a littered legal landscape, we set aside the district court's determination that the parties' dispute over the applicability of state jurisdiction is not yet ripe for adjudication and hold that Congress's grant of jurisdiction to the state in the Rhode Island Indian Claims Settlement Act of 1978, 25 U.S.C. §§ 1701-1716 (the Settlement Act), remains valid. We also hold, contrary to the Tribe's importuning, that the grant includes civil regulatory jurisdiction.
At that juncture, the tide turns. We conclude, despite the state's vehement protests, that the Gaming Act does not specially exempt the lands in question; that the Narragansetts have concurrent jurisdiction over, and exercise governmental power with respect to, those lands, and, therefore, are entitled to invoke the Gaming Act; and that, to the extent of the jurisdictional conflict between the Settlement Act and the Gaming Act, the former is impliedly repealed. In the end, we affirm both the district court's directive that Rhode Island enter into good faith negotiations to draft a tribal-state compact under which gaming operations can be mounted and its refusal to grant relief to various governmental figures and entities who have challenged the Tribe's entitlement to the extraordinary prophylaxis of the Gaming Act.
I. THE SETTLEMENT LANDS
We begin with a thumbnail sketch of how the land mass that is the breeding ground for this dispute came to be held in trust for the Tribe.
In the late 1970s, the Tribe asserted title claims to certain lands in Charlestown, Rhode Island, and, encountering resistance, pursued these claims in the federal courts. See Town of Charlestown v. United States, 696 F. Supp. 800, 801-05 (D.R.I. 1988) (recounting history of dispute), aff'd, 873 F.2d 1433 (1st Cir. 1989) (table). In 1978, the Tribe, the state, and the Town of Charlestown signed a joint memorandum of understanding (J-MEM) purporting to settle their differences. The Tribe agreed, inter alia, to the extinguishment of its title claims. In return, it obtained valuable consideration, including a lump-sum payment and effective control over roughly 1800 acres in Charlestown (the settlement lands), half donated by the state and half by private landowners.1 The titleholders agreed to deed the property to a nascent corporation which would be formed to hold title for the Tribe's benefit.
Because Congress possesses plenary power over Indian matters, see Morton v. Mancari, 417 U.S. 535, 551-52, 41 L. Ed. 2d 290, 94 S. Ct. 2474 (1974), the parties sought its blessing. In response, Congress passed the Settlement Act, a law that, for the most part, tracks the J-MEM. In 1978, the state legislature approved the Narragansett Indian Land Management Corporation Act, 6A R.I. Gen. Laws §§ 37-18-1 to 37-18-15 (1990) (the State Act), thereby creating the nominee corporation that would hold title to the settlement lands. The necessary conveyancing followed.
The next five years passed without relevant incident. Then, in 1983, the Secretary of the Interior, acting pursuant to departmental regulations, see 25 C.F.R., Part 83 (1993), officially recognized the Narragansetts as an Indian tribe. See 48 Fed. Reg. 6177-78 (Feb. 2, 1983). On the heels of federal recognition, the settlement lands changed hands twice more. In 1985, the Rhode Island General Assembly amended the State Act to permit the holding company to transfer title to the Tribe.2 The corporation complied. In September of 1988, less than a month before the Gaming Act became law, the Tribe deeded the settlement lands to the federal Bureau of Indian Affairs (the Bureau) as trustee.
II. THE GAMING ACT
The Gaming Act is an expression of Congress's will in respect to the incidence of gambling activities on Indian lands. The statute sets in place a sophisticated regulatory framework, defining a species of gambling, called "gaming," and dividing it into tiers, called "classes." Each class connotes a different level of gambling activity and, consequently, each class is regulated to a varying degree of stringency. See 25 U.S.C. §§ 2703(6) - 2703(8).
Class I gaming - which consists, essentially, of Indian ritual gambling - always can be conducted on Indian lands. See 25 U.S.C. § 2710(a)(1). Class II gaming - which encompasses bingo - can be conducted as of right on Indian lands in any state, such as Rhode Island, that does not generally proscribe activities of that type. See 25 U.S.C. § 2710(b)(1)(A). Class III gaming - a residual category that includes what is commonly thought of as casino gambling - is permitted by compact; and, moreover, a state is obliged to negotiate such a compact in good faith with a sponsoring tribe unless the state bans all persons throughout its territory from conducting class III gaming. See 25 U.S.C. § 2710(d). Short of an outright ban - and few state legislatures have indicated a willingness to go that far3 - the tribal-state compact is the exclusive method of regulating class III gaming. The method of the Gaming Act prevents a state from frustrating the introduction of class III gaming by an endless filibuster, for there are tight time parameters within which compact negotiations must be brought to fruition once a federal court finds that a state has failed to bargain in good faith. See id. § 2710(d)(7)(B). As a practical matter, then, a state ordinarily may regulate casino gambling on Indian lands only in pursuance of a consensual compact.
Because the case at bar revolves around class III gaming, the centrality of this last point cannot be overstated. One of the Gaming Act's fundamental policies is that "Indian tribes have the exclusive right to regulate gaming activity on Indian lands. . . ." Id. § 2701(5). The legislative history of the statute draws out the implications of this policy:
The mechanism for facilitating the unusual relationship in which a tribe might affirmatively seek the extension of State jurisdiction and the application of state laws to activities conducted on Indian land is a tribal-State compact. In no instance does [the Gaming Act] contemplate the extension of State jurisdiction or the application of State laws for any other purpose.
S. Rep. No. 446, 100th Cong., 2d Sess. 3, reprinted in 1988 U.S.C.C.A.N. 3071, 3075-76.
Under our tripartite system of government, Congress, not the courts, is empowered to make such policy choices. See Irons v. FBI, 811 F.2d 681, 689 (1st Cir. 1987) (acknowledging that the legislative branch sets policy by means of statutes and the courts must honor the legislature's policy choices and enforce the statutes). Thus, the courts have not focused on the wisdom of the policies underlying the Gaming Act, but have followed the legislative lead and recognized that the very structure of the Gaming Act forbids the assertion of state civil or criminal jurisdiction over class III gaming except when the tribe and the state have negotiated a compact that permits state intervention. See, e.g., United Keetoowah Band of Cherokee Indians v. Oklahoma, 927 F.2d 1170, 1177 (10th Cir. 1991); Sucuan Band of Mission Indians v. Roache, 788 F. Supp. 1498, 1504 (S.D. Cal. 1992).
III. PROCEEDINGS BELOW
On January 15, 1992, the Tribe formally requested that Rhode Island enter into good faith bargaining designed to produce a tribal-state compact that would allow construction and operation of a casino, i.e., inauguration of class III gaming, on the settlement lands. Rhode Island declined to negotiate, instead filing suit in the federal district court.4 The state asked the court to declare that the Gaming Act does not apply to the settlement lands, and that, therefore, those lands are subject to Rhode Island's general criminal and civil laws (including its civil regulatory laws). The state also sought to enjoin the development of gambling facilities on the settlement lands and to block negotiations antecedent to a tribal-state compact. The Tribe answered and counterclaimed for declaratory and injunctive relief that would pave the way for casino gambling on the settlement lands. The tribe requested, among other things, a declaration that the state's civil regulatory laws do not apply to the settlement lands; a declaration that the Narragansetts are entitled to operate a class III casino on those lands in conformance with the Gaming Act; and a mandatory injunction commanding the state to negotiate in good faith toward a compact.
The district court considered cross motions for summary judgment premised on a joint statement of uncontroverted facts. After pondering the parties' proffers, the court deferred substantive consideration of the dispute over the general applicability of state and local jurisdiction, citing ripeness concerns. See Rhode Island v. Narragansett Tribe of Indians, 816 F. Supp. 796, 799-800 (D.R.I. 1993). The court then assumed, for argument's sake, that the state had been granted jurisdiction over the settlement lands by virtue of the Settlement Act. See id. at 804. Proceeding on that assumption, the court concluded that any such grant was "preempted" by the Gaming Act, and, consequently, had no enduring force or effect. Id. Based on these findings, the court ordered the state to enter into good faith negotiations to formulate a tribal-state compact. See id. at 806.
In reaching the Conclusion that the Gaming Act controlled, the court divided its reasoning into four parts. First, it cited with approval three precedents holding that the Gaming Act overrode other federal statutes of earlier vintage. See id. at 801-02. Second, it found the evidence of Congress's intent to place Rhode Island beyond the Gaming Act's sphere to be unpersuasive in the absence of textual support in the statute. See id. at 802-04. Third, it dismissed the suggestion that the Gaming Act could not trump the Settlement Act because the former was the more general of the two statutory schemes. See id. at 804. Fourth, it determined that the Tribe "had jurisdiction" and "exercised governmental power" over the settlement lands in sufficient measure to animate the Gaming Act. See id. at 805-06.
After the plaintiffs filed a timely notice of appeal, the district court stayed its order.5
IV. THE DECISIONAL FRAMEWORK
The search for statutory meaning inevitably reduces to a pure question of law. Thus, the issues on appeal engender de novo review, to be conducted without special deference to the district court's views. See, e.g., FDIC v. Keating, 12 F.3d 314, (1st Cir. 1993) (per curiam) [No. 93-1230, slip op. at 4]; Liberty Mut. Ins. Co. v. Commercial Union Ins. Co., 978 F.2d 750, 757 (1st Cir. 1992). When a court interprets statutes that touch on Indian sovereignty, general rules of construction apply, but they must be visualized from a distinctive perspective. The Court has described this coign of vantage:
The underlying premise is that congressional intent will control. In determining this intent, we are cautioned to follow "the general rule that 'doubtful expressions are to be resolved in favor of [Indians]". . . . . But the "general rule" does not command a determination . . . in the face of congressionally manifested intent to the contrary. In all cases, "the face of the Act," the "surrounding circumstances," and the "legislative history," are to be examined with an eye toward determining what congressional intent was.
Rosebud Sioux Tribe v. Kneip, 430 U.S. 584, 586-87, 51 L. Ed. 2d 660, 97 S. Ct. 1361 (1977) (citations omitted); see also South Carolina v. Catawba Indian Band, 476 U.S. 498, 506 & n.16, 90 L. Ed. 2d 490, 106 S. Ct. 2039 (1986) (collecting cases).
Our search follows an odd trajectory in this case, because it starts down a road that the district court chose not to explore, and, once that journey is ended, proceeds to trace the path of an argument that ultimately proves to be a dead end. Although this approach is unorthodox, we think it facilitates a systematic testing of the appellants' core contention: that the settlement lands lie beyond the Gaming Act's reach.
The nature of our approach makes it desirable that we set out a roadmap. We propose, in the following two parts, to deal with the Settlement Act, for, if that statute did not confer state jurisdiction in respect to the settlement lands, or if state jurisdiction, once conferred, vanished before the Gaming Act materialized, then the state's case would necessarily founder. To this end, we discuss in Part V whether the jurisdictional inquiry is ripe; finding that it is, we discuss in Part VI the validity and scope of the jurisdiction ceded to the state by the Settlement Act.
Next, we must consider the Gaming Act's effect on the state's jurisdiction. In Part VII, we deal with - and rebuff - the state's suggestion that the settlement lands are entirely exempt from the Gaming Act. In Part VIII, we deal with - and rebuff - the state's argument that, even absent a categorical exemption, the Tribe's relationship to the settlement lands does not possess the attributes needed to trigger the Gaming Act's provisions. The final curtain falls at the Conclusion of Part IX, where we confront the interface between the Settlement Act and the Gaming Act, and test the district court's remedial order in the crucible of our understanding.
The lower court declined to resolve the issue of state and local jurisdiction, finding no "case of actual controversy" sufficient to satisfy the Declaratory Judgment Act, 28 U.S.C. § 2201 (1988). Narragansett Tribe, 816 F. Supp at 800. Though we are mindful of the deference due to a district court's decision to withhold a grant of declaratory relief, see El Dia, Inc. v. Hernandez Colon, 963 F.2d 488, 492 (1st Cir. 1992), we believe that the court abused its discretion in this instance.
A. The Applicable Standards.
When faced with questions of ripeness in the declaratory judgment context, this court employs the test developed in Abbott Laboratories v. Gardner, 387 U.S. 136, 149, 18 L. Ed. 2d 681, 87 S. Ct. 1507 (1967). The test, as phrased in our cases, contains two parts:
First, we consider whether an issue is fit for review, e.g., whether a challenged government action is final and whether determination of the merits turns upon facts which may not yet be sufficiently developed. Second, we consider the question of hardship, a question which typically turns upon whether the challenged action creates a direct and immediate dilemma for the parties.
El Dia, 963 F.2d at 495 (citation and internal quotation marks omitted); accord W.R. Grace & Co. v. United States EPA, 959 F.2d 360, 364 (1st Cir. 1992). The key consideration in this analysis "is the extent to which the claim involves uncertain and contingent events that may not occur as anticipated, or indeed may not occur at all." Lincoln House, Inc. v. Dupre, 903 F.2d 845, 847 (1st Cir. 1990) (citation and internal quotation marks omitted).
Applying this test in the declaratory judgment context often requires custom tailoring, for there are at least two salient differences between declaratory actions and the mine-run of other cases: first, declaratory relief is more likely to be discretionary; and, second, declaratory actions contemplate an "ex ante determination of rights" that "exists in some tension with traditional notions of ripeness." Step-Saver Data Systems, Inc. v. Wyse Technology, 912 F.2d 643, 647 (3d Cir. 1990). Our opinion in El Dia responded to the first of these differences. See El Dia, 963 F.2d at 491-93. We believe that our opinion today responds to the second difference.
The linchpin of ripeness under the Declaratory Judgment Act, as in all Article III cases, is adverseness. In a declaratory judgment action adverseness must be appraised in a practical, commonsense way. Thus, satisfying the adverseness requirement demands that "the facts alleged, under all the circumstances, show that there is a substantial controversy, between parties having adverse legal interests, of sufficient immediacy and reality to warrant the issuance of a declaratory judgment." Maryland Casualty Co. v. Pacific Coal & Oil Co., 312 U.S. 270, 273, 85 L. Ed. 826, 61 S. Ct. 510 (1941), citing Aetna Life Ins. Co. v. Haworth, 300 U.S. 227, 239-42, 81 L. Ed. 617, 57 S. Ct. 461 (1937). This requirement should not be applied woodenly. Most litigation has idiosyncratic features, and the adverseness criterion invites careful calibration on a case-by-case basis. The line is often difficult to draw. While a declaratory judgment should not be granted "in speculative situations," Public Affairs Assocs., Inc. v. Rickover, 369 U.S. 111, 112 (1962), a litigant "does not have to await the consummation of threatened injury to obtain preventive relief. If the injury is certainly impending that is enough." Pacific Gas & Elec. Co. v. State Energy Resources Conserv. & Dev't Comm'n, 461 U.S. 190, 201, 75 L. Ed. 2d 752, 103 S. Ct. 1713 (1983) (citation omitted).
One sound way of gauging adverseness is to evaluate the nature of the relief requested. The controversy must be such that it admits of "specific relief through a decree of conclusive character, as distinguished from an opinion advising what the law would be upon a hypothetical state of facts." Aetna Life, 300 U.S. at 240-41. Some courts call this measure of adverseness "conclusivity" and treat it as a separate requirement. See, e.g., Armstrong World Indus., Inc. v. Adams, 961 F.2d 405, 421-23 (3d Cir. 1992).
The second part of the ripeness inquiry evoked by declaratory judgment actions is concerned with the hardship to the parties that would result from a refusal to consider granting relief. We believe that this part of the inquiry should focus on the judgment's usefulness. Rather than asking, negatively, whether denying relief would impose hardship, courts will do well to ask, in a more positive vein, whether granting relief would serve a useful purpose, or, put another way, whether the sought-after declaration would be of practical assistance in setting the underlying controversy to rest. See Step-Saver, 912 F.2d at 647.
This formulation is hardly a radical departure from Abbott Laboratories and its progeny, for the one question may always be transformed into the other. For example, to say that denying relief is tolerable where an adequate state remedy has been realized is tantamount to saying that granting the requested relief in such a situation would be pointless. See El Dia, 963 F.2d at 495. Indeed, the Court some time ago observed that one reason the legal issues must be crystallized in a declaratory action is to enable the trial Judge to see "some useful purpose to be achieved in deciding them." Public Service Comm'n v. Wycoff Co., 344 U.S. 237, 244, 97 L. Ed. 291, 73 S. Ct. 236 (1952). Furthermore, framing the hardship question in a positive fashion best comports with the spirit of the Declaratory Judgment Act. As Judge Becker explained: "The idea behind the Act was to clarify legal relationships so that plaintiffs (and possibly defendants) could make responsible decisions about the future." Step-Saver, 912 F.2d at 649 (citing legislative history).
B. Applying the Standards.
Here, the district court's ruling on ripeness flowed from the notion that neither the state nor the town would have any occasion to exercise reserved jurisdiction until the compact negotiation process ended, thereby clearing the way for class III gaming. Narragansett Tribe, 816 F. Supp. at 799-800. The court seemed to focus on the uncertainty of the situation, suggesting that the need for the relief requested depended on the occurrence of speculative events. We disagree.
Whether state and local authorities retain any jurisdiction over the settlement lands is a question of immediate importance to all parties, separate and apart from the question of precisely what state and local jurisdiction survives. We think, in fairness to the parties, that the former question must be settled before they are ordered to commence negotiations for a tribal-state compact. Because the required interpretation of the Settlement Act projects a "purely legal" issue, W.R. Grace, 959 F.2d at 364, the resolution of which will not be changed by further factual development, and because it is of critical importance to the negotiation process in which the parties must engage, see infra Part IX(B), the case for a finding of adverseness is very powerful. And, moreover, while it is true that the compact negotiations may bear on the timing of class III gaming and the allocation of regulatory responsibilities, the negotiations cannot effect the existence vel non of state and local jurisdiction.
The impetus for reaching the merits is strengthened because the other characteristics traditionally associated with ripeness are also extant. We have no serious reservation about whether the proper parties are before the court or whether the requested ruling will, if granted, conclusively define the parties' baseline legal rights. By like token, such a ruling would be of great near-term utility, facilitating the course of future tribal-state compact negotiations and clarifying to some extent the legal status of the settlement lands at a time when substantially expanded use seems highly probable. Accordingly, we rule that the basic issue of state and local jurisdiction (although not the specific, fact-intensive permutations of that issue, see infra Part IX(C)) is ripe for declaratory judgment purposes.
VI. STATE AND LOCAL JURISDICTION
Addressing the merits of this issue entails an examination of the validity and scope of the Settlement Act. The Act states that, with two exceptions not relevant here,6 "the settlement lands shall be subject to the civil and criminal laws and jurisdiction of the State of Rhode Island." 25 U.S.C. § 1708. The Tribe maintains that this pronouncement was nugatory when made, or, if initially effective, was relegated to the scrap heap well before Congress enacted the Gaming Act. The Tribe also maintains that, validity aside, any grant of jurisdiction excludes civil regulatory jurisdiction, and, therefore, has no bearing upon the proposed operation of a gambling casino. We are not persuaded.
The Tribe's basic position is that, even prior to the Gaming Act, section 1708 of the Settlement Act did not constitute a valid conferral of jurisdiction because, until federal recognition occurred in 1983, the Tribe had no jurisdiction to relinquish.
This resupinate reasoning stands logic on its ear. The Tribe did not surrender jurisdiction in 1978. Rather, the Tribe, the state, and the town came to an agreement, spelled out in the J-MEM, to ask Congress, among other things, to grant jurisdiction to the state. The Tribe has articulated no reason why, regardless of its legal status, Congress lacked the power to effectuate this jurisdictional grant.
In any event, the Tribe is mistaken in its professed belief that it lacked jurisdictional power at the time of the Settlement Act. Federal recognition is just that: recognition of a previously existing status. The purpose of the procedure is to "acknowledge that certain American Indian tribes exist." 25 C.F.R. § 83.2 (1993). The Tribe's retained sovereignty predates federal recognition - indeed, it predates the birth of the Republic, see Santa Clara Pueblo v. Martinez, 436 U.S. 49, 56, 56 L. Ed. 2d 106, 98 S. Ct. 1670 (1978) - and it may be altered only by an act of Congress, see Morton, 417 U.S. at 551-52.7
The Tribe has two other arrows in its jurisdictional quiver. First, it hypothesizes that section 1708 did not survive federal recognition. This hypothesis is the mirror image of the hypothesis just considered: rather than being cast as the prerequisite for the conferral of jurisdiction, recognition is taken to have nullified that conferral. The two hypotheses suffer from the same infirmity. Tribal sovereignty (and, hence, jurisdiction) may be neither augmented nor diminished except through congressional enactment. Second, the Tribe suggests that, if section 1708 survived recognition, it did not survive the subsequent alienation of the settlement lands. This suggestion goes nowhere. Supposing that the jurisdictional grant contained in section 1708 could have been jettisoned by the state, the Tribe, or the Bureau without congressional sanction - a supposition we do not share - the fact is that, at every salient moment, the parties in interest took pains to reaffirm section 1708.8 We conclude, therefore, that the grant of jurisdiction contained in section 1708 of the Settlement Act was valid when made, and was undiluted at the time Congress passed the Gaming Act.
Validity notwithstanding, it is an open question whether the jurisdictional grant contained in section 1708 extends to civil regulatory jurisdiction. The Tribe insists not It tells us that the enacting Congress intended to copy the distinction between civil regulatory and civil adjudicatory jurisdiction limned two years earlier in Bryan v. Itasca County, 426 U.S. 373, 48 L. Ed. 2d 710, 96 S. Ct. 2102 (1976). We find this tale to be both unsupported and unsupportable.
The Tribe's argument runs along the following lines. In its view, there are salient discrepancies in respect to jurisdiction among the J-MEM, the original Senate bill leading to the Settlement Act, and the final version of the Act itself. To illustrate the Tribe's point, we list the three versions of the jurisdictional clause side by side, in order of drafting. According to three documents, the settlement lands were to be subject to:
All laws of the state . . . including but not limited to state and local building, fire and safety codes [J-MEM, P 13];
the complete civil and criminal jurisdiction of the State. . . . [Joint Hearing on S.3153 and H.R. 12860, 95th Cong., 2d Sess., at 36, 51 (June 20, 1978)];
the civil and criminal laws and jurisdiction of the State. . . . [18 U.S.C. § 1708].
Analogizing to Bryan, the Tribe posits that this progression signals Congress's intent to limit the jurisdictional grant.
This proposed interpretation finds no succor in the legislative history. Without such support, we think it is evident that the Narragansetts read too much into too little. Considering the overall context, the deviations from one document to another do not strike us as especially significant. The progressive development of the jurisdictional language can more plausibly be interpreted as intended to clarify the breadth of the grant, rather than to narrow it. Perhaps the drafters feared that "all laws of the state" might suggest regulatory jurisdiction alone, and that "civil and criminal jurisdiction" might imply only jurisdiction in the judicial sense. "Civil and criminal laws and jurisdiction" more obviously includes all sorts of jurisdiction, and can fairly lay claim to being the broadest of the three formulations.
The only change arguably suggesting a diminution in the scope of jurisdiction is the removal of the word "complete" from the draft version of the bill. We think that this change, too, may be understood as an attempt at clarification: the word "complete" could well have been removed simply to avoid any suggestion that the grant of jurisdiction was intended to be exclusive. Cf. United States v. Cook, 922 F.2d 1026, 1032 (2d Cir.) (suggesting that "exclusive jurisdiction" and "complete jurisdiction" may have the same connotation), cert. denied, 111 S. Ct. 2235 (1991). For that reason, the discrepant language is at best inconclusive.
The small changes in phraseology pinpointed by the Tribe, floated without visible means of support, place this case at considerable remove from Bryan, a case in which the Court confronted a genuinely suggestive lingual discrepancy, and interpreted the final version of the statute in line with clearly articulated legislative history. See Bryan, 426 U.S. at 379-87. Here, by contrast, the discrepancies that the Tribe perceives are more conjectural than suggestive; there is absolutely nothing in the legislative history of the Settlement Act that indicates congressional intent either to limit the scope of state jurisdiction or to carve a jurisdictional distinction along civil regulatory/civil adjudicatory lines.9 Thus, Bryan is not a fair congener.
Our assessment is reinforced by a commonsense tenet of statutory construction. Relatively minor differences between an Indian agreement and the ratifying act of Congress needed to give it effect, without more, do not give rise to an inference that Congress intended to modify the agreement. See Rosebud Sioux, 430 U.S. at 599 (holding that a 1904 act of Congress did not modify a 1901 Indian agreement, despite a suggestive minor change in language). At least when an "implied continuity in purpose" exists between the antecedent agreement and the subsequently enacted statute, courts should construe the latter to effectuate the former, notwithstanding differing linguistic choices. Id. So it is here, for the Settlement Act was designed to implement the agreement embodied in the J-MEM. See, e.g., 25 U.S.C. § 1701(d) (declaring that the J-MEM "requires implementing legislation"); Joint Hearing at 97 (acknowledging that "the legislation as drafted intends to implement the settlement agreement") (statement of Alan R. Parker, Gen. Counsel, Sen. Select Comm. on Indian Affairs).
We need not belabor the obvious. Since the self-serving inference drawn by the Tribe is plainly at odds with the discernible intention undergirding the Settlement Act, and, in the bargain, plays havoc with the statutory text, we decline gratuitously to limit the scope of section 1708 in order to parallel the holding in Bryan.10 Cf. United States v. Dakota, 796 F.2d 186, 188 (6th Cir. 1986) (refusing to extend Bryan distinction to 18 U.S.C. § 1955, because it would be inappropriate to apply a test "developed in a different context to address different concerns"). Hence, we conclude that the Settlement Act granted civil regulatory jurisdiction, as well as civil adjudicatory jurisdiction, to the state.11
C. Local Jurisdiction.
We digress to add a few words about local jurisdiction, mindful that the Town of Charles town and certain municipal officials are parties to this lawsuit.
Although we recognize both the town's desire to assert jurisdiction in respect to the settlement lands and the Tribe's opposition, we see nothing to be gained by giving separate treatment to the question of local jurisdiction. As a general matter, municipal authority is entirely derivative of state authority, see 7A R.I. Gen. Laws § 45-2-1 (1991); and in the exercise of governmental powers (as opposed to proprietary powers), municipalities act only as the agents of the state, see Buckhout v. City of Newport, 68 R.I. 280, 27 A.2d 317, 320 (R.I. 1942).
It follows that if the state chooses to cede a portion of its sovereignty to the town, the town may use that authority to the extent of the power delegated See, e.g., Vukic v. Brunelle, 609 A.2d 938, 941 (R.I. 1992). But delegated powers, of necessity, cannot exceed those possessed by the delegator. The town has cited no independent basis upon which it might exercise municipal jurisdiction, and none is apparent to us. Thus, Charlestown's concerns are necessarily subsumed in our Discussion of the state's jurisdiction.
VII. THE REACH OF THE GAMING ACT
Before addressing the Tribe's ultimate argument - that the Gaming Act cancels whatever jurisdiction the Settlement Act granted - we must first consider both furcula of the state's assertion that the settlement lands are specifically exempted from the Gaming Act's domain.
A. The Consensual Transfer Provision.
The Gaming Act's so-called "consensual transfer" provision, familiarly known as "section 23(d)," is the site of the next battle. It states in relevant part:
The United States shall have exclusive jurisdiction over criminal prosecutions of violations of State gambling laws that are made applicable under this section to Indian country, unless an Indian tribe pursuant to a Tribal-State compact . . . or under any other provision of Federal law, has consented to the transfer to the State of criminal jurisdiction with respect to gambling on the lands of the Indian tribe.
18 U.S.C. § 1166(d). This proviso, Rhode Island asseverates, presages an exemption applicable to the settlement lands. On this theory, section 23(d) allows a state lawfully to assert civil and criminal jurisdiction over gaming under either a tribal-state compact or "any other provision of Federal law" that embodies a consensual transfer of jurisdiction. And it portrays section 1708 of the Settlement Act as constituting such an agreed transfer.
This interpretation signifies a promiscuous elevation of hope over reason, for it completely overlooks two limitations that are apparent on the face of the statute. First, section 23(d) is a penal provision that in terms deals only with criminal prosecutions; it has no implications for civil jurisdiction (whether regulatory or adjudicatory). Second, section 23(d) pertains only to "gambling," which is defined for purposes of that section as excluding any kind of "gaming." See 18 U.S.C. § 1166(c). Thus, properly understood, section 23(d) allows states to exercise jurisdiction pursuant to a consensual transfer only to enforce criminal laws that proscribe gambling activities falling outside the sanctuary of the Gaming Act. This is of no assistance to Rhode Island, which seeks to assert unfettered jurisdiction (including civil regulatory jurisdiction) over activities constituting class II and class III gaming.12
B. Decrypting the Legislative History.
Next, the state attempts a flanking maneuver. Without meaningful citation to the Gaming Act's text, the state hawks the proposition that Congress, in passing the Act, intended to leave intact the grant of jurisdiction tendered a decade earlier in the Settlement Act. And to fill the forensic void left by the utter absence of any statutory language to this effect, the state pushes forward carefully selected snippets of legislative history. There are two significant problems with this approach.
In the first place, courts must look primarily to statutory language, not to legislative history, in determining the meaning and scope of a statute. See, e.g., United States v. Turkette, 452 U.S. 576, 580, 69 L. Ed. 2d 246, 101 S. Ct. 2524 (1981); Consumer Prod. Safety Comm'n v. GTE Sylvania, Inc., 447 U.S. 102, 108, 64 L. Ed. 2d 766, 100 S. Ct. 2051 (1980); United States v. Charles George Trucking Co., 823 F.2d 685, 688 (1st Cir. 1987); see also Felix Frankfurter, The Reading of Statutes, reprinted in Of Law and Men 60 (Philip Elman ed. 1956) (noting importance of statutory language and explaining that legislative intent "is not drawn, like nitrogen, out of the air"). When a statute's text is encompassing, clear on its face, and productive of a plausible result, it is unnecessary to search for a different, contradictory meaning in the legislative record. See Charles George Trucking, 823 F.2d at 688; United States v. Meyer, 808 F.2d 912, 915 (1st Cir. 1987); Massachusetts Fin. Servs., Inc. v. Securities Investor Protection Corp., 545 F.2d 754, 757 (1st Cir. 1976), cert. denied, 431 U.S. 904, 52 L. Ed. 2d 388, 97 S. Ct. 1696 (1977). This is precisely such a case.
In the second place, legislative history that is in itself inconclusive will rarely, if ever, overcome the words of a statute. In a case such as this one, an inquiring court, at most, should resort to legislative history only to determine "whether there is a 'clearly expressed legislative intention' contrary to [the statutory] language, which would require [the court] to question the strong presumption that Congress expresses its intent through the language it chooses." INS v. Cardoza-Fonseca, 480 U.S. 421, 432 n.12, 94 L. Ed. 2d 434, 107 S. Ct. 1207 (1987) (quoting GTE Sylvania, 447 U.S. at 108). After careful consideration of Rhode Island's extratextual arguments, we conclude that the material it musters fails to establish such a clearly expressed legislative intention.
The state begins this phase of its case by pointing to a preliminary version of the Gaming Act that contained a provision (former section 23) safeguarding the Settlement Act from implied repeal.13 Once that foundation is poured, the state then brings to the forefront a colloquy on the floor of the Senate involving Rhode Island's two senators, Messrs. Pell and Chafee, and Senator Inouye, sponsor and floor manager of the bill that became the Gaming Act, regarding the eventual deletion of former section 23 from the bill:
Mr. PELL. Mr. President, I would like to thank the managers of S.555, the Indian Gaming Regulatory Act, and particularly the chairman of the Select Committee on Indian Affairs [Mr. Inouye], for their hard work and patience in achieving a consensus on this important measure.
In the interests of clarity,14 I have asked that language specifically citing the protections of the Rhode Island Claims Settlement Act (Public Law 95-395) be stricken from S.555. I understand that these protections clearly will remain in effect.
Mr. INOUYE. I thank my colleague, the senior Senator from Rhode Island [Mr. Pell], and assure him that the protections of the Rhode Island Claims Settlement Act (P.L. 95-395), will remain in effect and that the Narragansett Indian Tribe clearly will remain subject to the civil, criminal, and regulatory laws of the State of Rhode Island.
Mr. CHAFEE. Mr. President, I too would like to thank the chairman [Mr. Inouye] and members of the Select Committee on Indian Affairs for their cooperation and assistance. The chairman's statement makes it clear that any high stakes gaming, including bingo, in Rhode Island will remain subject to the civil, criminal and regulatory laws of our State.
134 Cong. Rec. S12,650 (daily ed. Sept. 15, 1988).
Although we give full faith and credit to the earnestness of the senators involved in this exchange, we are unable to accept the colloquy at face value. In the game of statutory interpretation, statutory language is the ultimate trump card. Consequently, the overarching rule is that "statements by individual legislators should not be given controlling effect"; rather, such statements are to be respected only to the extent that they "are consistent with the statutory language." Brock v. Pierce County, 476 U.S. 253, 263, 90 L. Ed. 2d 248, 106 S. Ct. 1834 (1986).
This interpretive rule applies fully to the special case of statements by those members of Congress most intimately associated with a bill: its floor manager and its sponsors. The Court has so stated in unmistakable terms: "The contemporaneous remarks of a sponsor of legislation are certainly not controlling in analyzing legislative history." Weinberger v. Rossi, 456 U.S. 25, 35 n.15, 71 L. Ed. 2d 715, 102 S. Ct. 1510 (1982); see also Brock, 476 U.S. at 263; GTE Sylvania, 447 U.S. at 118; Chrysler Corp. v. Brown, 441 U.S. 281, 311, 60 L. Ed. 2d 208, 99 S. Ct. 1705 (1979) cf. Grove City Coll. v. Bell, 465 U.S. 555, 567, 79 L. Ed. 2d 516, 104 S. Ct. 1211 (1984) (explaining that remarks of a sponsor may be taken as authoritative to the extent that they are consistent with plain language).15 Various courts of appeals, this court included, repeatedly have echoed the same theme. See, e.g., North & South Rivers Watershed Ass'n v. Scituate, 949 F.2d 552, 555 n.6 (1st Cir. 1991); United States v. Tabacca, 924 F.2d 906, 911 (9th Cir. 1991); Devargas v. Mason & Hanger-Silas Mason Co., 911 F.2d 1377, 1387 (10th Cir. 1990), cert. denied, 498 U.S. 1074, 112 L. Ed. 2d 860, 111 S. Ct. 799 (1991); United States v. McGoff, 265 U.S. App. D.C. 312, 831 F.2d 1071, 1090-91 (D.C. Cir. 1987); Northern Colo. Water Conservancy Dist. v. FERC, 235 U.S. App. D.C. 79, 730 F.2d 1509, 1518 (D.C. Cir. 1984).
This overarching rule makes good sense, for floor statements afford solid evidence of congressional intent only when they jibe with the final version of the statutory text. "To permit . . . clear statutory language to be materially altered by such colloquies, which often take place before the bill has achieved its final form, would open the door to the inadvertent, or perhaps even planned, undermining of the language actually voted on by Congress and signed into law by the President." Regan v. Wald, 468 U.S. 222, 237, 82 L. Ed. 2d 171, 104 S. Ct. 3026 (1984).
Another reason that this overarching rule makes sense is that, as a practical matter, most members of the enacting Congress will be familiar only with the bill as it stands when the vote occurs and, perhaps, with the committee reports, in broad outline of purpose; they cannot be expected to be familiar with every stray floor statement, with every twist and turn of the bill's prior history, or with every other legislator's thoughts as to what the bill accomplishes (or stops short of accomplishing). Cf., e.g., Hirschey v. FERC, 250 U.S. App. D.C. 1, 777 F.2d 1, 7-9 & n.1 (D.C. Cir. 1985) (Scalia, J., Concurring) (observing that members of Congress cannot be held accountable for lacking knowledge of minute details in committee reports). It is particularly unrealistic to attribute knowledge of statements made on the Senate floor to House members, who gave their seal of approval to the Senate bill only after the Rhode Island exemption provision had been deleted, without mentioning Rhode Island's parochial concern. See 134 Cong. Rec. H8146, H8426 (daily ed. Sept. 27, 1988) (commemorating passage by the House of Representatives). For much the same reason, it is unrealistic to attribute such knowledge to the President.
In our republican form of government, legislators make laws by writing statutes - an exercise that requires putting words on paper in a way that conveys a reasonably definite meaning. Once Congress has spoken, it is bound by what it has plainly said, notwithstanding the nods and winks that may have been exchanged in floor debates and committee hearings. After all, it is not the proper role of legislators to use unwritten assurances or side arrangements to alter the clear meaning of agreed language. And the judiciary must stand as the ultimate guarantor of the integrity of an enacted statute's text.
In sum, once Congress has spoken, a court cannot override the unambiguous words of an enacted statute and substitute for them the court's views of what individual legislators likely intended. Any other rule imports a virulent strain of subjectivity into the interpretive task and, in the process, threatens to transfer too large a slice of legislative power from Congress to the courts. See Frankfurter, supra, at 60 (warning that courts should not be "led off the trail by tests that have overtones of subjective design").
Here, the colloquy upon which the state relies is an especially slender reed because it offers an explanation of Congress's action that defies a widely accepted principle of statutory construction. When Congress includes limiting language in an early version of proposed legislation, and then rewrites the bill prior to enactment so as to scrap the limitation, the standard presumption is that Congress intended the proviso to operate without limitation. See Cardoza-Fonseca, 480 U.S. at 432; Russello v. United States, 464 U.S. 16, 23-24, 78 L. Ed. 2d 17, 104 S. Ct. 296 (1983); United States v. Sisseton-Wahpeton Sioux Tribe, 897 F.2d 358, 362 n.8 (8th Cir. 1990).16 Deletion, without more, suggests that Congress simply had a change of heart.
In this case, there is no "more." The state tries to buttress its argument by touting a comment in the Senate report to the effect "that nothing in the [Gaming Act] will supersede any specific restriction or specific grant of Federal authority or jurisdiction to a State which may be encompassed in another Federal statute, including the Rhode Island Claims Settlement Act [and the Maine] Indian Claims Settlement Act." 1988 U.S.C.C.A.N. at 3082 (citations omitted). But this patch of legislative history is also threadbare. The Senate report speaks only to the bill as reported out of committee; it was composed before the deletion of the special Rhode Island exemption provision, former section 23, from the final version of the bill during floor debate. Thus, the quoted statement sheds no light on Congress's intent regarding the law it actually enacted.17
VIII. DOES THE GAMING ACT APPLY?
Our odyssey is not yet finished, as the state and the amici construct a plausible textual argument as an amulet to ward off the Gaming Act. This argument stems from the language limiting the applicability of the Gaming Act's key provisions to "any Indian tribe having jurisdiction over Indian lands," or, stated differently, to "Indian lands within such tribe's jurisdiction." See 25 U.S.C. §§ 2710(d)(3)(A), 2710(b)(1). These are dual limitations, for one element of the definition of "Indian lands" requires that an Indian tribe "exercise governmental power" over them. 25 U.S.C. § 2703(4). Rhode Island claims that the Narragansetts do not "have jurisdiction" over, and do not exercise "governmental power" with respect to, the settlement lands; and, thus, that the Gaming Act does not pertain. Evaluating the state's thesis requires an exploration into another aspect of congressional intent.
A. Having Jurisdiction.
In the state's view, the phrase "having jurisdiction," as used in the Gaming Act, must, insofar as the settlement lands are concerned, be gauged in light of the Settlement Act. We agree. But the mere fact that the Settlement Act cedes power to the state does not necessarily mean, as Rhode Island suggests, that the Tribe lacks similar power and, thus, lacks "jurisdiction" over the settlement lands. Although the grant of jurisdictional power to the state in the Settlement Act is valid and rather broad, see supra Parts V(B), VI, we do not believe that it is exclusive. To the contrary, we rule that the Tribe retains concurrent jurisdiction over the settlement lands and that such concurrent jurisdiction is sufficient to satisfy the corresponding precondition to applicability of the Gaming Act.
In undertaking the task of determining whether the Settlement Act's jurisdictional grant is exclusive in nature, it must be remembered that Indian sovereignty is "a backdrop against which the applicable . . . federal statutes must be read." McClanahan v. State Tax Comm'n, 411 U.S. 164, 172, 36 L. Ed. 2d 129, 93 S. Ct. 1257 (1973). This backdrop is a necessary adjunct to the search for legislative intent in the context of Indian-related legislation. See Cotton Petroleum Corp. v. New Mexico, 490 U.S. 163, 176, 104 L. Ed. 2d 209, 109 S. Ct. 1698 (1989). Consequently, we paint the backdrop before placing the statute at center stage.
1. The Backdrop. Indian tribes are "distinct, independent political communities, retaining their original natural rights" in matters of local governance. Santa Clara Pueblo, 436 U.S. at 55, quoting Worcester v. Georgia, 31 U.S. (6 Pett.) 515, 559, 8 L. Ed. 483 (1832). While tribal rights are retained at congressional sufferance and are subject to defeasance should Congress so elect, tribes retain their sovereign powers in full measure unless and until Congress acts to circumscribe them. See United States v. Wheeler, 435 U.S. 313, 323, 55 L. Ed. 2d 303, 98 S. Ct. 1079 (1978). As the Supreme Court has explained, "Indian tribes still possess those aspects of sovereignty not withdrawn by treaty or statute, or by implication as a necessary result of their dependent status." Id.; accord Bottomly v. Passamaquoddy Tribe, 599 F.2d 1061, 1065-66 (1st Cir. 1979).
We believe that jurisdiction is an integral aspect of retained sovereignty. After all, the Court has held that retained sovereignty includes the power of Indians to make and enforce their own substantive law in internal matters, including matters such as membership rules, inheritance rules, and the regulation of domestic relations. See Santa Clara Pueblo, 436 U.S. at 56 (citing cases). Jurisdiction is cut from much the same fabric.
Of course, the shape of retained sovereignty has never been precisely defined. Thus, it cannot be said with assurance whether or not criminal, civil adjudicatory, and civil regulatory jurisdiction, in whole cloth, are aspects of retained sovereignty. But we have no need today to map such far-flung frontiers. For present purposes, so long as the portion of jurisdiction encompassed within the natural rights of the Narragansetts is substantial enough to satisfy the Gaming Act's "having jurisdiction" prong, our inquiry is satisfied.
The state has not contended that any treaty impinges upon the Tribe's jurisdiction. By like token, the record will not support a finding of jurisdiction abandoned or lost through implicit divestiture, see Wheeler, 435 U.S. at 326. We are left, then, with the relatively confined question of whether the Tribe's retained jurisdiction has been forfeited by statute. It is against this backdrop that we focus the lens of our inquiry on the Settlement Act.
2. The Settlement Act. By its terms, the Settlement Act purposes to do no more than grant jurisdiction to the state; it does not expressly strip the Tribe of jurisdiction, transfer jurisdiction from the Tribe to the state, or employ suggestive adjectives like "exclusive" or "complete" in describing the jurisdictional grant. The omission of the word "exclusive" looms particularly large in light of the use of that word elsewhere. For instance, the word is used to modify the general jurisdictional grant in 18 U.S.C. § 1162 (1988), one of the few analogous statutes granting "civil and criminal jurisdiction" over Indian lands to an individual state. Even more tellingly, the word is used in the Settlement Act itself, which characterizes as "exclusive" the grant to federal courts of jurisdiction to entertain certain constitutional challenges. 25 U.S.C. § 1711. This phenomenon commands our utmost attention, for where "Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion." Rodriguez v. United States, 480 U.S. 522, 525, 94 L. Ed. 2d 533, 107 S. Ct. 1391 (1987) (citations omitted).
We are not alone in our reluctance to infer exclusivity absent some suggestion to that effect in the statutory text. At least one other court has found the omission of words such as "exclusive" or "complete" in a similar context to be meaningful. See Cook, 922 F.2d at 1026, 1032-33 (concluding from the omission of any such language that a grant to New York of jurisdiction over Indian lands, embodied in 25 U.S.C. § 232, is non-exclusive).
Comparative analysis is also instructive. We think it is sensible to compare the jurisdictional grant embedded in the Settlement Act with the jurisdictional grants encased in two other Indian claims settlement acts that were to some extent modeled after the Settlement Act. Both of the latter pieces of legislation - one involving Massachusetts, one involving Maine - contain grants of jurisdiction parallel to section 1708, expressed in similar language. See 25 U.S.C. § 1771g (1988); 25 U.S.C. § 1725 (1988) Yet both acts also contain corresponding limits on Indian jurisdiction, conspicuously absent from the Settlement Act. See 25 U.S.C. § 1771e(a) ; 25 U.S.C. § 1725(f). By placing stated limits on the retained jurisdiction of the affected tribes, these newer acts imply that an unadorned grant of jurisdiction to a state - such as is embodied in the Settlement Act - does not in and of itself imply exclusivity.
We find these factors to be of decretory significance. Given the strong congressional bias, especially noticeable in the past generation, against policies that would promote Indian assimilation, see Bryan, 426 U.S. at 387-88 & n.14, and also given Congress's fortunate penchant for great clarity when expressing its intent in this area, see id. at 389 ("Congress knows well how to express its intent directly when that intent [is] to subject . . . Indians to the full sweep of state laws."); Mattz v. Arnett, 412 U.S. 481, 504 n.22, 37 L. Ed. 2d 92, 93 S. Ct. 2245 (1973) (observing that Congress generally employs "clear language of express termination when that result is desired") (collecting examples), we are of the view that acts diminishing the sovereign rights of Indian tribes should be strictly construed. So here. Since the Settlement Act does not unequivocally articulate an intent to deprive the Tribe of jurisdiction, we hold that its grant of jurisdiction to the state is non-exclusive. The Narragansetts, therefore, have made the necessary threshold showing. They retain that portion of jurisdiction they possess by virtue of their sovereign existence as a people - a portion sufficient to satisfy the Gaming Act's "having jurisdiction" prong.
B. Exercising Governmental Power.
In addition to having jurisdiction, a tribe must exercise governmental power in order to trigger the Gaming Act. Meeting this requirement does not depend upon the Tribe's theoretical authority, but upon the presence of concrete manifestations of that authority. Consequently, an inquiring court must assay the jurisdictional history of the settlement lands.18 Cf., e.g., De Coteau v. District County Court, 420 U.S. 425, 442, 43 L. Ed. 2d 300, 95 S. Ct. 1082 (1975).
The inquiry into governmental power need not detain us. In the post-recognition period, the Tribe has taken many strides in the direction of self-government. It has established a housing authority, recognized as eligible to participate in the Indian programs of the federal Department of Housing and Urban Development, see 24 C.F.R., Part 905 (1993). It has obtained status as the functional equivalent of a state for purposes of the Clean Water Act, after having been deemed by the Environmental Protection Agency as having "a governing body carrying out substantial governmental duties and powers," 33 U.S.C. § 1377(e) (1988), and as being capable of administering an effective program of water regulation, see 40 C.F.R. § 130.6(d) (1993). It has taken considerable advantage of the Indian Self-Determination and Education Assistance Act (ISDA), a statute specifically designed to help build "strong and stable tribal governments." 25 U.S.C. § 450(a)(b) (1988). The Tribe administers health care programs under an ISDA pact with the Indian Health Service, and, under ISDA contracts with the Bureau, administers programs encompassing job training, education, community services, social services, real estate protection, conservation, public safety, and the like. These activities adequately evince that the Tribe exercises more than enough governmental power to satisfy the second prong of the statutory test.
IX. THE INTERFACE
Because we have concluded that the settlement lands, under the Tribe's auspices, meet both prerequisites of the Gaming Act, those lands are subject to the Act's benefits and burdens. The task remaining is to determine how the Gaming Act and the Settlement Act operate in tandem.
A. Principles Governing the Interface.
In warming to this reconciliatory task, we abjure the preemption analysis undertaken below, see Narragansett Tribe, 816 F. Supp. at 804. The doctrine of preemption is derived from the Supremacy Clause, U.S. Const., Art VI, cl. 2, and therefore applies only to conflicts between federal provisions, on one hand, and state or local provisions, on the other hand. See Cipollone v. Liggett Group, 120 L. Ed. 2d 407, 112 S. Ct. 2608, 2617 (1992). The proper mode of analysis for cases that involve a perceived conflict between two federal statutes is that of implied repeal. See Cook, 922 F.2d at 1033 (rejecting preemption analysis as inappropriate in resolving a conflict between the Gaming Act and an earlier federal statute); see also 1A Norman J. Singer, Sutherland on Stat. Const. § 23.09 (5th ed. 1993). Hence, we follow that analytic path.
We start by reiterating the bedrock principle that implied repeals of federal statutes are disfavored. In the absence of a contrary legislative command, when two acts of Congress touch upon the same subject matter the courts should give effect to both, if that is feasible. See Pipefitters Local 562 v. United States, 407 U.S. 385, 432 n.43 (1972); United States v. Tynen, 78 U.S. (11 Wall.) 88, 92, 20 L. Ed. 153 (1871). In other words, so long as the two statutes, fairly construed, are capable of coexistence, courts should regard each as effective. See Traynor v. Turnage, 485 U.S. 535, 547-48, 99 L. Ed. 2d 618, 108 S. Ct. 1372 (1988). However, "if the two [acts] are repugnant in any of their provisions, the latter act, without any repealing clause, operates to the extent of the repugnancy as a repeal of the first." Tynen, 78 U.S. (11 Wall.) at 92. Even absent outright repugnancy, a repeal may be implied in cases where the later statute covers the entire subject "and embraces new provisions, plainly showing that it was intended as a substitute for the first act." Id.; see also Posadas v. National City Bank, 296 U.S. 497, 503-04, 80 L. Ed. 351, 56 S. Ct. 349 (1936); Natural Resources Defense Council v. EPA, 824 F.2d 1258, 1278 (1st Cir. 1987).19
The doctrine of implied repeal operates without special embellishment in the Indian law context. See, e.g., Blackfeet Indian Tribe v. Montana Power Co., 838 F.2d 1055, 1058 (9th Cir.), cert. denied, 488 U.S. 828, 102 L. Ed. 2d 56, 109 S. Ct. 79 (1988). The rationale for encouraging preemption in the Indian context - that the federal government is a more trustworthy guardian of Indian interests than the states - has no relevance to a conflict between two federal statutes.
B. Applying the Principles.
It is evident that the Settlement Act and the Gaming Act are partially but not wholly repugnant. The Settlement Act assigned the state a number of rights. Among those rights - and by no means one of the rights at the epicenter of the negotiations leading up to the Act - was the non-exclusive right to exercise jurisdiction, in all customary respects save two, see supra note 6, over the settlement lands. The Gaming Act leaves undisturbed the key elements of the compromise embodied in the Settlement Act. It also leaves largely intact the grant of jurisdiction - but it demands an adjustment of that portion of jurisdiction touching on gaming.
Even in respect to jurisdiction over gaming, the two laws do not collide head-on. Thus, in connection with class III gaming, the Gaming Act does not in itself negate the state's jurisdiction, but, instead, channels the state's jurisdiction through the tribal-state compact process. It is only with regard to class I and class II gaming that the Gaming Act ex proprio vigore bestows exclusive jurisdiction on qualifying tribes.20 And it is only to these small degrees that the Gaming Act properly may be said to have worked a partial repeal by implication of the preexisting statute.
In the area in which the two laws clash, the Gaming Act trumps the Settlement Act for two reasons. First, the general rule is that where two acts are in irreconcilable conflict, the later act prevails to the extent of the impasse.21 See Watt v. Alaska, 451 U.S. 259, 266, 68 L. Ed. 2d 80, 101 S. Ct. 1673 (1981); Tynen, 78 U.S. (11. Wall.) at 92; see also 2B Singer, Sutherland on Stat. Const., supra, § 51.02, at 121. Second, in keeping with the spirit of the standards governing implied repeals, courts should endeavor to read antagonistic statutes together in the manner that will minimize the aggregate disruption of congressional intent. Here, reading the two statutes to restrict state jurisdiction over gaming honors the Gaming Act and, at the same time, leaves the heart of the Settlement Act untouched. Taking the opposite tack - reading the two statutes in such a way as to defeat tribal jurisdiction over gaming on the settlement lands - would honor the Settlement Act, but would do great violence to the essential structure and purpose of the Gaming Act. Because the former course keeps disruption of congressional intent to a bare minimum, that reading is to be preferred.
Based on our understanding of the statutory interface, we hold that the provisions of the Indian Gaming Regulatory Act apply with full force to the lands in Rhode Island now held in trust by the United States for the Narragansett Indian Tribe.22
C. Some Unanswered Questions.
Despite this holding - a holding that resolves the case before us - it would be disingenuous to pretend that all the relevant questions have been answered. While the Tribe retains all aspects of its retained sovereignty, as that term is commonly comprehended in our jurisprudence, Congress, after having granted to the state non-exclusive jurisdiction over the settlement lands via the Settlement Act, impliedly withdrew from that grant, via the Gaming Act, the state's jurisdiction over gaming.23 Yet, the withdrawal of jurisdiction over gaming cannot be interpreted to signify a withdrawal of all residual jurisdiction.
This means that the state continues to possess a quantum of regulatory authority. Of course, any effort by the state to exercise this residual authority is hedged in by barriers on both sides: on one side, by the Tribe's retained rights of sovereignty; on the other side, by the Tribe's congressionally approved authority over a specific subject matter, namely, gaming. Testing the sturdiness of one or the other of these barriers in a given case will require "a particularized inquiry into the nature of the state, federal, and tribal interests at stake." White Mountain Apache Tribe v. Bracker, 448 U.S. 136, 145, 65 L. Ed. 2d 665, 100 S. Ct. 2578 (1980). We cannot undertake such an inquiry in the abstract, and, thus, the jurisdictional status of the settlement lands remains ill-defined in certain respects. But that is the nature of litigation; Article III of the Constitution forbids courts from issuing advisory opinions or answering hypothetical questions. See, e.g., International Longshoremen's & Warehousemen's Union v. Boyd, 347 U.S. 222, 224, 98 L. Ed. 650, 74 S. Ct. 447 (1954); United Public Workers v. Mitchell, 330 U.S. 75, 89, 91 L. Ed. 754, 67 S. Ct. 556 (1947). Having exhausted the limits of the case in controversy, we must depart the stage, leaving it set for the possibility of future litigation.
In parting, we offer a few words of guidance. The crucial questions which must yet be answered principally deal with the nature of the regulable activities which may - or may not - be subject to state control, e.g., zoning, traffic control, advertising, lodging. It is true that nondiscriminatory burdens imposed on the activities of non-Indians on Indian lands are generally upheld. See, e.g., Washington v. Confederated Tribes of Colville Indian Reservation, 447 U.S. 134, 151, 65 L. Ed. 2d 10, 100 S. Ct. 2069 (1980) (discussing tax burdens). But it is also true that a comprehensive federal regulatory scheme governing a particular area typically leaves no room for additional state burdens in that area. See White Mountain Apache Tribe, 448 U.S. at 148 (finding state timber regulation to be preempted). Which activities are deemed regulable, therefore, will probably depend, in the first instance, on which activities are deemed integral to gaming. Although the core functions of class III gaming on the settlement land are beyond Rhode Island's unilateral reach, the distinction between core functions and peripheral functions is tenebrous, as is the question of exactly what Rhode Island may and may not do with respect to those functions that eventually are determined to be peripheral.
If these criss-crossing lines prove agonizingly difficult to decipher, let alone to administer, they "are no more or less so than many of the classifications that pervade the law of Indian jurisdiction." Washington v. Yakima Indian Nation, 439 U.S. 463, 502, 58 L. Ed. 2d 740, 99 S. Ct. 740 (1979). And in all events, the jurisdictional issues remain subject to further judicial intervention, pursuant to the Gaming Act, in a more fact-specific context, if the parties' compact negotiations collapse.
We can go no further at this time. We add, however, that although our opinion today answers some questions and raises others, we do not mean to encourage the protagonists to litigate ad infinitum. The parties' baseline power need not be defined with exactitude by judicial decree where, as here, they are compelled to enter negotiations out of which will emerge a new balance of power. The next step in the allocation of jurisdiction over gaming is in the hands of the parties, through negotiations designed to produce a tribal-state compact as contemplated by the Gaming Act, see 25 U.S.C. § 2710(d). If cool heads and fair-minded thinking prevail, that step may be the last.
The district court's issuance of a mandatory injunction compelling Rhode Island to commence good faith negotiation of a tribal-state compact is affirmed. The declaratory judgment entered in the district court shall, however, be modified as may be necessary to reflect the holdings contained in this opinion. Costs to appellees.
- Dissent follows -
COFFIN, Senior Circuit Judge (dissenting). With understandable respect for the effort evident in the court's opinion, and with full recognition of the closeness of this case, I reluctantly am unable to accept its evaluation of legislative history and its Conclusion that the Gaming Act worked an implied repeal of the Settlement Act.
The court, in my view, errs in two respects. First, it invokes a generally applicable principle of statutory interpretation -- statutory language, if clear, forecloses recourse to legislative history -- in an area where this teaching is not so absolute, i.e., when two federal statutes, literally read, are in tension. Second, when the court deigns to consider the legislative history of the later, supposedly impliedly repealing statute, the Gaming Act, it undervalues it, characterizing it as "carefully selected snippets" that "fail to establish . . . a clearly expressed legislative intention." Ante pp. 25-26.
The most apposite recent authority of which I am aware is Watt v. Alaska, 451 U.S. 259, 68 L. Ed. 2d 80, 101 S. Ct. 1673 (1981), in which two federal statutes contained irreconcilably different formulae for the distribution of revenues from the lease or sale of minerals from wildlife refuges. The Court acknowledged that the consolidated cases before it "involved two statutes each of which by its literal terms applies to the facts before us." Id. at 266. There, as here, the argument was made that the plain language of the later statute controlled and made improper any resort to legislative history.
The Court agreed that the statutory language was the starting point, but stated that "ascertainment of the meaning apparent on the face of a single statute need not end the inquiry . . . because the plain meaning rule is 'rather an axiom of experience than a rule of law, and does not preclude consideration of persuasive evidence if it exists.' The circumstances of the enactment of particular legislation may persuade a court that Congress did not intend words of common meaning to have their literal effect." Id. at 265-66 (citations and footnote omitted).
The Court then stated:
Without depreciating this general rule [that the more recent of two irreconcilably conflicting statutes governs], we decline to read the statutes as being in irreconcilable conflict without seeking to ascertain the actual intent of Congress. Our examination of the legislative history is guided by another maxim: "'repeals by implication are not favored,'" Morton v. Mancari, 417 U.S. at 549, quoting Posadas v. National City Bank, 296 U.S. 497, 503, 80 L. Ed. 351, 56 S. Ct. 349 (1936). "The intention of the legislature to repeal must be 'clear and manifest.'" United States v. Borden Co., 308 U.S. 188, 198, 84 L. Ed. 181, 60 S. Ct. 182 (1939), quoting Red Rock v. Henry, 106 U.S. 596, 602, 27 L. Ed. 251, 1 S. Ct. 434 (1883). We must read the statutes to give effect to each if we can do so while preserving their sense and purpose. Mancari, supra, at 551; see Haggar Co. v. Helvering, 308 U.S. 389, 394, 84 L. Ed. 340, 60 S. Ct. 337 (1940).
Watt, 451 U.S. at 266-67.
The Court then, despite the absence of any explanation in the legislative history for adding the word "minerals" to the later legislation, after studying "the few legislative materials pertinent," was persuaded "that Congress intended to work no change in the pre-existing framework." Id. at 267.
The Dissenting three Justices argued as does the court in this case, but they did not prevail. So far as I have been able to ascertain, Watt has not been eroded since its issuance.
It seems clear to me, at least, that the legislative history in this record reveals not merely the lack of a "clear and manifest" Congressional intent to repeal, but an affirmative intent that the pre-existing legislation should remain intact. An examination of the history reveals an express explanation, a deliberate, pre-planned colloquy with the floor manager of the legislation (the chairman of the Select Committee on Indian Affairs) as the very first exchange with interested Senators following his introductory presentation.
In his presentation, Senator Inouye referred both to the objective of "determining what patterns of jurisdiction and regulation should govern the conduct of gaming activities on Indian lands" and affirmed the principle "that by virtue of their original tribal sovereignty, tribes reserved certain rights when entering into treaties with the United States, and that today, tribal governments retain all rights that were not expressly relinquished." 134 Cong. Rec. S12649 (daily ed. Sept. 15, 1988).
Immediately at the Conclusion of the chairman's presentation, the following colloquy took place:
Mr. President, I would like to thank the managers of S. 555, the Indian Gaming Regulatory Act, and particularly the chairman of the Select Committee on Indian Affairs [Mr. Inouye], for their hard work and patience in achieving a consensus on this important measure.
In the interests of clarity, I have asked that language specifically citing the protections of the Rhode Island Indian Claims Settlement Act (Public Law 95-395) be stricken from S. 555. I understand that these protections clearly will remain in effect.
I thank my colleague, the senior Senator from Rhode Island [Mr. Pell], and assure him that the protections of the Rhode Island Indian Claims Settlement Act (P.L. 95-395), will remain in effect and that the Narragansett Indian Tribe clearly will remain subject to the civil, criminal, and regulatory laws of the State of Rhode Island.
Mr. President, I too would like to thank the chairman [Mr. Inouye] and members of the Select Committee on Indian Affairs for their cooperation and assistance. The chairman's statement makes it clear that any high stakes gaming, including bingo, in Rhode Island will remain subject to the civil, criminal, and regulatory laws of our State.
134 Cong. Rec. S12650 (daily ed. Sept. 15, 1988).
Following the colloquy other senators made comments or asked questions. A colloquy similar to that quoted above, between Senator Reid and the chairman, established the extent to which an earlier piece of legislation dealing with gambling devices would be altered by the bill under Discussion. 134 Cong. Rec. S12650 (daily ed. Sept. 15, 1988). Still another exchange concerned the scope of actions allowed under a grandfather clause. 134 Cong. Rec. S12651.
Were such responsible and calculated floor exhanges with managers of legislation to be rendered of little or no account, the character of the legislating process would suffer a substantial constriction, and a valued opportunity for clarification, minor correction, and fine tuning would be lost. I do not think the judiciary should be a party to any such result.
Certainly in the instant case, this legislative history supports the Conclusion that the Rhode Island Senators thought the implied repeal language unnecessary because they did not believe that the jurisdictional provisions of the Gaming Act applied to the Settlement Act. That this accorded with the intent of the Senate seems equally clear, unless we are to proclaim this traditional kind of colloquy with leadership mere smoke and mirrors.
I think it also worth noting that the colloquy includes a statement by the bill's sponsor and floor manager, whose remarks usually are afforded substantial weight. See North Haven Bd. of Educ. v. Bell, 456 U.S. 512, 526-27, 72 L. Ed. 2d 299, 102 S. Ct. 1912 (1982); United States v. Mass. Maritime Academy, 762 F.2d 142, 149 (1st Cir. 1985).
I add that the colloquy, although clear and to the point, is not the only evidence of Congressional intent. The Senate report also mentions that "nothing in the [Gaming Act] will supersede any specific restriction or specific grant of Federal authority or jurisdiction to a State which may be encompassed in another Federal statute, including the Rhode Island Claims Settlement Act [and the Maine] Indian Claims Settlement Act." S.Rep. No. 446, 100th Cong. 2d Sess. 12 (1988), reprinted in 1988 U.S.C.C.A.N. 3071, 3082 (citations omitted). While the court concludes that the report, issued before the deletion proposed by Senator Pell, is of no present relevance, I do not think it can so easily be discounted. Rather, it seems to me entirely consistent with the colloquy with the Rhode Island senators.
If, therefore, we assign proper weight to the legislative history, I think it unavoidable that we would have to conclude that the Gaming Act had effected no implied repeal of the Settlement Act. If, of course, the Congress were to feel that an inJustice had been done to appellees, it could provide a remedy through supplemental legislation.24
I therefore, with great reluctance, Dissent.
1. The provenance of the two parcels remains of continuing legal relevance because the 900 acres donated by the state may be used only for conservation purposes. See 6A R.I. Gen. Laws § 37-18-14 (1990). Thus, the development plan for high-stakes gambling is of necessity limited to the so-called "private" portion of the settlement lands.
2. The State Act amendments themselves suggest that congressional approval of the land transfer is "required and appropriate," 6A R.I. Gen. Laws § 37-18-14, and the case law is in accord, see Oneida Indian Nation v. Oneida County, 414 U.S. 661, 667-68, 39 L. Ed. 2d 73, 94 S. Ct. 772 (1974) (explaining that, as a general rule, Indian tribes may not alienate their land without congressional consent). Yet, Congress never ratified the State Act amendments. Because the validity of the title transfer is not directly in issue in this litigation, and because appellants have not acknowledged, much less relied upon, the absence of ratification, we do not explore the consequences of this omission.
3. There are strong economic and political disincentives to an outright ban, for class III gaming encompasses, among other things, such popular sources of state revenue as lotteries, and such familiar fundraising devices as "Las Vegas nights" to benefit churches and other charities.
4. As matters now stand, the named plaintiffs (appellants before us) include the state, the town, and various state and municipal officials. For simplicity's sake, we refer to the plaintiffs, collectively, as "Rhode Island" or "the state." Similarly, we refer to the defendants, collectively, as "the Narragansetts" or "the Tribe," noting, however, that plaintiffs' suit also names two tribal hierarchs as defendants.
5. The Tribe originally cross-appealed from the stay and from the denial of its motion for relief therefrom. In a separate order, we today dismiss that cross-appeal, the Tribe having failed to present any developed argumentation in support thereof.
6. The exceptions relate to the Tribe's general exemption from state taxation, 25 U.S.C. § 1715(a), and its exemption from state regulations anent fishing and hunting, 25 U.S.C. § 1706(a)(3).
7. This legal principle also disposes of certain other arguments raised by the Tribe. Thus, the Tribe's sovereignty could not have been eviscerated by an act of the Rhode Island General Assembly, unratified by Congress, that purported to extinguish tribal status, see 1879-1880 Acts, Resolves and Reports of the General Assembly of the State of Rhode Island and Providence Plantations, Chap. 800, at 101-06, or by the issuance of mere administrative notices, see, e.g., 48 Fed. Reg. 6177-78 (Feb. 2, 1983).
8. The 1985 State Act amendments transferring title from the holding company to the Tribe contained a provision for state jurisdiction substantially identical to that contained in section 1708, see 6A R.I. Gen. Laws § 37-18-13(b); and the deeds conveying the settlement lands from the Tribe to the Bureau in 1988 explicitly confirmed the applicability of section 1708.
9. We do not believe the Tribe's cause is aided by the Bureau's tentative expression of support for the position that section 1708 excludes civil regulatory jurisdiction. See Southeast Regional Solicitor's Memorandum Opinion (April 30, 1992). The Bureau's views are not entitled to any special weight in the interpretation of statutory provisions that it is not charged to execute. See Crandon v. United States, 494 U.S. 152, 177, 108 L. Ed. 2d 132, 110 S. Ct. 997 (1990).
10. To the extent that the district court's opinion in Maynard v. Narrangansett Tribe, 798 F. Supp. 94, 98-99 (D.R.I. 1992), aff'd on other grounds, 984 F.2d 14, 15 (1st Cir. 1993), suggests a contrary view, we reject it.
11. Because our analysis is specific to the Settlement Act, we need not join the debate over the general applicability of the Bryan distinction. See United Keetoowah Band, 927 F.2d at 1176 n.13 (surveying debate); see also Yavapai-Prescott Indian Tribe v. Arizona, 796 F. Supp. 1292, 1294-96 (D. Ariz. 1992) (discussing applicability of Bryan distinction in respect to Gaming Act).
12. The state's hopelessly mistaken interpretation of the consensual transfer provision apparently derives from a dictum in Lac du Flambeau Band of Lake Superior Chippewa Indians v. Wisconsin, 743 F. Supp. 645, 653-54 (W.D. Wis. 1990) (suggesting that the final version of section 23 was intended to exempt Rhode Island). The Lac du Flambeau court was misled by the Senate report's gloss on the soon-to-be-deleted Rhode Island exemption provision, see infra pp. 26 n.13, 31-32. The case, therefore, lacks precedential value.
13. In the original bill, former section 23 read as follows: Nothing in this Act may be construed as permitting gaming activities, except to the extent permitted under the laws of the State of Rhode Island, on lands acquired by the Narragansett Indian Tribe under the Rhode Island Indian Claims Settlement Act or on any lands held by, or on behalf of, such Tribe. 134 Cong. Rec. S12,649 (daily ed. Sept. 15, 1988). After the Senate eliminated this provision it renumbered the remaining provisions. As a result, former section 23 and section 23 as enacted, 18 U.S.C. § 1166, discussed supra Part VII(A), bear no relation to one another.
14. We are constrained to note that whatever interests Congress may have been serving when it deleted the former section 23, "the interests of clarity" were not among them.
15. While statements by legislative sponsors are sometimes described as "an authoritative guide to the statute's construction," North Haven Bd. of Educ. v. Bell, 456 U.S. 512, 526-27, 72 L. Ed. 2d 299, 102 S. Ct. 1912 (1982) (citing cases), that description is appropriate only when a statute's text leaves room for differing interpretations. See, e.g., DeBartolo Corp. v. Florida Gulf Coast Trades Council, 485 U.S. 568, 585, 99 L. Ed. 2d 645, 108 S. Ct. 1392 (1988) ("It is the sponsors that we look to when the meaning of the statutory words is in doubt.") (emphasis supplied).
16. To be sure, it might be postulated that Congress deleted former section 23 because it feared that a specific reference to Rhode Island would give rise to the inference that other individual states with special grants of jurisdiction were not similarly exempted from the Gaming Act. Congress could have avoided such ambiguity in a myriad of ways short of striking former section 23, say, by substituting a generic exemption for a category of states including Rhode Island, or by citing Rhode Island in a non-exhaustive list of states that would be exempted. To accept the suggested rationale would be to indulge in sheer speculation.
17. We sympathize with the predicament in which Rhode Island's senators found themselves - being asked to take the word of a powerful committee chairman - but sympathy alone cannot carry the day. Our Dissenting brother puts the very best face on the state's argument, yet the Dissent, though gracefully written, contains nothing to shake our view of either the controlling legal principles or the legislative history. While we plead guilty to the charge of literalism, placing strong emphasis on the statutory text is a court's proper function. See, e.g., GTE Sylvania, 447 U.S. at 108 (declaring that statutory language "must ordinarily be regarded as conclusive"); Caminetti v. United States, 242 U.S. 470, 490, 61 L. Ed. 442, 37 S. Ct. 192 (1917) (explaining that "when words are free from doubt they must be taken as the final expression of the legislative intent"). Moreover, we have made a point of considering the legislative history of the Gaming Act on its own terms, in recognition of the continuing influence of less text-based theories of statutory interpretation, such as that underpinning Watt v. Alaska, 451 U.S. 259, 68 L. Ed. 2d 80, 101 S. Ct. 1673 (1981). We believe that our result is compelled by any acceptable mode of interpretation. Finally, although we share Judge Coffin's reticence to discredit responsible floor exchanges, we fail to see how a floor exchange utterly at odds with the words of an enacted statute can be given primacy in the interpretive process. If legislative bodies desire to accomplish particular results, they must use their tools with greater care.
18. An historical perspective is also relevant to the "having jurisdiction" inquiry. A "longstanding assumption of jurisdiction . . . not only demonstrates the parties' understanding of the meaning of the Act, but has created justifiable expectations which should not be upset. . . ." Rosebud Sioux, 430 U.S. at 604-05.
19. We addressed this point in United States v. Brien, 617 F.2d 299 (1st Cir.), cert. denied, 446 U.S. 919 (1980). We caution, however, that while Brien suggests that statutes may never be impliedly repealed in part, see id. at 309, more recent cases clarify the point, see, e.g., Bristol Energy Corp. v. New Hampshire PUC, 13 F.3d 471, (1st Cir. 1994) [No. 93-1824, slip op. at 16-17]. The rule is that, generally, there can be no partial implied repeal absent repugnancy. This is simply another way of stating that congressional intent to substitute a later act for an earlier one will ordinarily be implied only if the later act usurps the whole ground occupied by the first. See Posadas, 296 U.S. at 504. If repugnancy is found, however, then a partial repeal is in most cases preferred - indeed, mandated - for only that part of the earlier statute which is plainly anathematic should be nullified.
20. We take no view on whether, apart from the Gaming Act, a state might have regulated the activities that comprise class I gaming without violating the Free Exercise Clause.
21. The state argues that the Settlement Act should prevail because it is the more specific statute. There are two cracks in this palladium. As noted by the court below, it is arguable which statute is the more specific. See Narragansett Tribe, 816 F. Supp. at 804. More fundamentally, the canon upon which the state relies is rooted in the presumption that, when legislatures enact general laws, they do not have in mind every preexisting statute that touches on some specific aspect of the general subject. Where, as here, the enacting Congress is demonstrably aware of the earlier law at the time of the later law's enactment, there is no basis for indulging the presumption.
22. We decline to address certain constitutional claims advanced by the amici, for these claims were not urged by the plaintiffs in the court below. According to well established authority, amici can do no more than "assist the court in achieving a just resolution of issues raised by the parties." Lane v. First Nat'l Bank of Boston, 871 F.2d 166, 175 (1st Cir. 1989). In the court of appeals, amici cannot usurp the litigants' prerogative and introduce new issues or issues not properly preserved for appeal.
23. It is important to note, however, that jurisdiction over class III gaming is subject to restoration, in whole or in part, as a negotiated by-product of a tribal-state compact.