with whom Justice Brennan joins, dissenting.
The Court today holds that the Klamath Tribe has no special right to hunt and fish on certain lands although it has done so undisturbed from time immemorial. Instead, the Tribe is determined to be subject to state regulation to the same extent as any other person in the State of Oregon. This Court has in the past recognized that Indian hunting and fishing rights — even if nonexclusive, and even if existing apart from reservation lands — are valuable property rights, not fully subject to state regulation and not to be deemed abrogated without explicit indication.1 Although all agree that hunting and fishing have historically been vital to the continued prosperity of the Klamath, the Court today assumes that the Klamath Tribe silently gave up its rights to hunt and fish on these lands in a 1901 Agreement, approved by Congress in 1906, that had no purpose other than to benefit the Tribe for a previous injustice. It reaches this conclusion even though there is no historical evidence that any party to the Agreement envisioned it as having the effect of altering tribal hunting and fishing practices, and even though hunting and fishing practices did not in fact change as a result of the Agreement. Although I agree that the boilerplate language of the Agreement can be read as the Court does, I also believe that such a reading is not necessary, ignores the Agreement’s historical context, and is not faithful to the well-established principles that Indian treaties are to be interpreted as they were likely understood by the tribe and that doubts concerning the meaning of a treaty should be resolved in favor of the tribe.2 Accordingly, I dissent.
*776r-H
I will only briefly summarize the relevant history of the Klamath Reservation. As the Court explains, in 1864 the Klamath Tribe entered into a treaty with the United States whereby it agreed to settle on a reservation of 1.9 million acres in south central Oregon. Treaty of Oct. 14, 1864, 16 Stat. 707. This land was a small part of the 22 million acres of land to which the Klamath had held aboriginal title. As the Court points out: “The 1864 Treaty also provided that the [Klamath Tribe] would have. . . ‘the exclusive right of taking fish in the streams and lakes, included in said reservation, and of gathering edible roots, seeds, and berries within its limits.’” Ante, at 755. Although the borders of the reservation soon became the subject of some dispute, the purposes of the Treaty have always been clear. These purposes, and the importance of Indian hunting and fishing rights to their accomplishment, were well stated in a report to Congress by a Commission appointed to study the later boundary dispute:
“It was evidently a principal object of the treaty to draw the Indians in from the large extent of territory over which they were roaming, subject to constant collisions with the steadily encroaching whites, and to concentrate them on an area much more limited, but which would still be ample to provide them with the means of subsistence.
“To attain this, the marked tendency of the treaty and the emphatic testimony of the Indians seek to make all the boundaries mountain ridges, a purpose of which the nature of the country renders easy of accomplishment on all sides except the north.
“There is no provision in the treaty, however, for the support of the Indians by the Government, and as the *777high altitude and the severity of the climate are unfavorable to the cultivation of cereals, their subsistence depended upon natural products, consisting principally of game, fish, wild roots, and seeds. These mountain barriers, therefore, must include a territory frequented by game, streams stocked with fish, and ground producing the roots and seeds which formed so large a portion of the subsistence of the Indians.” S. Doc. No. 93, 54th Cong., 2d Sess., 6-7 (1897) (Klamath Boundary Commission Report).
The boundaries of the reservation that was eventually established pursuant to the Treaty, however, contained only about two-thirds of the land promised the Klamath Tribe, and among the areas left outside the reservation were tribal hunting, fishing, and gathering grounds of substantial importance. These areas had been specifically included in the Treaty’s definition of the planned reservation at the Tribe’s insistence; but, as the result of an erroneous 1871 survey, over 617,000 acres of land promised to the Tribe were excluded from the newly established reservation. As a result of the erroneous survey and in violation of the Treaty, non-Indians began to enter on the land for stock grazing and, to a lesser extent, for settlement. See, e. g., S. Exec. Doc. No. 129, 53d Cong., 2d Sess., 4-6, 8-9, 11, 17 (1894) (various documents noting grazing uses and relatively light settlement); see also n. 5, infra. The Klamath vehemently and repeatedly protested these entrances, but nevertheless continued to hunt and fish on the excluded land. See S. Doc. No. 93, supra, at 11, 15-16, 18. The protests continued for decades, and eventually led to Congress’ establishment of a Boundary Commission to determine the proper boundaries of the reservation and to determine the value of the erroneously excluded land. Act of June 10, 1896, ch. 398, 29 Stat. 321, 342.
The Boundary Commission went to the reservation and interviewed large numbers of Klamath. Tribal elders all *778insisted that they were sure that the disputed land was supposed to be in the reservation. They had explicitly demanded the land’s inclusion in the 1864 Treaty, they explained, because of the land’s traditional importance in the Tribe’s essential hunting, fishing, and gathering activities. The Commissioners inspected the land and found a tribal fishing site upon which a stone dam had been constructed and maintained by the Tribe to aid in gathering large numbers of fish. The Commission concluded that the Klamath’s complaints were largely justified and deserving of redress.3
The Commission determined, pursuant to the Tribe’s desires, that redress would take the form of officially ceding the excluded land back to the United States for compensation, leaving the border of the reservation where it had been erroneously set. As the Court notes, however, the Commission determined the value of the excluded land with no reference to its use for hunting, fishing, or gathering — basing valuation on its use for timber and stock grazing. Yet the Commission knew the land’s importance to the Tribe for hunting and fishing, since this was the basis of the Commission’s finding that it had been erroneously excluded from the reservation. Similarly, during the course of the two years of negotiations toward an agreement, there was no reference to any cessation of hunting, fishing, or gathering activity on the land in question, nor, it is true, to the continuing of such activity. The *779issue was simply never mentioned, and there is certainly no specific evidence that anyone, whether Klamath or Government official, envisioned that the Agreement would compel the Tribe to in any way alter the important hunting and fishing activities that it had been engaged in since the initial establishment of the reservation. During that time the Tribe had been forced to accept that others were entering and using the land, but the Tribe also had continued to fish and hunt as it always had done.
The Court is correct that the Tribe seemed fully satisfied with the possibility that the excluded land would be ceded to the United States for compensation, and there were no protests raised concerning loss of fishing, hunting, and gathering rights. Ante, at 759. But I cannot conclude from this silence that the Tribe understood and agreed to the extinguishing of hunting and fishing rights on the ceded land. Ante, at 770. Given the historical context of the 1901 Agreement, its proper interpretation is that, first, it compensated the Tribe for the fact that its position since the reservation’s establishment had been less than the Tribe had been promised, and, second, it preserved the Tribe’s position as it had actually existed since the erroneous survey. The Tribe’s actual position between the erroneous survey and the 1901 Agreement included no ability to exercise exclusive possession of the erroneously excluded lands, although they had been promised that right in the 1864 Treaty; but the Tribe’s position did include the ability to hunt and fish on those lands, and there is no reason to believe that a goal of the 1901 Agreement was to terminate such activities.
h-< I
A
As the Court notes, the case focuses on two provisions of the 1901 Agreement. Article I of the Agreement contained a broad cession by the Tribe of “all their claim, right, title, *780and interest in and to” the excluded land. 34 Stat. 367. In contrast, Article IV of the Agreement broadly declared that “nothing in [the] agreement shall be construed to deprive the said Klamath ... of any benefits to which they are entitled under existing treaties, not inconsistent with the provisions of this agreement.” Respondent and the courts below argued that the language of Article IV can reasonably be interpreted as a reservation by the Indians of a nonexclusive right to hunt and fish on those parts of the ceded land not in private hands.4
The Court rejects this construction of Article IV because of its unexplained insistence that the 1901 Agreement must be understood in terms of the structure of the 1864 Treaty, which envisioned no nonexclusive or off-reservation hunting rights. Indeed, as the Court emphasizes, a provision of the 1864 Treaty obligated the Tribe’s members to remain on the reservation established by its terms. 16 Stat. 708. Thus, in the Court’s view, because the reservation was diminished by the 1901 Agreement, and because the 1864 Treaty envisioned that the Tribe would hunt and fish only on its reservation, the 1901 Agreement must also have diminished the area where hunting and fishing rights existed. To allow nonexclusive *781hunting and fishing rights on the ceded lands would, in the Court’s view, create a “glaring inconsistency” with the 1864 Treaty, because to exercise such a right would have required the Tribe to leave the borders of its now-diminished reservation, in violation of the 1864 Treaty obligation to remain on reservation land. Ante, at 770.
B
This overly formal approach to treaty interpretation ignores the fundamental presumptions that Indian treaties are to be construed as the tribes would have understood them, Choctaw Nation v. Oklahoma, 397 U. S. 620, 631 (1970), and that ambiguities should be resolved in favor of the tribe. Washington v. Washington Commercial Passenger Fishing Vessel Assn., 443 U. S. 658, 675-676 (1979). I would have thought that an inquiry into the 1901 Agreement’s meaning would focus, not primarily on the formal structure of the 1864 Treaty — leaving both documents abstracted from their actual purposes and historical contexts — but instead on the problems that arose since 1864 that gave rise to the need for the 1901 Agreement. Certainly, the latter approach is better suited to the goal of determining the purposes of the parties, and especially, to the goal of determining the understandings of the Tribe.
When looking at the 1901 Agreement in terms of its own historical setting, the evidence clearly supports two conclusions —first, that the Tribe had no expectation that it was losing its ability to continue those fishing and hunting practices that it had been pursuing from time immemorial on the ceded lands, and second, that the United States had no particular interest in terminating such fishing and hunting activities.
(1)
The Tribe’s perspective is not difficult to divine. At the time of the 1901 Agreement, as well as at the time of the 1906 *782Act of Congress which ratified this Agreement, “[hjunting, fishing, gathering and trapping [were] crucial to the survival of the Klamath Indians.” App. 19 (stipulated facts). The Tribe had received, under the 1864 Treaty, the right to hunt and fish on the specific lands that were ceded in the 1901 Agreement, and had received that right because it had insisted on the particular importance to the Tribe of its ability to hunt and fish on those specific lands. Although these lands had not been included within the erroneous borders of the original reservation, the Tribe nevertheless entered them to hunt and fish.
The 1864 Treaty had also granted the Tribe the exclusive right to possess the lands in question, and particularly prohibited the use of these lands by non-Indians. 16 Stat. 708. But the Tribe had never been able to exercise this right to exclude others. The erroneous boundaries had opened the lands to others; thus, the Tribe’s ability to hunt and fish had become nonexclusive and its ability to exercise exclusive possession had disappeared. This was what it had lost, and accordingly, tribal members’ complaints had focused only on the presence of non-Indians on their lands. They never asserted an interference with their ability to hunt and fish. It is clear that the Tribe envisioned the 1901 Agreement only as providing compensation for the loss that the Tribe had suffered. And there is certainly nothing in the record to indicate that the Agreement in any way was working a further loss on the Tribe. In this context, Article IV makes clear that the Tribe was not to lose any benefits that it had actually possessed as it entered the 1901 Agreement.
(2)
The United States’ purposes were similarly clear, as the 1901 Agreement was entirely a result of Indian demands for the redress of an unfortunate mistake. The United States fully understood that the land in question was ill-suited for agriculture and settlement, and the record reflects no other *783collateral purposes of Congress. Indeed, there is no evidence of any pressures on Congress from non-Indians urging the cession at issue.5 There is simply no reason to believe that the United States — acting as trustee and seeking to compensate the Tribe for an unjust and accidental diminishment of their reservation — intended silently to effectuate a further diminution of tribal rights. We should not lightly assume that Congress, acting as a trustee of the Tribe’s interests, wished to deprive the Tribe of access to food supplies that it might need and had always utilized.
It is likely that the United States’ interests in 1901 had little to do with preserving the formal structure of the 1864 Treaty, an interest that the Court today assumes. Although the 1864 Treaty required the Tribe to stay on the land reserved to it by the Treaty, the alternative in 1864 was the Tribe’s continued presence on over 22 million acres of land to which it held aboriginal title. The land on which the Tribe was to stay, although poor land for settlement, was known to contain game, fish, and vegetation in such quantities as to allow the Tribe to be self-sufficient with no reason to wander. By 1901, there was no longer an issue as to whether the Tribe would ever again wander over the 22 million acres they had once held under aboriginal title — the Klamath had fully accepted that they would remain on a much smaller area. But the issue of retaining the Tribe’s self-sufficiency was still a concern.
In 1901, the Klamath were not viewed as hostile Indians, see n. 5, infra, and the surrounding land was minimally settled at best. For the United States to prohibit all tribal *784access to the ceded areas would have served no interest that the United States ever publicly declared, and it would have compromised the Klamath’s ability to remain self-sufficient. It is thus unreasonable to believe that the United States, while purporting to act for the benefit of the Indians, placed a high priority on assuring that the Klamath be strictly confined to the now-diminished area of their reservation, even if that would mean less access to food. The United States’ interests would have been fully served by reading the 1864 Treaty to require only that the Tribe not leave the area that was initially specified as the reservation. Article IV of the 1901 Agreement can thus easily be seen as an effort to preserve the Tribe’s right to travel, hunt, and fish on the full area of the original reservation, so long as those activities are consistent with the Tribe’s loss of exclusive possessory rights in the ceded lands. So long as the ceded lands were not opened to significant settlement, this resolution would fully serve what interest there still was in containing the Klamath and would not compromise the shared interest in continuing the Klamath’s self-sufficiency.
(3)
This interpretation of the parties’ perspectives fully conforms to what we know of the parties’ subsequent behavior.6 Congress never opened the ceded lands to settlement, and in fact, by the time it had ratified the 1901 Agreement, “[virtually all the land ceded by the Tribe was . . . closed to entry and placed in either national forests or parks.” App. 13-14 (stipulated facts). No argument has been made that continued hunting and fishing by the Indians is incompatible with the land’s uses. The Tribe’s behavior is also fully consistent *785with its current interpretation of the Agreement. The parties have stipulated that the Tribe has in fact “continued to hunt, fish and trap on the excluded lands from the time of their cession to the present,” id., at 14 (stipulated facts). Thus, no subsequent behavior of the United States or of the Tribe reflects an expectation that the Tribe would alter its hunting and fishing patterns as a result of the cession.
(4)
Last, the 1901 Agreement’s treatment of the issue of compensation also provides evidence that the parties did not envision that the Agreement denied the Klamath continued access to these traditional hunting and fishing grounds. The parties have stipulated that the Commission in no way-considered the land’s value for hunting or fishing when it calculated the proper compensation to the Tribe. Id., at 12. Yet, the Commission was well aware that the land was a hunting and fishing ground of some importance to the Tribe. Similarly, when the Indian Claims Commission reviewed and supplemented the compensation awarded the Klamath— more than six decades after the ratification of the Agreement — it never assigned any value to hunting or fishing rights. Id., at 14; see also Klamath and Modoc Tribes v. United States, 20 Ind. Cl. Comm’n 522 (1969). The silence of both these bodies is not surprising, if one accepts that the cession did not envision that Indian hunting and fishing would cease. We do not normally assume that the United States, without providing compensation, intended to deprive a tribe of valued hunting and fishing rights. Menominee Tribe of Indians v. United States, 391 U. S. 404, 413 (1968) (will not lightly assume that Congress meant to abrogate hunting and fishing rights without provision of compensation); cf. United States v. Sioux Nation of Indians, 448 U. S. 371, 422-424 (1980) (will not assume that compensation designed to ensure Tribe’s survival after it gave up traditional *786hunting activities was intended to cover both the taking of hunting rights and the taking of land). Nor should we lightly assume that the Tribe silently accepted the lack of specific compensation because its members understood that their valued hunting and fishing rights were merely incidental to land ownership.7
C
The analysis of the Agreement offered here is fully consistent with this Court’s prior cases regarding Indian hunting and fishing rights. We have accepted that nonexclusive hunting and fishing rights have often existed independently from rights of exclusive possession of land. Thus, there have been many treaties in which Indians have explicitly reserved nonexclusive hunting and fishing rights while ceding the corresponding lands. See nn. 1 and 4, supra. Similarly, Congress has explicitly reserved to a Tribe continued hunting and fishing rights even after a reservation has been fully terminated. See, e. g., 25 U. S. C. §564m(b) (fishing rights explicitly reserved upon termination of Klamath Reservation in 1954); see also Kimball v. Callahan, 590 F. 2d 768, 772 (CA9), cert. denied, 444 U. S. 826 (1979). But most importantly, this Court has held that hunting and fishing rights can by implication survive the full termination of a reservation, even where the enactment terminating the reservation is written in broad language and makes no reference to those rights’ survival. Menominee Tribe of Indians v. United States, supra.
In this case, as a result of the erroneous survey there was a de facto separation of the Klamath’s hunting and fishing *787rights from their rights of exclusive possession of the land. The former rights existed to the extent they could, consistent with the loss of the latter rights. In essence, the Tribe was left with off-reservation rights to hunt and fish on land from which it could not exclude others. The 1901 Agreement, which preserved to the Klamath all “benefits to which they are entitled under existing treaties, not inconsistent with the provisions of the [cession],” was not meant to take from them what was left of their right of access to their traditional hunting and fishing grounds.
Ill
In light of this Court’s repeated statements that the abrogation of Indian rights should not be lightly inferred, and that treaties be interpreted as they would have been understood by the Indians, I find the Court’s opinion today disturbing. Rather than follow the sort of historical inquiry that these canons should call for, the Court analyzes the case as one involving little more than the plain meaning of boilerplate language. It turns to history only to determine if its perceived “plain meaning” would be an impossible one. Ultimately, this produces a largely insensitive and conclusory historical inquiry that ignores how events almost certainly appeared to the Tribe.
The decision today represents another erroneous deprivation of the Klamath’s tribal rights. The Court has offered no reason to believe the 1901 Agreement was designed to accomplish anything other than the redress of the wrong that had already been done to the Tribe. The Court has certainly offered no reason to believe that it was designed to effectuate a further diminution of the Klamath’s rights.
I respectfully dissent.
See, e. g., United States v. Sioux Nation of Indians, 448 U. S. 371, 422-423 (1980); Menominee Tribe of Indians v. United States, 391 U. S. 404 (1968); Tutee v. Washington, 315 U. S. 681 (1942).
See Washington v. Washington Commercial Passenger Fishing Vessel Assn., 443 U. S. 658, 675-676 (1979); Choctaw Nation v. Oklahoma, 397 U. S. 620, 631 (1970); see also Menominee Tribe of Indians v. United *776States, supra, at 413 (the intention to abrogate treaty rights is not to be lightly imputed to Congress).
The Boundary Commission concluded its report as follows:
“In conclusion, we respectfully submit that during all this long period of thirty-two years these Indians have exhibited a patient and unwavering confidence in the justice of the Government demanding the highest commendation.
“Believing themselves to be grievously wronged by the white settlements on land they considered secured to them by solemn pledge of the Government and from which their subsistence was largely drawn, they yet endured all the aggravating conditions of these years, resisting all the allurements of the adjacent and kindred tribes during the [recent war] and remained loyal and true.” S. Doc. No. 93, 54th Cong., 2d Sess., 11 (1897).
As the Court notes, ante, at 764-765, n. 15, the Klamath claim a hunting and fishing right quite similar to the right of nonexclusive, off-reservation hunting and fishing expressly reserved by many of the Indians of the Pacific Northwest when they entered cession agreements. See Puyallup Tribe v. Department of Game of Washington, 391 U. S. 392 (1968); United, States v. Winans, 198 U. S. 371 (1905). I would also agree with the Court that such a right is not an “absolute freedom from state regulation.” See ante, at 765, n. 16. It should also be emphasized, however, that the right is nonetheless a valuable one, placing significant limits on permissible state regulation. See Antoine v. Washington, 420 U. S. 194, 207 (1975) (State must demonstrate that its regulation is a reasonable and necessary conservation measure and that its application to the Indians is necessary in the interest of conservation); see also Department of Game of Washington v. Puyallup Tribe, 414 U. S. 44 (1973); Tulee v. Washington, 315 U. S., at 684.
As the Court points out, see ante, at 759-760, n. 8, the United States’ first negotiator considered the excluded land “practically worthless” and believed that Congress should restore to the reservation the unentered excluded acreage rather than purchase it. The Tribe resisted this recommendation, preferring the compensated cession that was eventually accepted by Congress.
This Court has accepted that subsequent history of Indian lands can give “additional clue[s] as to what Congress expected would happen [with respect to] land on a particular reservation.” Solem v. Bartlett, 465 U. S. 463, 472 (1984).
The Court speculates that the right to hunt and fish was simply not viewed by the Indians as a right separate from the right to possess the land. But the Indians clearly did value the hunting and fishing, and both before and after the 1901 Agreement, the Indians continued to hunt and fish without interference even though, during both periods, they knew that they did not exercise exclusive possession of the land. I decline to assume that the Indians were simply consciously violating the law.