gress' use of the phrase "directly related" implies that student groups directly related to the subject matter of courses offered by the school do not fall within the "noncurriculum related" category and would therefore be considered "curriculum related."

The logic of the Act also supports this view, namely, that a curriculum-related student group is one that has more than just a tangential or attenuated relationship to courses offered by the school. Because the purpose of granting equal access is to prohibit discrimination between religious or political clubs on the one hand and other noncurriculum-related student groups on the other, the Act is premised on the notion that a religious or political club is itself likely to be a noncurriculum-related student group. It follows, then, that a student group that is "curriculum related" must at least have a more direct relationship to the curriculum than a religious or political club would have.

Although the phrase "noncurriculum related student group" nevertheless remains sufficiently ambiguous that we might normally resort to legislative history, see, e. g., James, supra, at 606, we find the legislative history on this issue less than helpful. Because the bill that led to the Act was extensively rewritten in a series of multilateral negotiations after it was passed by the House and reported out of committee by the Senate, the Committee Reports shed no light on the language actually adopted. During congressional debate on the subject, legislators referred to a number of different definitions, and thus both petitioners and respondents can cite to legislative history favoring their interpretation of the phrase. Compare 130 Cong. Rec. 19223 (1984) (statement of Sen. Hatfield) (curriculum-related clubs are those that are "really a kind of extension of the classroom"), with ibid. (statement of Sen. Hatfield) (in response to question whether school districts would have full authority to decide what was curriculum related, "[w]e in no way seek to limit that discretion"). See Laycock, Equal Access and Moments of Silence: The Equal

Status of Religious Speech by Private Speakers, 81 Nw. U. L. Rev. 1, 37-39 (1986).

We think it significant, however, that the Act, which was passed by wide, bipartisan majorities in both the House and the Senate, reflects at least some consensus on a broad legislative purpose. The Committee Reports indicate that the Act was intended to address perceived widespread discrimination against religious speech in public schools, see H. R. Rep. No. 98-710, p. 4 (1984); S. Rep. No. 98-357, pp. 10–11 (1984), and, as the language of the Act indicates, its sponsors contemplated that the Act would do more than merely validate the status quo. The Committee Reports also show that the Act was enacted in part in response to two federal appellate court decisions holding that student religious groups could not, consistent with the Establishment Clause, meet on school premises during noninstructional time. See H. R. Rep. No. 98-710, supra, at 3-6 (discussing Lubbock Civil Liberties Union v. Lubbock Independent School Dist., 669 F. 2d 1038, 1042-1048 (CA5 1982), cert. denied, 459 U. S. 1155-1156 (1983), and Brandon v. Guilderland Bd. of Ed., 635 F. 2d 971 (CA2 1980), cert. denied, 454 U. S. 1123 (1981)); S. Rep. No. 98-357, supra, at 6-9, 11-14 (same). A broad reading of the Act would be consistent with the views of those who sought to end discrimination by allowing students to meet and discuss religion before and after classes.

In light of this legislative purpose, we think that the term "noncurriculum related student group" is best interpreted broadly to mean any student group that does not directly relate to the body of courses offered by the school. In our view, a student group directly relates to a school's curriculum if the subject matter of the group is actually taught, or will soon be taught, in a regularly offered course; if the subject matter of the group concerns the body of courses as a whole; if participation in the group is required for a particular course; or if participation in the group results in academic

credit. We think this limited definition of groups that directly relate to the curriculum is a commonsense interpretation of the Act that is consistent with Congress' intent to provide a low threshold for triggering the Act's requirements. For example, a French club would directly relate to the curriculum if a school taught French in a regularly offered course or planned to teach the subject in the near future. A school's student government would generally relate directly to the curriculum to the extent that it addresses concerns, solicits opinions, and formulates proposals pertaining to the body of courses offered by the school. If participation in a school's band or orchestra were required for the band or orchestra classes, or resulted in academic credit, then those groups would also directly relate to the curriculum. The existence of such groups at a school would not trigger the Act's obligations.

On the other hand, unless a school could show that groups such as a chess club, a stamp collecting club, or a community service club fell within our description of groups that directly relate to the curriculum, such groups would be "noncurriculum related student groups" for purposes of the Act. The existence of such groups would create a "limited open forum" under the Act and would prohibit the school from denying equal access to any other student group on the basis of the content of that group's speech. Whether a specific student group is a “noncurriculum related student group" will therefore depend on a particular school's curriculum, but such determinations would be subject to factual findings well within the competence of trial courts to make.

Petitioners contend that our reading of the Act unduly hinders local control over schools and school activities, but we think that schools and school districts nevertheless retain a significant measure of authority over the type of officially recognized activities in which their students participate. See, e. g., Hazelwood School Dist. v. Kuhlmeier, 484 U. S. 260 (1988); Bethel School Dist. No. 403 v. Fraser, 478 U. S.

675 (1986). First, schools and school districts maintain their traditional latitude to determine appropriate subjects of instruction. To the extent that a school chooses to structure its course offerings and existing student groups to avoid the Act's obligations, that result is not prohibited by the Act. On matters of statutory interpretation, "[o]ur task is to apply the text, not to improve on it." Pavelic & LeFlore v. Marvel Entertainment Group, 493 U. S. 120, 126 (1989). Second, the Act expressly does not limit a school's authority to prohibit meetings that would "materially and substantially interfere with the orderly conduct of educational activities within the school." 84071(c)(4); cf. Tinker v. Des Moines Independent Community School Dist., 393 U. S. 503, 509 (1969). The Act also preserves "the authority of the school, its agents or employees, to maintain order and discipline on school premises, to protect the well-being of students and faculty, and to assure that attendance of students at meetings is voluntary." §4071(f). Finally, because the Act applies only to public secondary schools that receive federal financial assistance, §4071(a), a school district seeking to escape the statute's obligations could simply forgo federal funding. Although we do not doubt that in some cases this may be an unrealistic option, Congress clearly sought to prohibit schools from discriminating on the basis of the content of a student group's speech, and that obligation is the price a federally funded school must pay if it opens its facilities to noncurriculum-related student groups.

The dissent suggests that "an extracurricular student organization is 'noncurriculum related' if it has as its purpose (or as part of its purpose) the advocacy of partisan theological, political, or ethical views." Post, at 276; see also post, at 271, 290 (Act is triggered only if school permits "controversial” or "distasteful" groups to use its facilities); post, at 291 ("noncurriculum" subjects are those that "cannot properly be included in a public school curriculum'"). This interpretation of the Act, we are told, is mandated by Congress' intention to

“track our own Free Speech Clause jurisprudence," post, at 279, n. 10, by incorporating Widmar's notion of a "limited public forum" into the language of the Act. Post, at 271-272.

This suggestion is flawed for at least two reasons. First, the Act itself neither uses the phrase "limited public forum" nor so much as hints that that doctrine is somehow "incorporated” into the words of the statute. The operative language of the statute, 20 U. S. C. §4071(a), of course, refers to a "limited open forum," a term that is specifically defined in the next subsection, § 4071(b). Congress was presumably aware that “limited public forum," as used by the Court, is a term of art, see, e. g., Perry Ed. Assn. v. Perry Local Educators' Assn., 460 U. S. 37, 45-49 (1983), and had it intended to import that concept into the Act, one would suppose that it would have done so explicitly. Indeed, Con

gress' deliberate choice to use a different term-and to define that term-can only mean that it intended to establish a standard different from the one established by our free speech cases. See Laycock, 81 Nw. U. L. Rev., at 36 (“The statutory limited open forum' is an artificial construct, and comparisons with the constitutional ['limited public forum'] cases can be misleading"). To paraphrase the dissent, “[i]f Congress really intended to [incorporate] Widmar for reasons of administrative clarity, Congress kept its intent well hidden, both in the statute and in the debates preceding its passage." Post, at 281-282, n. 15.

Second, and more significant, the dissent's reliance on the legislative history to support its interpretation of the Act shows just how treacherous that task can be. The dissent appears to agree with our view that the legislative history of the Act, even if relevant, is highly unreliable, see, e. g., post, at 274-275, n. 5, and 281–282, n. 15, yet the interpretation it suggests rests solely on a few passing, general references by legislators to our decision in Widmar, see post, at 274, and n. 4. We think that reliance on legislative history is hazardous at best, but where "not even the sponsors of the bill