concurring.
I concur in the opinion of the Court. I write separately to suggest a clarification of our Establishment Clause doctrine. The suggested approach leads to the same result in this case as that taken by the Court, and the Court’s opinion, as I read it, is consistent with my analysis.
I
The Establishment Clause prohibits government from making adherence to a religion relevant in any way to a person’s standing in the political community. Government can run afoul of that prohibition in two principal ways. One is ex*688cessive entanglement with religious institutions, which may interfere with the independence of the institutions, give the institutions access to government or governmental powers not fully shared by nonadherents of the religion, and foster the creation of political constituencies defined along religious lines. E. g., Larkin v. Grendel’s Den, Inc., 459 U. S. 116 (1982). The second and more direct infringement is government endorsement or disapproval of religion. Endorsement sends a message to nonadherents that they are outsiders, not full members of the political community, and an accompanying message to adherents that they are insiders, favored members of the political community. Disapproval sends the opposite message. See generally Abington School District v. Schempp, 374 U. S. 203 (1963).
Our prior cases have used the three-part test articulated in Lemon v. Kurtzman, 403 U. S. 602, 612-613 (1971), as a guide to detecting these two forms of unconstitutional government action. * It has never been entirely clear, however, *689how the three parts of the test relate to the principles enshrined in the Establishment Clause. Focusing on institutional entanglement and on endorsement or disapproval of religion clarifies the Lemon test as an analytical device.
II
In this case, as even the District Court found, there is no institutional entanglement. Nevertheless, the respondents contend that the political divisiveness caused by Pawtucket’s display of its creche violates the excessive-entanglement prong of the Lemon test. The Court’s opinion follows the suggestion in Mueller v. Allen, 463 U. S. 388, 403-404, n. 11 (1983), and concludes that “no inquiry into potential political divisiveness is even called for” in this case. Ante, at 684. In my view, political divisiveness along religious lines should not be an independent test of constitutionality.
Although several of our cases have discussed political divisiveness under the entanglement prong of Lemon, see, e. g., Committee for Public Education & Religious Liberty v. Nyquist, 413 U. S. 756, 796 (1973); Lemon v. Kurtzman, supra, at 623, we have never relied on divisiveness as an independent ground for holding a government practice unconstitutional. Guessing the potential for political divisiveness inherent in a government practice is simply too speculative an enterprise, in part because the existence of the litigation, as this case illustrates, itself may affect the political response to the government practice. Political divisiveness is admittedly an evil addressed by the Establishment Clause. Its existence may be evidence that institutional entanglement is excessive or that a government practice is perceived as an endorsement of religion. But the constitutional inquiry should focus ultimately on the character of the government activity that might cause such divisiveness, not on the divisiveness itself. The entanglement prong of the Lemon test is properly limited to institutional entanglement.
*690I — j ) — I I — I
The central issue in this case is whether Pawtucket has endorsed Christianity by its display of the creche. To answer that question, we must examine both what Pawtucket intended to communicate in displaying the creche and what message the city’s display actually conveyed. The purpose and effect prongs of the Lemon test represent these two aspects of the meaning of the city’s action.
The meaning of a statement to its audience depends both on the intention of the speaker and on the “objective” meaning of the statement in the community. Some listeners need not rely solely on the words themselves in discerning the speaker’s intent: they can judge the intent by, for example, examining the context of the statement or asking questions of the speaker. Other listeners do not have or will not seek access to such evidence of intent. They will rely instead on the words themselves; for them the message actually conveyed may be something not actually intended. If the audience is large, as it always is when government “speaks” by word or deed, some portion of the audience will inevitably receive a message determined by the “objective” content of the statement, and some portion will inevitably receive the intended message. Examination of both the subjective and the objective components of the message communicated by a government action is therefore necessary to determine whether the action carries a forbidden meaning.
The purpose prong of the Lemon test asks whether government’s actual purpose is to endorse or disapprove of religion. The effect prong asks whether, irrespective of government’s actual purpose, the practice under review in fact conveys a message of endorsement or disapproval. An affirmative answer to either question should render the challenged practice invalid.
A
The purpose prong of the Lemon test requires that a government activity have a secular purpose. That requirement *691is not satisfied, however, by the mere existence of some secular purpose, however dominated by religious purposes. In Stone v. Graham, 449 U. S. 39 (1980), for example, the Court held that posting copies of the Ten Commandments in schools violated the purpose prong of the Lemon test, yet the State plainly had some secular objectives, such as instilling most of the values of the Ten Commandments and illustrating their connection to our legal system, but see 449 U. S., at 41. See also Abington School District v. Schempp, 374 U. S., at 223-224. The proper inquiry under the purpose prong of Lemon, I submit, is whether the government intends to convey a message of endorsement or disapproval of religion.
Applying that formulation to this case, I would find that Pawtucket did not intend to convey any message of endorsement of Christianity or disapproval of non-Christian religions. The evident purpose of including the créche in the larger display was not promotion of the religious content of the creche but celebration of the public holiday through its traditional symbols. Celebration of public holidays, which have cultural significance even if they also have religious aspects, is a legitimate secular purpose.
The District Court’s finding that the display of the créche had no secular purpose was based on erroneous reasoning. The District Court believed that it should ascertain the city’s purpose in displaying the créche separate and apart from the general purpose in setting up the display. It also found that, because the tradition-celebrating purpose was suspect in the court’s eyes, the city’s use of an unarguably religious symbol “raises an inference” of intent to endorse. When viewed in light of correct legal principles, the District Court’s finding of unlawful purpose was clearly erroneous.
B
Focusing on the evil of government endorsement or disapproval of religion makes clear that the effect prong of the Lemon test is properly interpreted not to require invalidation of a government practice merely because it in fact causes, *692even as a primary effect, advancement or inhibition of religion. The laws upheld in Walz v. Tax Comm’n, 397 U. S. 664 (1970) (tax exemption for religious, educational, and charitable organizations), in McGowan v. Maryland, 366 U. S. 420 (1961) (mandatory Sunday closing law), and in Zorach v. Clauson, 343 U. S. 306 (1952) (released time from school for off-campus religious instruction), had such effects, but they did not violate the Establishment Clause. What is crucial is that a government practice not have the effect of communicating a message of government endorsement or disapproval of religion. It is only practices having that effect, whether intentionally or unintentionally, that make religion relevant, in reality or public perception, to status in the political community.
Pawtucket’s display of its creche, I believe, does not communicate a message that the government intends to endorse the Christian beliefs represented by the créehe. Although the religious and indeed sectarian significance of the creche, as the District Court found, is not neutralized by the setting, the overall holiday setting changes what viewers may fairly understand to be the purpose of the display — as a typical museum setting, though not neutralizing the religious content of a religious painting, negates any message of endorsement of that content. The display celebrates a public holiday, and no one contends that declaration of that holiday is understood to be an endorsement of religion. The holiday itself has very strong secular components and traditions. Government celebration of the holiday, which is extremely common, generally is not understood to endorse the religious content of the holiday, just as government celebration of Thanksgiving is not so understood. The créche is a traditional symbol of the holiday that is very commonly displayed along with purely secular symbols, as it was in Pawtucket.
These features combine to make the government’s display of the creche in this particular physical setting no more an endorsement of religion than such governmental “acknowl*693edgments” of religion as legislative prayers of the type approved in Marsh v. Chambers, 463 U. S. 783 (1983), government declaration of Thanksgiving as a public holiday, printing of “In God We Trust” on coins, and opening court sessions with “God save the United States and this honorable court.” Those government acknowledgments of religion serve, in the only ways reasonably possible in our culture, the legitimate secular purposes of solemnizing public occasions, expressing confidence in the future, and encouraging the recognition of what is worthy of appreciation in society. For that reason, and because of their history and ubiquity, those practices are not understood as conveying government approval of particular religious beliefs. The display of the creche likewise serves a secular purpose — celebration of a public holiday with traditional symbols. It cannot fairly be understood to convey a message of government endorsement of religion. It is significant in this regard that the creche display apparently caused no political divisiveness prior to the filing of this lawsuit, although Pawtucket had incorporated the créche in its annual Christmas display for some years. For these reasons, I conclude that Pawtucket’s display of the créche does not have the effect of communicating endorsement of Christianity.
The District Court’s subsidiary findings on the effect test are consistent with this conclusion. The court found as facts that the créche has a religious content, that it would not be seen as an insignificant part of the display, that its religious content is not neutralized by the setting, that the display is celebratory and not instructional, and that the city did not seek to counteract any possible religious message. These findings do not imply that the créche communicates government approval of Christianity. The District Court also found, however, that the government was understood to place its imprimatur on the religious content of the créche. But whether a government activity communicates endorsement of religion is not a question of simple historical fact. *694Although evidentiary submissions may help answer it, the question is, like the question whether racial or sex-based classifications communicate an invidious message, in large part a legal question to be answered on the basis of judicial interpretation of social facts. The District Court’s conclusion concerning the effect of Pawtucket’s display of its créche was in error as a matter of law.
HH <
Every government practice must be judged m its unique circumstances to determine whether it constitutes an endorsement or disapproval of religion. In making that determination, courts must keep in mind both the fundamental place held by the Establishment Clause in our constitutional scheme and the myriad, subtle ways in which Establishment Clause values can be eroded. Government practices that purport to celebrate or acknowledge events with religious significance must be subjected to careful judicial scrutiny.
The city of Pawtucket is alleged to have violated the Establishment Clause by endorsing the Christian beliefs represented by the créche included in its Christmas display. Giving the challenged practice the careful scrutiny it deserves, I cannot say that the particular créche display at issue in this case was intended to endorse or had the effect of endorsing Christianity. I agree with the Court that the judgment below must be reversed.
The Court wrote in Lemon v. Kurtzman that a statute must pass three tests to withstand Establishment Clause challenge.
“First, the statute must have a secular legislative purpose; second, its principal or primary effect must be one that neither advances nor inhibits religion; finally, the statute must not foster ‘an excessive government entanglement with religion.’” 408 U. S., at 612-613 (citations omitted). Though phrased as a uniformly applicable test for constitutionality, this three-part test “provides ‘no more than [a] helpful signpos[t]’ in dealing with Establishment Clause challenges.” Mueller v. Allen, 463 U. S. 388, 394 (1983) (quoting Hunt v. McNair, 413 U. S. 734, 741 (1973)).
Moreover, the Court has held that a statute or practice that plainly embodies an intentional discrimination among religions must be closely fitted to a compelling state purpose in order to survive constitutional challenge. See Larson v. Valente, 456 U. S. 228 (1982). As the Court’s opinion observes, ante, at 687, n. 13, this case does not involve such discrimination. The Larson standard, I believe, may be assimilated to the Lemon test in the clarified version I propose. Plain intentional discrimination should give rise to a presumption, which may be overcome by a showing of compelling purpose and close fit, that the challenged government conduct constitutes an endorsement of the favored religion or a disapproval of the disfavored.