OF CROSSES AND CRECHES: THE ESTABLISHMENT CLAUSE AND PUBLICLY SPONSORED DISPLAYS OF RELIGIOUS SYMBOLS* INTRODUCTION Religions often communicate their beliefs and teachings through symbolism.' Religious symbols are an integral part of religious2 practice and are inextricably linked to the beliefs they represent. Consequently, if a government or private group displays a religious symbol on public property, or if the government uses public funds to underwrite such a display, the exhibit serves not only to communicate a religious message, but also implies governmental support for that message. 3 The government's support may be an unconstitutional establishment of the religion whose symbolism is publicly 4 displayed. * © 1985 Joshua D. Zarrow. 1. A symbol communicates ideas and beliefs by representing or suggesting them; it is an authoritative synopsis of a faith or doctrine. WEBSTER'S NINTH NEW COLLEGIATE DICrIONARY 1195 (1985). From the earliest days of civilization, mankind has relied on religious symbolism to express beliefs and convey religious convictions. SeeJ. Moss, MAN'S RELIGIONS 5-10 (1969) (observing that religious symbols in form of cave paintings and stone carvings played important roles in early religious life). 2. The Supreme Court has recognized that religious symbols convey more than their immediate meaning: "The Church speaks through the Cross, the crucifix, the altar and shrine." West Virginia State Bd. of Educ. v. Barnette, 319 U.S. 624, 632 (1943). According to psychologist CarlJung, when an individual views an object in the external world, his unconscious triggers instinctual processes that create connotations beyond the viewer's control. See C.G.JuNG, 9 COLLECTED WORKS 141-42 (1978). Thus, a symbol conveys more than its obvious and immediate meaning. 3. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 701 (1984) (Brennan, J., dissenting) (including creche in city display places government's imprimatur on religious beliefs represented); Stone v. Graham, 449 U.S. 39, 42 (1980) (per curiam) (posting Ten Commandments in public school connotes governmental support for religion); Gilfillan v. City of Philadelphia, 637 F.2d 924, 930 (3d Cir. 1980) (using city funds in connection with Papal visit implies state support of religion), cert. denied, 451 U.S. 987 (1981); Fox v. City of Los Angeles, 22 Cal. 3d 794, 794, 587 P.2d 663, 665. 150 Cal. Rptr. 867, 869 (1978) (displaying cross on public property connotes support for religion represented); see also Chase, Litigating A Nativity Scene Case, 24 ST. Louis U.LJ. 237, 239 (1980) (reasoning that public displays of nativity scenes represent governmental support of Christian dogma). 4. U.S. CONsT. amend. I. The establishment clause to the United States Constitution forbids Congress from making laws respecting an establishment of religion. Id. The Supreme Court has emphasized that this clause not only prohibits the establishment of a national or state religion, but also prohibits laws with respect to that forbidden objective. See Lemon v. 477 478 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 The first amendment to the United States Constitution forbids the government from establishing religion. 5 To determine whether particular governmental actions violate this establishment clause, the Supreme Court employs a three-part test. 6 The Court's application of this tripartite test proscribes governmental activities that have either the purpose or primary effect of advancing religion, 7 or that excessively entangle the government in religion.8 Courts have applied the test to the display of "religious" 9 symbols on public property with varying results.10 The Supreme Court has eroded this traditional establishment clause analysis in recent years.' 1 Indeed, in recent decisions, the Supreme Court has elected to forego application of the tripartite Kurtzman, 403 U.S. 602, 612 (1971) (noting that Constitution forbids laws respecting establishment as well as laws establishing religion); Engel v. Vitale, 370 U.S. 421, 429-80 (1962) (stating that first amendment bars more than establishing state religion); Everson v. Board of Educ., 330 U.S. 1, 15-16 (1947) (holding that first amendment bars state or federal church as well as other state or federal incursions and influences on religion). See generally C. KRUSE, THE HISTORICAL MEANING AND JUDICIAL CONSTRUCTION OF THE FIRST AMENDMENT 85 (1962) (noting that although framers feared state-established church, judiciary construes establishment clause to forbid mere facilitation of religion). 5. U.S. CONST. amend. I. The Supreme Court made the establishment clause applicable to the states through the due process clause of the fourteenth amendment. Everson v. Board of Educ., 330 U.S. 1, 15 (1947) (incorporating establishment clause into fourteenth amendment). 6. See Lemon v. Kurtzman, 403 U.S. 602, 612-13 (1971) (establishing analytical framework to determine when government action violates establishment clause); infra notes 34-63 and accompanying text (discussing establishment clause doctrine). 7. Lemon v. Kurtzman, 403 U.S. 602, 612 (1971). Prior to Lemon, the Supreme Court often applied a "purpose and effect" test. See, e.g., Epperson v. Arkansas, 393 U.S. 97, 107 (1968) (using purpose and effect test to invalidate state law barring theory of evolution from public schools); Board of Educ. v. Allen, 392 U.S. 236, 238 (1968) (employing purpose and effect test to uphold religiously neutral book loaning program); School Dist. of Abington Township v. Schempp, 374 U.S. 203, 222 (1963) (invoking purpose and effect test to prohibit mandatory bible reading). The Court combined the purpose and effect test with an entanglement test in Lemon. Lemon v. Kurtzman, 403 U.S. 602, 612-13 (1971). 8. Lemon v. Kurtzman, 403 U.S. 602, 613 (1971). The Supreme Court first articulated the entanglement prong in Walz v. Tax Comm'n, 397 U.S. 664, 674-75 (1970). 9. Determining whether a symbol is religious or secular may not only influence the application of the tripartite Lemon test, but may actually be outcome determinative. Compare Lynch v. Donnelly, 465 U.S. 668, 687 (1984) (creche is not religious and survives Lemon review) with American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1339 (E.D. Mich. 1984) (creche is religious, therefore fails Lemon review). 10. See, e.g., Stone v. Graham, 449 U.S. 39, 41 (per curiam) (1980) (Ten Commandments on public school walls fails tripartite review); Allen v. Morton, 495 F.2d 65, 67 (D.C. Cir. 1973) (per curiam) (creche on public property cannot withstand establishment clause analysis); Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1020-23, 558 P.2d 338, 34748 (1976) (cross on public property survives Lemon review), cert. denied sub nom. Lowe v. Eugene Sand & Gravel, Inc., 434 U.S. 876 (1977). 11. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 680-85 (1984) (rejecting inquiry into religious effect of symbol in favor of review of secular purpose of symbol); Marsh v. Chambers, 463 U.S. 783, 786-92 (1983) (using historical analysis in lieu of Lemon test to justify religious practice); Larson v. Valente, 456 U.S. 228, 246-55 (1982) (applying strict scrutiny rather than Lemon analysis); see also infra notes 66-77 and accompanying text (discussing Supreme Court's limitations to traditional doctrine). 1986] OF CROSSES AND CRECHES 479 test completely. 12 In Lynch v. Donnelly, 13 the Supreme Court discarded the traditional tripartite test in favor of both an historical analysis' 4 and a muted version of the traditional establishment clause test' 5 to uphold the display of a publicly owned creche in a private park.' 6 The Lynch decision has caused considerable confusion among the circuits. 17 It is now unclear which analysis applies to establishment clause issues,' 8 and what constitutes a religious symbol.' 9 Local governments, therefore, are left with little guidance in determining whether they constitutionally may display certain symbols. 20 The result has been political controversy and religious divisiveness in communities where allegedly religious symbols have 21 been displayed. 12. See Marsh v. Chambers, 463 U.S. 783, 786-92 (1983) (utilizing historical analysis to uphold prayer in legislature); Larson v. Valente, 456 U.S. 228, 244-46 (1982) (using strict scrutiny test to strike down law discriminating against one religion). 13. 104 S. Ct. 1355 (1984). 14. Id. at 1359-61. The Court justified a publicly sponsored creche display by reciting instances of governmental recognition of religion. Id. See also infra notes 114-18 and accompanying text (discussing Court's use of history in Lynch). 15. Id. at 1363-65. See also infra notes 119-34 and accompanying text (discussing Court's limitations on traditional test in Lynch). 16. Lynch v. Donnelly, 465 U.S. 668, 671 (1984). The creche display was owned and operated by the City of Pawtucket, Rhode Island, but exhibited in a privately owned park. Id. 17. See, e.g., McCreary v. Stone, 739 F.2d 716, 723-30 (2d Cir. 1984) (interpreting Supreme Court decision in Lynch as supporting publicly sponsored creche display without seasonal figures), aff'd mem. by an equally divided Court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985); Fausto v. Diamond, 589 F. Supp. 451, 464-70 (D.R.I. 1984) (basing application of Lemon test on Lynch decision to uphold monument in public park); American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 231-38 (S.D. Tex. 1984) (applying three different tests to display of religious symbols); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1338-40 (E.D. Mich. 1984) (finding it necessary to distinguish Lynch to apply traditional Lemon analysis to religious displays). 18. See McCreary v. Stone, 739 F.2d 716, 723-30 (2d Cir. 1984) (applying Lynch analysis to publicly sponsored creche display), aff'd mem. by an equally divided Court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1338-40 (E.D. Mich. 1984) (applying traditional Lemon analysis to publicly sponsored creche display); American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 231-38 (S.D. Tex. 1984) (applying three different tests to publicly sponsored display of religious symbols). 19. Compare Lynch v. Donnelly, 465 U.S. 668, 679 (1984) (concluding that symbol must be viewed in context to determine religious nature) and McCreary v. Stone, 739 F.2d 716, 729 (2d Cir. 1984) (viewing symbol in its seasonal context to determine its religious nature), aff'd mem. by an equally divided Court sub nona. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985) with American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1340 (E.D. Mich. 1984) (judging symbol in physical context to determine its religious nature). See also infra notes 152-61 and accompanying text (discussing ad hoc definitions for religious symbols). 20. Compare American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1339-40 (E.D. Mich. 1984) (finding city's creche display unconstitutional) and American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 241 (S.D. Tex. 1984) (declaring publicly sponsored display of crosses and Star of David unconstitutional) with Fausto v. Diamond, 589 F. Supp. 451, 470 (D.R.I. 1984) (upholding constitutionality of right-to-life monument to unknown child jointly supported by town and group with religious affiliation). 21. See N.Y. Times, Dec. 20, 1984, at B2, col. 1 (discussing conflict along religious lines caused by Scarsdale creche); N.Y. Times, Dec. 20, 1984, at B2, col. 4 (discussing religious THE 480 AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 Part I of this Comment examines the development of the establishment clause analysis and recent limitations to traditional doctrine. Part II discusses the application of traditional doctrine to religious symbols displayed on public property. Part III examines the Lynch decision. In Part IV, this Comment reviews the impact of Lynch on courts deciding religious symbol cases. Part V then analyzes the doctrinal and theoretical problems inherent in establishment clause analyses of government-sponsored, religious symbols. Finally, Part VI suggests a two-part analysis that accommodates the policies behind the first amendment and provides a uniform method for resolving establishment clause disputes. I. THE DEVELOPMENT OF ESTABLISHMENT CLAUSE DOCTRINE The United States Constitution forbids Congress from establish- ing or inhibiting religious practices. 22 The Framers' experiences with religious intolerance 23 compelled them to separate church controversy caused by nativity scene in Central Park); Wash. Post, Dec. 18, 1984, at A13, col. 6 (discussing plans of American Civil Liberties Union to oppose erection of twenty-foot tall menorah in Grand Rapids, Michigan); Wash. Post, Dec. 12, 1984, at C3, col. 2 (noting that display of creche in Washington, D.C. sparked small holy war); Wash. Post, Nov. 28, 1984, at Alb, col. 1 (notingJewish opposition to, and Christian support of, creche on White House ellipse). 22. U.S. CONST. amend. I. The Constitution provides that "Congress shall make no law respecting an establishment of religion or prohibiting the free exercise thereof...." Id. 23. See Everson v. Board of Educ., 330 U.S. 1, 9-14 (1947) (discussing Framers' experiences with religious intolerance). Although not always dispositive in resolving constitutional issues, the views of the Framers serve to illuminate the motivations behind the first amendment. Professor Tribe describes three views of church-state separation prevalent among the Framers: (1) the evangelical view-without a separation of church and state, the church would be corrupted; (2) the Jeffersonian view-separation would safeguard secular interests; and (3) the Madisonian view-religion and state are advanced by competition between religious sects, rather than domination by one. L. TRIBE, AMERICAN CONSTITUIONAL LAw, § 14-3, at 816-17 (1978). See also M. HowE, THE GARDEN AND THE WILDERNESS: RELIGION AND GOVERNMENT IN AMERICAN HISTORY 1-31 (1965) (describing prevalent views of church-state relations during early American history). Although there is some debate as to who actually wrote the first amendment, compare Engel v. Vitale, 370 U.S. 421, 436 (1962) (Madison wrote first amendment) and Everson v. Board of Educ., 330 U.S. 1, 33 (1947) (RutledgeJ., dissenting) (same) with C. KRUSE, supra note 4, at 72 (first amendment written by committee), it is, nevertheless, clear that both Madison and Jefferson were strong proponents of the separation of church and state. See Van Alstyne, Trends in the Supreme Court: Mr.Jefferson's Crumbling Wall - A comment on Lynch v. Donnelly, 1984 DUKE LJ. 770, 773 (Madison and Jefferson both advocated separation of church and state). The religious domination of the Anglican Church in Great Britain and the establishment of colonial and state churches from the seventeenth through eighteenth centuries in America contributed to a perceived need for church-state separation among the founding fathers. In his Memorial and Remonstrance Against Religious Assessments, James Madison decried the religious intolerance caused by close church-state ties: "Torrents of blood have been spilt in the old world, by vain attempts of the secular arm to extinguish Religious discord, by proscribing all difference in Religious opinions." J. MADISON, THE COMPLETE MADISON, His BASIC WRITINGS 304 (1971). See generally THE FEDERALIST No. 10 J. Madison) (Tudor ed. 1937) (deploring all factionalism that leads to conflict in society). Lessons taught by centuries of religious persecution convinced the Framers to "disestab- 1986] OF CROSSES AND CRECHES 481 from state24 through the two religion clauses contained in the first amendment: the establishment clause and the free exercise clause. 25 Although the two clauses often overlap in theory, the Supreme Court traditionally has reviewed laws burdening the practice of religion under the free exercise clause, and laws aiding a particular religion under the establishment clause. 26 Because displays of religious symbols on public property imply governmental support for the underlying religious message of the symbol displayed, 27 the applicable review is under the establishment clause. The Supreme Court first thoroughly analyzed the establishment clause in Everson v. Board of Education.28 In Everson, the Court re- viewed the constitutionality of a New Jersey program reimbursing lish" religion. See generally L. PFEFFER, CHURCH, STATE AND FREEDOM (1967) (discussing evolution of church-state separation in America and abroad); W. SwEET, THE STORY OF RELIGION IN AMERICA (1950) (discussing religion in early America). 24. See 16 THE WRITINGS OF THOMAS JEFFERSON 281-82 (Library ed. 1904) (describing ideal church-state relations). In his letter to the Danbury Baptist Association, Thomas Jefferson used a "wall of separation" as a metaphor to describe the relationship between church and state. Id. The Supreme Court subsequently adopted this metaphor. See Everson v. Board of Educ., 330 U.S. 1, 18 (1947) (stating that first amendment has erected "wall" between church and state). 25. U.S. CONST. amend. I. See supra note 22 (text of religion clauses). 26. Although the Supreme Court has developed distinct analyses for each clause, the free exercise and establishment clauses often overlap. See Abington School Dist. v. Schempp, 374 U.S. 203, 222 (1963) (describing overlap of two clauses); L. TRIBE, supra note 23, § 14-2, at 813-15 (detailing relationship of two clauses). See generally, Moore, The Supreme Court and the Relationship Between the "Establishment" and "Free Exercise" Clauses, 42 TEX. L. REV. 142 (1963) (two clauses are not entirely independent); Pfeffer, Freedom and/or Separation: The Constitutional Dilemma of the First Amendment, 64 MINN. L. REV. 561 (1980) (discussing relationship between two clauses); Comment,A Non-Conflict Approach to the FirstAmendment Religion Clauses, 131 U. PA. L. REV. 1175 (1983) (discussing Supreme Court's analyses of conflict between clauses). Despite the possible overlap between the two clauses, the Supreme Court has generally addressed cases in terms of either the establishment clause or the free exercise clause. Compare Wisconsin v. Yoder, 406 U.S. 205, 233-34 (1972) (free exercise clause prohibits penalty on Amish for refusing to send children to school) and Sherbert v. Verner, 374 U.S. 398, 409-10 (1963) (free exercise clause bars denial of unemployment benefits for woman who refused to work on Sabbath although establishment clause not offended) with Larkin v. Grendel's Den, Inc., 459 U.S. 116, 126-27 (1982) (establishment clause invalidates law giving church right to control liquor licensing) and Levitt v. Committee for Pub. Educ. & Religious Liberty, 413 U.S. 472,479-82 (1973) (establishment clause prohibits state reimbursement of parochial schools). The Supreme Court generally applies a free exercise analysis to state laws that burden the practice of religion. The Court balances the state's compelling interest in the law burdening religion against the individual's interest in the free exercise of his religion. See, e.g., Wisconsin v. Yoder, 406 U.S. 205, 221 (1972) (religious beliefs of Amish outweigh state interest in education); Braunfield v. Brown, 366 U.S. 599, 607 (1961) (state's interest in Sunday closing law outweighs OrthodoxJew's desire to work on Sunday); cf. Cantwell v. Connecticut, 310 U.S. 296, 311 (1940) (restrictions on freedom of religion permissible only to prevent grave and immediate danger); L. PFEFFER, supra note 23, at 617 (interest of community only supersedes free exercise if interest cannot be protected without incursion into religion). 27. See supra notes 1-3 and accompanying text (discussing implications of publicly sponsored or displayed religious symbols). 28. 330 U.S. 1 (1947). Although several establishment clause cases preceded Everson, they have had little impact on establishment clause doctrine. See Quick Bear v. Leupp, 210 U.S. 50, 81-82 (1908) (federal money can be used for Indian education at Catholic mission); Bradfield v. Roberts, 175 U.S. 291, 299-300 (1899) (federal money may fund church-run hos- THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 parents for their children's transportation costs to parochial or public schools, an action that equally benefited a religious institution and a nonreligious institution.2 9 The Court reviewed the history of religion in the United States3" to support its conclusion that the government must be strictly neutral towards religion. 3 ' The Court held, however, that the New Jersey reimbursement practice maintained the requisite neutrality 3 2 In subsequent cases, the Court used the historical analysis it had developed in Everson to strengthen the establishment clause test by holding that the government cannot 33 aid religion either directly or indirectly. A. The Tripartite Test Based on the principle of neutrality articulated in Everson, and the subsequent use of historical analysis to support that principle, the Supreme Court developed a doctrinal approach to resolve establishment clause issues. 34 In Lemon v. Kurtzman35 the Court crystallized the doctrine into a tripartite test. 36 In Lemon, the Court ruled that state actions must not have the purpose of advancing religion, 3 7 the primary effect of advancing religion, 38 or excessively entangle the pital); Watson v. Jones, 80 U.S. (13 Wall.) 679, 733-34 (1871) (civil courts cannot decide issues of religious doctrine). 29. Everson v. Board of Educ., 330 U.S. 1, 29 (1947). 30. Id. at 9-15. The Court traced religious persecution from Europe through the colonial period. Id. 31. Id. at 18. 32. Id at 29. 33. See Engel v. Vitale, 370 U.S. 421, 431 (1962) (history dictates that government cannot aid religion); McGowan v. Maryland, 366 U.S. 420, 453 (1961) (history establishes that state's coercive power cannot aid religion); McCullom v. Board of Educ., 333 U.S. 203, 212 (1948) (history teaches that government may not indirectly aid religion). In contrast to this prohibition against aid to religion was the Supreme Court's conclusion that "we are a religious people whose institutions presuppose a Supreme Being." Zorach v, Clauson, 343 U.S. 306, 313 (1952) (upholding state law allowing public school students to leave class to pray and receive religious instructions). Courts continuously have cited this proposition from Zorach to uphold state action. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 675 (1984) (supporting creche); Marsh v. Chambers, 463 U.S. 783, 792 (1983) (upholding legislative prayer); Fausto v. Diamond, 589 F. Supp. 451, 465 (D.R.I. 1984) (upholding monument in public park). 34. See supra note 7 (discussing post-Everson,. pre-Lemon development of traditional establishment clause doctrine). The Supreme Court's first articulations of the establishment clause doctrine prohibited governmental actions that had a purpose or primary effect of advancing religions. School Dist. of Abington Township v. Schempp, 374 U.S. 203, 222 (1963). The Court later prohibited government and religion from becoming excessively entangled with one another. Walz v. Tax Comm'n, 397 U.S. 664, 670 (1970). The Court eventually combined the three elements-purpose, effect, and entanglement-into a tripartite test. Lemon v. Kurtzman, 403 U.S. 602, 612-13 (1971). 35. 403 U.S. 602 (1971). 36. Id. at 612-13. 37. Id. at 612. 38. Id. 1986] OF CROSSES AND CRECHES 483 government in religious matters.3 9 If a state action violates any of the three prongs in the Lemon analysis, that action is unconstitutional. 40 Courts have applied the Lemon tripartite test for more than 41 a decade. The first prong of the Lemon test requires that the government must have a secular purpose to justify its actions. 4 2 If a court does not believe the secular purpose proposed by the state, it will attempt to discern for itself the state's "actual" intent. 43 The Supreme Court, for example, concluded that Kentucky had religious motivations for requiring the posting of the Ten Commandments in public school rooms, even though the purpose of the requirement offered 44 by Kentucky was to highlight the United States' legal foundations. This purpose prong of the Lemon test has rarely been dispositive in establishment clause cases. 4 5 Indeed, recent Supreme Court deci39. Id. at 613. Conversely, religious groups cannot exercise control over government. Larkin v. Grendel's Den, Inc., 459 U.S. 116, 127 (1982). 40. See Larkin v. Grendel's Den, Inc., 459 U.S. 116, 123 (1982) (to withstand establishment clause scrutiny, all three criteria must be met); Stone v. Graham, 449 U.S. 39, 40-41 (1980) (per curiam) (violation of any prong requires invalidation); Wolman v. Walter, 433 U.S. 229, 235-36 (1977) (each prong must be met); Committee for Pub. Educ. v. Nyquist, 413 U.S. 756, 772-73 (1973) (mandatory to pass each prong). 41. See, e.g., Committee for Pub. Educ. v. Regan, 444 U.S. 646, 653-63 (1980) (using tripartite test to uphold state practice of reimbursing parochial schools); Levitt v. Committee for Pub. Educ. & Religious Liberty, 413 U.S. 472, 482 (1973) (applying test to forbid certain reimbursements to private schools); Nartowicz v. Clayton County School Dist., 736 F.2d 646, 648-50 (11th Cir. 1984) (holding that Lemon criteria supported district court's preliminary injunction); Florey v. Sioux City Falls School Dist., 619 F.2d 1311, 1314-19 (8th Cir.) (using Lemon test to uphold use of religious themes in secular instruction), cert. denied, 449 U.S. 987 (1980); Bonjour v. Bonjour, 592 P.2d 1233, 1242-44 (Alaska 1979) (using tripartite test to invalidate child custody determination based on religion); Keegan v. University of Del., 349 A.2d 14, 16 (Del. 1975) (upholding use of university facilities for religious meetings under Lemon analysis), cert. denied, 424 U.S. 934 (1976); Americans United v. Rogers, 538 S.W.2d 711, 716-18 (Mo. 1976) (using Lemon analysis to uphold state tuition grants to private colleges), cert. denied, 429 U.S. 1029 (1977). 42. Lemon v. Kurtzman, 403 U.S. 602, 612 (1971). The Supreme Court generally has been able to find secular motivations for governmental actions, even when such actions subsequently were declared unconstitutional. See, e.g., Larkin v. Grendel's Den, Inc., 459 U.S. 116, 123-27 (1982) (holding that although purpose of law was acceptable, effect of law was not); Meek v. Pittenger, 421 U.S. 349, 367-70 (1975) (ruling that educational development of children was valid purpose underlying law invalidated on other grounds); Lemon v. Kurtzman, 403 U.S. 602, 613 (1971) (concluding that promoting secular education was valid purpose although promotional means were unacceptable). 43. See Stone v. Graham, 449 U.S. 39, 39 (1980) (per curiam) (looking beyond avowed secular purpose of state action to action's "plainly" religious purpose); Epperson v. Arkansas, 393 U.S. 97, 107 (1968) (finding religious purpose for state prohibition on evolution instruction); Gilfillan v. City of Philadelphia, 637 F.2d 924, 930 (3d Cir. 1980) (characterizing city's stated secular purpose as suspect and finding its action to be religiously motivated), cert. denied, 451 U.S. 987 (1981); Hall v. Bradshaw, 630 F.2d 1018, 1020 (4th Cir. 1980) (rejecting state's argument that prayer on state map had secular purpose and finding religious purpose). 44. Stone v. Graham, 449 U.S. 39, 41 (1980) (per curiam). 45. In the Supreme Court, the purpose prong has disposed of only three cases. See Wallace v. Jaffree, 105 S. Ct. 2479, 2490 (1985) (holding silent prayer statute unconstitutional based on finding that purpose of law was religious); Stone v. Graham, 449 U.S. 39, 41 (1980) (per curiam) (holding statute requiring posting of Ten Commandments in classroom uncon- 484 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 sions have diminished significantly the impact of this prong by neglecting to go beyond a superficial review of the government's 46 stated purpose. Under the second prong of Lemon, a state action may have the impermissible primary effect of advancing religion and, therefore, violating the establishment clause, even though the court finds a valid secular purpose for the action. 47 The Supreme Court initially held that a religious effect is permissible only if it remotely or incidentally benefits religion. 48 Over the years, however, the Supreme Court has increasingly limited the application of the effect prong.49 The Court, for example, recently has suggested that empirical evistitutional on grounds that law had no secular purpose); Epperson v. Kansas, 393 U.S. 97, 107 (1968) (declaring law prohibiting instruction in evolution unconstitutional based on conclusion that sole purpose of law was religious). The purpose prong has played a greater role in the lower federal courts. See, e.g., American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1110 (11 th Cir. 1983) (determining that religious purpose to construction of state-maintained cross violated establishment clause); Gilfillan v. City of Philadelphia, 637 F.2d 924, 930 (3d Cir. 1980) (finding that city expenditures related to Papal visit had religious purpose and violated establishment clause), cert. denied, 451 U.S. 987 (1981). 46. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 680 (1984) (holding that purpose behind state action need not be exclusively secular so long as secular purpose also exists); Mueller v. Allen, 463 U.S. 388, 394-95 (1983) (refusing to search for motivation beyond stated purpose); Widmar v. Vincent, 454 U.S. 263, 271 (1981) (intejecting hypothetical purpose into analysis to uphold religious meetings on campus as constitutional); see also L. TRIBE, supra note 23, § 14-8, at 835 (discussing loose application of purpose prong); Comment, Publicly-FundedDisplay of Religious Symbols: The Nativity Scene Controversy, 51 U. CN. L. REv. 353, 358-60 (1982) (detailing limitations on purpose prong); infra notes 64-77 and accompanying text (discussing recent limitations to Lemon doctrine). 47. See Grand Rapids School Dist. v. Ball, 105 S. Ct. 3216, 3222 (1985) (stating that although school programs had secular purpose, they had invalid religious effect); Larkin v. Grendel's Den, Inc., 459 U.S. 116, 122 (1982) (reasoning that religious effect of zoning ordinance was impermissible despite valid secular legislative purpose); Wolman v. Walter, 433 U.S. 229, 250-54 (stating that ostensible secular purpose does not save statute that has primary effect of advancing religion); Committee for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 794-95 (1973) (reasoning that despite secular purpose, law has impermissible religious effect); cf. Sloan v. Lemon, 413 U.S. 825, 832 (1973) (determining that state law has secular purpose but religious effect). 48. Committee for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 783-84 n.39 (1973) (holding that legitimate secular objective does not save statute that has more than incidental effect on religion). 49. See, e.g., Roemer v. Board of Pub. Works, 426 U.S. 736, 755-59 (1976) (noting aid is permissible even though it has both secular and religious effects); Hunt v. McNair, 413 U.S. 734, 743 (1973) (holding unless religious effect is primary, it is not constitutionally offensive). The Supreme Court recently has broadened the scope of permissible effects by limiting the scope of its effect prong inquiry. In Lynch v. Donnelly, 465 U.S. 668 (1984), for example, the Court concluded simply that the effect at issue was no worse than other effects found permissible in the past. Id. at 683. Based on this conclusion derived without an analysis, the Court found the effect permissible. Id. See infra notes 125-29 (discussing Lynch's effect prong analysis). The Supreme Court never has held that the establishment clause bars all aid to religion. See Tilton v. Richardson, 403 U.S. 672, 679 (1971) (stating that there is no absolute bar of aid from state to church); Walz v. Tax Comm'n, 397 U.S. 664, 671 (1970) (reasoning that constitutional history does not require conclusion that any aid from state to church is impermissible). 1986] OF CROSSES AND CRECHES 485 dence that a governmental action aids religion is not, by itself, an acceptable means of establishing an unconstitutional effect. 50 This development limits dramatically the impact of the effect prong by requiring that a governmental action must obviously and unmistakably aid religion before it Will be invalidated. 5 1 The constitutionality of state action that has a religious effect often depends on the type of aid a government provides and on whom the aid benefits. 52 If the aid flows to an institution pervaded by religion, the assistance is likely to have an impermissible religious effect.5 3 Moreover, if the church-state involvement affects impressionable children, the law is unlikely to survive. 54 If, however, a 50. See Mueller v. Allen, 463 U.S. 388, 402 (1983) (rejecting empirical inquiry into effect of allowing parents of parochial students to take state tax deduction for certain educational expenses). 51. See The Supreme Court, 1982 Term, 97 HARV. L. REV. 154 (1983) (discussing impact of Mueller decision limiting use of empirical evidence to address effect prong of establishment clause test). 52. See, e.g., Larkin v. Grendel's Den, Inc., 459 U.S. 116, 117-18 (1982) (invalidating zoning ordinance that primarily benefited churches); Stone v. Graham, 449 U.S. 39, 41-42 (1980) (per curiam) (indicating that government benefits believers by symbolically supporting Ten Commandments posted on walls of public schools). The cases involving state aid to parochial education highlight the types of situations in which aid is acceptable or prohibited. The Supreme Court has found a number of constitutionally acceptable uses for state funds in parochial schools. See Committee for Pub. Educ. v. Regan, 444 U.S. 646, 648 (1980) (upholding use of state funds to provide secular textbooks, standardized tests, and certain services to parochial schools); Roemer v. Board of Pub. Works, 426 U.S. 736, 739, 753 (1976) (approving use of noncategorical grants for private colleges); Tilton v. Richardson, 403 U.S. 672, 675, 689 (1971) (upholding use of federal grants and loans to parochial schools for expanding facilities). In contrast, the Court has held a number of other forms of government aid to parochial schools unconstitutional. See Grand Rapids School Dist. v. Ball, 105 S. Ct. 3216, 3230-31 (1985) (invalidating state assistance in form of shared time and community education programs open to students at nonpublic schools); Wolman v. Walter, 433 U.S. 229, 255 (1977) (invalidating use of state money for field trips, instructional material, and equipment); Meek v. Pittenger, 421 U.S. 349, 372 (1975) (invalidating use of state funds for guidance counselors in parochial schools); Committee for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 799 (1973) (invalidating use of tax benefits by parochial school parents); Lemon v. Kurtzman, 403 U.S. 602, 606-07 (197 1) (invalidating state supplements to parochial school teachers and reimbursements for parochial school costs). 53. Hunt v. McNair, 413 U.S. 734, 743 (1973). In Hunt, the Court stated that aid has a primarily religious effect if it "flows to an institution in which religion is so pervasive that a " Id. See also substantial portion of its functions are subsumed in the religious mission .. Roemer v. Board of Pub. Works, 426 U.S. 736, 762 (1976) (noting that effect is not religious unless religion pervades institution). 54. See, e.g., Grand Rapids School Dist. v. Ball, 105 S. Ct. 3216,3222 (1985) (invalidating shared-time educational program by focusing on impact on children); Meek v. Pittenger, 421 U.S. 349, 349 (1975) (invalidating state funding program due to impermissible effect on school children); Levitt v. Committee for Pub. Educ. & Religious Liberty, 413 U.S. 472, 480 (1973) (invalidating state aid to sectarian schools because of danger that children will be inculcated by religion); cf Lemon v. Kurtzman, 403 U.S. 602, 616 (1971) (positing that religious effect of state aid to parochial schools is enhanced by impressionability of school children); McCullom v. Board of Educ., 333 U.S. 203, 227 (1948) (opinion of FrankfurterJ.) (reasoning that children are especially affected by church-state relations); Jaffree v. Wallace, 705 F.2d 1526, 1536-37 (1 1th Cir. 1983) (prohibiting moment of silence in public school), aff'dsub nor. 105 S. Ct. 2479 (1985). 486 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 government aids pervasively secular programs in religious universities, the religious effect is likely to be found inconsequential.55 The third prong of Lemon provides that, even though a state action has neither the purpose nor primary effect of advancing religion, it nevertheless may violate the establishment clause if it excessively entangles government and religion.5 6 Although courts strive for neutrality, the Supreme Court has concluded that absolute separation between church and state is impossible. 57 Consequently, the Court has attempted to prohibit only those state actions that excessively entangle government and religion.58 In Lemon, the Court articulated several factors for discerning the presence of excessive entanglement. 5 9 Courts must look to the character and purposes of the institutions benefited, the nature of the aid, and the resulting 60 relationship between church and state. The Supreme Court has delineated two types of entanglement: administrative and political. 6 1 Administrative entanglement involves government supervision or control over religious activities, or a religious entity's administration or control over government functions. 6 2 Political entanglement concerns the divisiveness en55. See generally Widmar v. Vincent, 454 U.S. 263, 277 (1981) (requiring university to allow religious groups to use facilities that are generally available to student groups); Roemer v. Board of Pub. Works, 426 U.S. 736, 764-65 (1976) (upholding constitutionality of grants to private colleges); Hunt v. McNair, 413 U.S. 734, 749 (1973) (upholding use of state money for secular program in religious college). Congress recently resolved the issue of whether school space must be made available to high school religious groups. See Equal Access Act, 20 U.S.C. § 4071 (1982) (requiring public schools receiving federal money to grant access of facilities during open forum to religious and nonreligious groups). 56. Lemon v. Kurtzman, 403 U.S. 602, 613 (1971). 57. See id. at 614 (stating that total separation of chuch and state is not feasible); Walz v. Tax Comm'n, 397 U.S. 664, 674-75 (1970) (reasoning that no perfect separation between church and state is possible); Zorach v. Clauson, 343 U.S. 306, 313-14 (1952) (indicating that absolute neutrality of church and state is constitutionally unnecessary). But see Note, Rebuilding the Wall The Casefor a Return to the Strict Interpretationof the Establishment Clause, 81 COLUM. L. REv. 1463, 1463 (1981) (arguing for total separation). 58. Lemon v. Kurtzman, 403 U.S. 602, 615 (1971). 59. Id. 60. A court's inquiry into the character and purpose of an institution benefiting from state aid focuses on whether such character and purpose are secular or religious. Id. The same inquiry is made into the nature of the aid provided. Id. at 616. States may provide secular services, facilities, and materials to all citizens. Id. Accordingly, a court may inquire as to the religious nature of the aid that a state provides. Id. at 617. Finally, courts review the relationship that arises between church and state as a result of state aid. A court, for example, may strike down direct payments to a parochial school if it determines that the resulting relationship between the government and the school, in which the government retains an interest in reviewing the school's records to ensure that the money is used for secular purposes, is excessively entangling. Id at 621. 61. See Lynch v. Donnelly, 465 U.S. 668, 684 (1984) (distinguishing between administrative and political entanglement). 62. See Larkin v. Grendel's Den, Inc., 459 U.S. 116, 126-27 (1982) (church may not exercise traditionally governmental power relating to distribution of liquor licenses); Lemon v. Kurtzman, 403 U.S. 602, 620 (1971) (government may not direct church schools). Adminis- 19861 OF CROSSES AND CRECHES 487 63 gendered in communities affected by the church-state relations. B. Recent Developments Following Lemon, the Supreme Court uniformly applied the tripartite criteria in establishment clause cases 64 despite the criticisms leveled against the test. 65 In cases other than those involving schools, trative entanglement exists where the government must supervise the administration of state aid to ensure that it is not used for sectarian purposes. See Aguilar v. Felton, 105 S. Ct. 3232, 3239 (1985) (invalidating financial aid to parochial schools because it requires ongoing supervision); Wolman v. Walter, 433 U.S. 229, 254 (1977) (noting that Constitution prohibits supervision of parochial teachers by public school authorities); Meek v. Pittenger, 421 U.S. 349, 372 (1975) (supervising use of state aid is excessively entangling); Walz v. Tax Comm'n, 397 U.S. 664, 675 (1970) (noting that church and state may not become administratively entangled). The Court also has recognized a type of "reverse" administrative entanglement. In Walz v. Tax Comm'n, 397 U.S. 664 (1970), the Court held that tax exemptions that benefitted owners of property used for religious, educational, or charitable purposes were not excessively entangling. Id. at 680. The Court noted that if the exemptions were forbidden, the government would have to administer tax valuations, liens, and foreclosures and that involvement would excessively entangle government in religion. Ide at 674. Similarly, the Court in Widmar v. Vincent, 454 U.S. 263 (1981), concluded that trying to avoid "establishment" would lead to greater entanglement than continuing to allow student religious groups to use university buildings. Id. at 272 n.11 (1981). 63. See Larson v. Valente, 456 U.S. 228, 254-55 (1982) (noting that discriminatory administration of religious solicitation engenders religious politicization); Meek v. Pittenger, 421 U.S. 349, 372 (1975) (discussing political fragmentation); Committee for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 797 (1973) (discussing divisive political consequences of state assistance to parochial schools); Lemon v. Kurtzman, 403 U.S. 602, 622 (1971) (noting that first amendment was written to prevent political divisiveness along religious lines); Everson v. Board of Educ., 330 U.S. 1, 8-9 (1947) (discussing political conflicts arising from religion); W. SWEET, RELIGION IN COLONIAL AMERICA 320-22 (1942) (noting political conflict in Colonies caused by close church-state ties); Curry, James Madison and the Burger Court. Converging Views of Church-State Separation, 56 IND. LJ. 615, 615 (1981) (discussing Madison's concern with political divisiveness); Freund, PublicAid to ParochialSchools, 82 HARV. L. REV. 1680, 1692 (1969) (discussing issue of political entanglement); see also supra note 21 and accompanying text (discussing situations of actual community division engendered by church-state controversy). The Supreme Court never has invalidated a state action for political entanglement alone. See Lynch v. Donnelly, 465 U.S. 668, 684 (1984) (noting that Supreme Court never has disposed of case on finding of political entanglement); Committee for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 797-98 (1973) (stating that political entanglement alone cannot invalidate statute); Gaffney, PoliticalDivisivenessAlong Religious Lines: The Entanglement of the Court is Sloppy History and Bad Public Policy, 24 ST. Louis U.L.J. 205, 226-27 (1980) (stating that political divisiveness has not been sole criterion for invalidating state actions that implicate establishment clause). 64. See, e.g., Larkin v. Grendel's Den, Inc., 459 U.S. 116, 123-27 (1982) (barring church power to regulate liquor licensing under Lemon test); Committee for Pub. Educ. & Religious Liberty v. Regan, 444 U.S. 646, 653-62 (1980) (using tripartite test to uphold disbursement of certain state funds to parochial schools); Meek v. Pittenger, 421 U.S. 349, 359-73 (1975) (applying Lemon test to analyze constitutionality of state aid to parochial schools); cf Widmar v. Vincent, 454 U.S. 263, 270-77 (1981) (applying Lemon test in conjunction with strict scrutiny analysis to uphold policy of providing religious groups with access to school facilities). 65. See, e.g., H. ABRAHAM, FREEDOM AND THE COURT 207 (2d ed. 1972) (describing establishment clause interpretations as "hornet's nest"); Choper, The Religious Clauses of the First Amendment: Reconciling the Conflict, 41 U. PrTr. L. REV. 673, 680 (1980) (criticizing ad hoc, unreconcilable judgments under Lemon test); Cornellius, Church and State - The Mandate of the Establishment Clause: Wall of Separation or Benign Neutrality?, 16 ST. MARY LJ. 1, 15-19 (1984) 488 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 however, the Court recently has indicated a willingness to narrow, and even abandon, traditional establishment clause doctrine. 66 In departing from an exclusive application of the Lemon test the Court, at different times, has employed an historical analysis and a strict scrutiny review of state practices challenged as violative of the estab67 lishment clause. With only a passing mention of the Lemon criteria, the Court in Marsh v. Chambers68 disposed of an establishment clause challenge primarily on the basis of historical analysis. 69 The Court in Marsh justified the state practice of opening a state legislature with prayer by reciting the long history of religious invocations before legislative gatherings. 70 The Court's sole reference to the traditional Lemon test came in its discussion of the appellate court's use of the (pointing to disarray resulting from Lemon doctrine); Howard, Up Against the Wall: The Uneasy Separation of Church and State, in CHURCH, STATE AND PoLrrmcs 5, 21 (J. Hensel ed. 1981) (claiming that Supreme Court establishment clause cases are nonsensical). The Supreme Court has suggested that Lemon ought not to be rigidly applied. See Mueller v. Allen, 463 U.S. 388, 394 (1983) (suggesting that Lemon test is only helpful signpost); Committee for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 773, n.31 (1973) (citing Tilton v. Richardson, 403 U.S. 672 (1971), on need for flexible standard in establishment clause cases); Hunt v. McNair, 413 U.S. 734, 741 (1973) (suggesting that Lemon test is not rigidly applied standard); Tilton v. Richardson, 403 U.S. 672, 677-78 (1971) (recognizing that establishment clause cases cannot be judged by rigid tests). 66. See Lynch v. Donnelly, 465 U.S. 668, 680-85 (1984) (narrowing impact of each prong of Lemon test to uphold city-sponsored creche display); Marsh v. Chambers, 463 U.S. 783, 795 (1983) (abandoning Lemon test in favor of historical analysis to uphold legislative prayer); Larson v. Valente, 456 U.S. 228, 252 (1982) (discussing Lemon test only after noting its nondispositive nature and after having used strict scrutiny analysis to invalidate charitable solicitation statute that benefited certain religions over others); see also infra notes 68-77 and accompanying text (discussing recent trends in Supreme Court establishment clause analysis in cases other than those involving public aid to parochial education). During its last term, the Supreme Court demonstrated that it will continue to apply the Lemon test in cases involving religion in public schools and state aid to parochial schools. See Aguilar v. Felton, 105 S. Ct. 3232, 3239 (1985) (relying on entanglement prong of Lemon test to invalidate use of federal funds to pay salaries of public school employees who taught in parochial schools); Grand Rapids School District v. City of Grand Rapids, 105 S. Ct. 3216, 3230 (1985) (utilizing effect prong of Lemon test to invalidate public school district's shared time and community education programs located in parochial schoools); Wallace v. Jaffree, 105 S. Ct. 2479, 2493 (1985) (employing purpose prong of Lemon test to invalidate state statute authorizing prayer and meditation in school). The Supreme Court traditionally has imposed strict standards in its review of cases involving religious influence in public schools. See supra note 54 and accompanying text (discussing Court's efforts to stem religious influences in public schools). 67. See, e.g., Marsh v. Chambers, 463 U.S. 783, 786 (1983) (utilizing historical analysis). 68. Id. 69. Id 70. Id. at 786-95. The Court used history, unlike the pre-Lemon cases, to stress the coextensive nature of church and state. Id. Prior to Marsh, the Court had used history to illustrate the persecution and divisiveness that historically attended any union of church and state. See, e.g., Engel v. Vitale, 370 U.S. 421, 431-32 (1962) (noting that governmentally established religions and religious persecution go hand-in-hand); Everson v. Board of Educ., 330 U.S. 1, 8-13 (1946) (discussing persecution and divisiveness engendered by state-supported reli- gions). For a discussion of the malleable nature of history as a doctrinal tool, see infra notes 115-17. See also supra note 33 and accompanying text (discussing Court's early use of history). 1986] OF CROSSES AND CRECHES 489 7 doctrine. ' Furthermore, the Supreme Court has suggested that the Lemon analysis applies only in cases where the state action uniformly benefits all religions. 72 Thus, in Larson v. Valente,73 the Court ruled that state actions that discriminate among religions must withstand a strict scrutiny examination. 74 Nevertheless, the Court in Larson invalidated a state charitable solicitation statute that discriminated against certain religions under both the tripartite test and strict scrutiny.7 5 Although the Larson decision appears to create a stricter establishment clause review, 76 it further obfuscates the application of 71. Marsh v. Chambers, 463 U.S. 783, 786 (1983). In his dissent, Justice Brennan suggested that had the Supreme Court applied the Lemon test to the state practice, the practice would have been declared unconstitutional. Id at 800-01 (Brennan, J., dissenting). Justice Brennan wrote: "In sum, I have no doubt that, if any group of law students were asked to apply the principles of Lemon to the question of legislative prayer, they would nearly unanimously find the practice to be unconstitutional." Idt (BrennanJ., dissenting). Justice Brennan argued that the Court based its decision on sentiment rather than analysis. Id. at 796 (Brennan, J., dissenting). See generally Note, Legislative Prayerand the Establishment Clause: An Exception to the TraditionalAnalysis, 17 CREIGHTON L. REV. 157, 160 (1983) (criticizing Court's failure to use Lemon test in its decision in Marsh v. Chambers). 72. Larson v. Valente, 456 U.S. 228, 252 (1982). Larson involved a Minnesota law that burdened some religions but not others. Id at 250-51. Although the state's action might have violated the free exercise clause by burdening religious practices, and the Court appeared to apply a free exercise analysis, the Court actually reviewed the case under the establishment clause. Id. at 245-51. For a discussion of free exercise analysis, see supra note 26. The potential impact of the Supreme Court's statement that Lemon only applies to laws giving a uniform benefit remains unclear. A literal interpretation of the statement would preclude application of the Lemon test to laws that benefit Christians more than Jews, or Methodists more than Baptists. It is difficult to conceive of a government action that benefits all religions, because any given action has a limited impact. The Court may have intended to require application of a strict scrutiny analysis to intentional state discrimination among religions, while limiting application of the Lemon test to state action that is not intended to benefit specific religions. The Court's assertion on the proper application of the Lemon test was not analytically supported, nor was it supported by prior case law. See infra note 77 (highlighting cases where benefits were not uniform but Lemon was applied). 73. 456 U.S. 228 (1982). 74. Id. at 252. To be upheld under Larson, a law that grants denominational preferences must have a compelling government interest supported by a means closely fitted to that end. Id. at 244-47. The fact that the Court appears to be using free exercise analysis may limit the precedential value of the case for establishment clause purposes. 75. Id. at 246-55. The Supreme Court also applied a combination of strict scrutiny and establishhnent clause analysis in Widmar v. Vincent, 454 U.S. 263, 275-77 (1981). Based on the Supreme Court's holding in Larson, however, it appeared that the Court in Widmar should have applied only the Lemon test to a state action that uniformly discriminated against all religions. Compare Larson v. Valente, 456 U.S. 228, 252 (1982) (requiring application of Lemon test to laws that have uniform application to all religions; limiting application of strict scrutiny analysis to laws that treat different religions differently) with Widmar v. Vincent, 454 U.S. 263, 275-77 (1981) (applying strict scrutiny and Lemon analysis to discriminatory policy that applies uniformly to all religions). 76. Larson v. Valente, 456 U.S. 228, 248 (1982) (applying strict scrutiny analysis to find violation of establishment clause). Strict scrutiny appears to be a more stringent standard for establishment clause analysis in view of the malleable nature of the Lemon criteria. See supra notes 46-63 (discussing malleable nature of Lemon criteria). The Lemon test requires merely that state action have a secular purpose, whereas strict scrutiny requires that state action have a compelling state purpose. Compare Lemon v. Kurtzman, 403 U.S. 602, 612 (1971) (requiring secular purpose) with Larson v. Valente, 456 U.S. 228, 246-47 (1982) (requiring compelling 490 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 establishment clause analysis. 7 7 II. THE TRIPARTITE TEST AND RELIGIOUS SYMBOLS Despite the recent departures from the Lemon analysis, courts generally have continued to apply the tripartite test to displays of religious symbols on public property.7 8 Religious symbols raise establishment clause concerns because they convey a religious meaning to the viewer. 79 By displaying a religious symbol on public property, or by financing such a display on private property, the sponsoring government may convey the impression of support for the religion represented. 80 Moreover, in choosing to finance or dispurpose). Moreover, Lemon does not require that the secular law be narrowly tailored to the secular purpose. Lemon v. Kurtzman, 403 U.S. 602, 611-25 (1971) (silent regarding "narrowly tailored" requirement). In strict scrutiny, however, there must be a close ends/means fit. Larson v. Valente, 456 U.S. 228, 248 (1982) (requiring that government end be supported by closely fitted means). 77. See Note, Another Brick in the Wall: DenominationalPreferences and Strict Scrutiny Under the Establishment Clause, 62 NEB. L. REv. 359, 383 (1983) (commenting on confusion over test now applicable to establishment clause cases). Despite its stringency, strict scrutiny fails to account for laws that have a primarily religious effect or are excessively entangling. Moreover, it is unclear when the Lemon tripartite test criteria, as opposed to the Larson strict scrutiny review, must apply. Id. To date, courts have applied Lemon in cases where parochial schools have benefited from state aid. See Committee for Pub. Educ. v. Regan, 444 U.S. 646, 653-62 (1980) (applying Lemon test to state reimbursements to parochial institution); Wolman v. Walter, 433 U.S. 229, 236-55 (1977) (noting that state funding for nonpublic school must withstand Lemon test); Meek v. Pittenger, 421 U.S. 349, 358-73 (1975) (analyzing services and funds to private schools under tripartite test). Such benefits could not have been uniform because many religions have no educational institutions. Consequently, under Larson, the courts in these cases should have applied strict scrutiny in lieu of the Lemon test. See Larson v. Valente, 456 U.S. 228, 252 (1982) (limiting application of Lemon test to state actions that apply uniformly to all religions). Displays of religious symbols on public property discriminate against those religions that have no symbols, or whose symbols are not displayed. See infra note 81 and accompanying text (discussing decisions that have noted discriminatory element of public displays of religious symbols). Nevertheless, courts have applied the Lemon test in these cases. See infra notes 78-113 and accompanying text (discussing use of Lemon test in religious symbols cases). 78. See, e.g., Stone v. Graham, 449 U.S. 39, 40 (1980) (per curiam) (using Lemon test to determine constitutionality ofTen Commandments posted on walls of public schools); American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1109 (11 th Cir. 1983) (applying Lemon test to determine constitutionality of cross in public park); Allen v. Morton, 495 F.2d 65, 68 (D.C. Cir. 1973) (per curiam) (analyzing constitutionality of creche display on White House ellipse under Lemon test); Paul v. Dade County, 202 So. 2d 833, 835 (Fla. Dist. Ct. App. 1967) (reviewing constitutionality of cross on courthouse building using Lemon test). 79. See supra notes 1-2 and accompanying text (discussing communicative qualities of symbols). A viewer's own beliefs impact on his or her perceptions of a symbol. See Greenwalt, Religion as a Concept in ConstitutionalLaw, 72 CAL. L. REV. 753, 794 (1984) (stating that viewer's beliefs determine perception of symbol as religious or secular). In a controversy involving the placement of a menorah and creche in Central Park, for example, the New York Times reported that although Jewish leaders perceived the menorah as comparable to a Christmas tree, Christian leaders saw the menorah as comparable to a creche. N.Y. Times, Dec. 19, 1984, at BI, col. 5. 80. See, e.g., Gilfillan v. City of Philadelphia, 637 F.2d 924, 930 (3d Cir. 1980) (state aid connotes state approval of religious sect), cert. denied, 451 U.S. 987 (1981); Donnelly v. Lynch, 525 F. Supp. 1150, 1171 (D.R.I. 1981) (publicly supported nativity scene affiliates city with 1986] OF CROSSES AND CRECHES 491 play one symbol, or several symbols, the government necessarily 8 discriminates against the unrepresented religions. ' The threshold problem facing the courts in applying establishment clause analysis, therefore, is determining whether an object is a religious symbol. This, however, has proved no easy task. Although the courts have decided that the words "In God We Trust" are not religious,8 2 the Supreme Court has ruled that the Ten Commandments are patently religious. 83 The absence of useful guidelines with which to gauge an object's religious nature has created conflict among the states and circuits and wildly different appli84 cations of the Lemon standard. A. Crosses 85 The courts applying the tripartite test to the display of crosses on public property have reached conflicting results.8 6 In cases striking down the display of crosses, courts have found violations of each prong of the Lemon analysis.8 7 Either explicitly or implicitly, these Christianity), aff'd, 691 F.2d 1029 (Ist Cir. 1982), rev'd, 465 U.S. 668 (1984); Fox v. City of Los Angeles, 22 Cal. 3d 792, 798, 587 P.2d 663, 665, 150 Cal. Rptr. 867, 869 (1978) (cross display connotes state support for Christianity); see also Chase, supra note 3, at 239 (publicly sponsored nativity scenes indicate governmental support for Christian dogma). 81. See Stone v. Graham, 449 U.S. 39, 41 (1980) (per curiam) (posting Ten Commandments represents only Judeo-Christian beliefs); Goldstein v. Fire Dep't, 559 F. Supp. 1389, 1391 (S.D.N.Y. 1983) (indicating that fire department sign saying "Keep Christ in Christmas" signals governmental support only for Christianity); Fox v. City of Los Angeles, 22 Cal. 3d 794, 798, 587 P.2d 663, 665, 150 Cal. Rptr. 867, 869 (1978) (stating that city display of cross is preferential toward Christianity to exclusion of other religions). 82. See Aronow v. United States, 432 F.2d 242, 243 (9th Cir. 1970) (stating that slogan "In God We Trust" has no religious impact); Opinion of theJustices, 108 N.H. 97, 102, 228 A.2d 161, 164 (1967) (advising New Hampshire Senate that plaque saying "In God We Trust" in public classrooms is constitutional); see also Lynch v. Donnelly, 465 U.S. 668, 676 (1984) (stating in dicta that slogan on currency is constitutional); c. 36 U.S.C. § 186 (1982) (establishing "In God We Trust" as national motto); 31 U.S.C. § 5112(d)(1) (1982) (requiring currency to proclaim our trust in God). But see Pfeffer, The Deity in American ConstitutionalHistory, 23 J. OF CHURCH & ST. 215, 238 (1981) (noting Framers' opposition to governmental invocation of deity). 83. Stone v. Graham, 449 U.S. 39,42-43 (1980) (per curiam) (noting sectarian nature of Ten Commandments). Contra Anderson v. Salt Lake City Corp., 475 F.2d 29, 34 (10th Cir.) (concluding that statue of Ten Commandments on public property is secular and does not violate establishment clause), cert. denied, 414 U.S. 879 (1973). 84. See infra notes 85-109 and accompanying text (discussing conflicting analyses and characterizations in religious symbol cases). 85. WEBSTER'S NINTH NEW COLLEGIATE DICTIONARY 308 (1985). Webster's identifies the cross as a symbol adopted by Christians because of its resemblance to the instrument ofJesus' crucifixion. 86. Compare American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1110 (1 1th Cir. 1983) (placement of cross in public park violates establishment clause under tripartite test) with Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1020-26, 558 P.2d 338, 346-48 (1976) (cross utilized as war memorial survives tripartite test), cert. denied sub nom. Lowe v. Eugene Sand & Gravel, Inc., 434 U.S. 876 (1977). 87. See, e.g., American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1110 (11 th Cir. 1983) (placement of cross in public park violates purpose 492 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 courts have determined that a cross is a religious symbol.88 The United States Court of Appeals for the Eleventh Circuit, for example, has held that the State of Georgia violated the establishment clause by erecting a twenty-six feet wide by thirty-five feet tall lighted cross in a state park.8 9 In ruling that the cross was erected out of "religious stirrings," the Eleventh Circuit's analysis simply assumed that a cross was a religious symbol. 90 The court, therefore, held that the religiously motivated display violated the purpose prong of the Lemon test.9 1 Despite the inherent religious nature of crosses, 9 2 some courts have held that their display survives the Lemon analysis. 93 In 1969, prong of Lemon test); Gilfillan v. City of Philadelphia, 637 F.2d 924, 929-34 (3d Cir. 1980) (papal platform with 36 foot tall cross violates each prong of Lemon test), cert. denied, 451 U.S. 987 (1981). 88. See American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1110 (11th Cir. 1983) (analysis of purpose prong predicated on assumption that cross is religious symbol); Gilfillan v. City of Philadelphia, 637 F.2d 924, 929-32 (3d Cir. 1980) (court's purpose, effect, and entanglement analyses presuppose religious nature of cross), cert. denied, 451 U.S. 987 (1981); Fox v. City of Los Angeles, 22 Cal. 3d 792, 795-96, 587 P.2d 663, 664, 100 Cal. Rptr. 867, 868 (1978) (takingjudicial notice of religious nature of cross). At least one court implicitly has revealed its recognition of the cross as a religious symbol by analyzing a cross case by refering to previous opinions dealing with religious symbols. American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1110 nn.22-23 (11th Cir. 1983). 89. American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1111 (11th Cir. 1983). 90. Id. at 1110. The state initially claimed that the purpose of the cross was to attract tourists. Id. After receiving objections to the cross, the government considered designating the cross as a war memorial but the state never adopted this designation. Id. at 1101-02. But see Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1020-26, 558 P.2d 338, 34648 (1976) (designating cross as war memorial makes it secular), cert. denied sub nom. Lowe v. Eugene Sand & Gravel, Inc., 434 U.S. 876 (1977). 91. American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1110 (11 th Cir. 1983). The court further held that, even if the state's purpose in erecting the cross was secular, using a religious means to reach a secular goal was unconstitutional. Id. at 1111. The court distinguished the Supreme Court's decision in Larson by contending that strict scrutiny review is applicable only to cases in which the state refuses to display other religious symbols. Id. at 1109, n.20. 92. See supra note 85 (discussing religious nature of crosses). 93. See Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1020-26, 558 P.2d 338, 346-48 (1976) (concluding that cross designated as war memorial survives Lemon analysis), cert. denied sub nom. Lowe v. Eugene Sand & Gravel, Inc., 434 U.S. 876 (1977); Meyer v. Oklahoma City, 496 P.2d 789, 792 (Okla.) (ruling that display of cross on public fair grounds is constitutional under state constitution), cert. denied, 409 U.S. 980 (1972). It is not clear why federal standards were not applied in Meyer. States are bound by the federal prohibition against establishment of religion, unless the states apply stricter standards than those imposed by the United States Constitution. See supra note 5 (discussing applicability of U.S. Constitution to states). The Oklahoma Constitution forbids use of public property or money for the support ofany sect, church, or sectarian institution. OKLAHOMA CONST. art. I, § 5. The court in Meyer ruled that the cross at issue did not violate the state constitution because it did not "display, articulate, or portray, except in a most evanescent form," any sectarian ideas. Meyer v. Oklahoma City, 496 P.2d 789, 792 (Okla.), cert. denied, 409 U.S. 980 (1972). The court asserted, without support, that the commercial setting obscured any religiosity the cross may have contained. Id. at 792-93. 1986] OF CROSSES AND CRECHES 493 for example, the Oregon Supreme Court ordered the City of Eugene to remove a large cross from a public park. 94 The city refused to dismantle the cross and subsequently designated it as a war memorial. 95 In 1976 the Oregon Supreme Court concluded that the same cross it had declared unconstitutional seven years earlier now satisfied the Lemon test.96 The court predicated its new conclusion 97 on the assumption that a war memorial is secular. B. Creches Similar definitional and doctrinal discrepancies are present in cases involving the displays of crosses can be found in cases involving the display of nativity scenes. Although some courts have found that the portrayal of Christ's birth is inherently religious, 98 other courts have characterized it as secular. 9 9 The experience of two federal courts illustrates this tension; both applied the Lemon criteria to creche displays but each arrived at a different result. 10 0 The United States Court of Appeals for the District of Columbia Circuit ruled that a creche display on the White House Ellipse violated the entanglement prong of the tripartite test.' 0 ' In contrast, however, the 94. Lowe v. City of Eugene, 254 Or. 518,463 P.2d 360,361 (1969) (en banc), cert. denied, 397 U.S. 1042 (1970). 95. Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1009, 558 P.2d 338, 340 (1976), cert. denied, 434 U.S. 876 (1977). 96. L The court ruled that (1) the city's purpose was to memorialize war veterans, (2) a commemorative plaque reduced any religious effect, and (3) there was no excessive entanglement. Id 97. Id at 346. The court did not, however, explain how changing the designation of a cross diluted its religious nature. Moreover, the court also failed to explain why only Christian soldiers were memorialized or why religious means were used to effectuate "secular" ends. See id. at 346-49 (focusing analysis on secular nature of war memorial). The Supreme Court has suggested that a religious means cannot be used to achieve a secular goal. See School Dist. of Abington Township v. Schempp, 374 U.S. 203, 294 (1963) (Brennan, J., concurring) (arguing that secular purpose cannot be met with religious means); see also American Civil Liberties Union v. Rabun County Chamber of Commerce, 698 F.2d 1098, 1111 (11th Cir. 1983) (ruling that state cannot use religious means to reach secular ends). 98. See, e.g., Allen v. Morton, 495 F.2d 65, 73 (D.C. Cir. 1973) (per curiam) (finding that creche is "obvious" religious symbol); Burrelle v. City of Nashua, 599 F. Supp. 792, 797 (D.N.H. 1984) (ruling that privately owned creche on public grounds is religious symbol); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1341 (E.D. Mich. 1984) (ruling that creche maintained and displayed by government was religious). 99. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 685 (1984) (creche in holiday display is secular); Allen v. Hickel, 424 F.2d 944, 949 (D.C. Cir. 1970) (religious nature of creche secularized by its historical significance); Citizens Concerned for Separation of Church and State v. City of Denver, 526 F. Supp. 1310, 1313-14 (D. Colo. 1981) (creche on steps of government building is secular). 100. Allen v. Morton, 495 F.2d 65 (D.C. Cir. 1973) (per curiam) (finding display of privately owned creche in federally owned park violation of establishment clause); Citizens Concerned for Separation of Church and State v. City of Denver, 526 F. Supp. 1310 (D. Colo. 1981) (finding display of publicly owned creche on steps of government building constitutional). 101. Allen v. Morton, 495 F.2d 65, 67 (D.C. Cir. 1973) (per curiam). 494 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 United States District Court for the District of Colorado concluded that a creche displayed on public property withstood each prong of the Lemon review.10 2 Although both courts noted that nativity scenes may be religious symbols, each concluded that Christmas has evolved into a secular holiday, which has limited the extent of the 10 3 nativity displays' religious nature. C. Other Symbols Courts confronted with public displays of arguably religious symbols, other than crosses or creches have applied the tripartite test with varying results. In one instance, the United States Court of Appeals for the Tenth Circuit ruled that a granite monolith installed on public grounds and inscribed with the Ten Commandments and symbols representing both the Star of David and Christ passed traditional Lemon review.10 4 Conversely, the Supreme Court has ruled that the Lemon test would not allow the posting of the Ten Commandments in public schools. 05 Other religious writings on public property have failed the Lemon review as well. Courts have held unconstitutional a sign on a local firehouse inscribed picting a tripartite pronouncing "Keep Christ in Christmas"' 1 6 and a prayer on an official state map. 10 7 A county seal, however, decross and sheep on a pastoral background, survived the analysis.' 0 8 Although the courts have applied the Lemon 102. Citizens Concerned for Separation of Church and State v. City of Denver, 526 F. Supp. 1310, 1311-15 (D. Colo. 1981). 103. See Allen v. Morton, 495 F.2d 65, 69 (D.C. Cir. 1973) (per curiam) (noting dilution of creche's religious nature by display in secular holiday setting); Citizens Concerned for Separation of Church and State v. City of Denver, 526 F. Supp. 1310, 1313-1,t (D. Colo. 1981) (finding religious nature of creche diminished by increasingly secular aspect of holiday season). Some commentators have suggested that the Christmas season invests the manger scene with religious meaning, rather than divesting it of religiosity. Plastic figures of a child, sheep, shepherds, angels, and two parents are not inherently religious. Rather, it is the seasonal celebration of Christ's birth that makes the scene religious. See generally R. BROWN, ThE BIRTH OF THE MESSIAH (1977) (discussing importance of Christ's nativity to Christian religion); Elliott, The Birth and Background ofJesus of Nazareth, 28 HIsT. TODAY 773 (1978) (explaining significance that story of Christ's birth holds for Christians). 104. Anderson v. Salt Lake City Corp., 475 F.2d 29, 34 (10th Cir.) (recognizing both secular and religious aspects of Ten Commandments before ruling that they were acceptable passive representation of America's legal foundation), cert. denied, 414 U.S. 879 (1973). The court did not explain, however, how the inscription of a Star of David, letters of the Hebrew alphabet, the Seeing Eye of God, and a symbol of Christ represented America's legal foundation. lId at 30. 105. Stone v. Graham, 449 U.S. 39, 41 (1980) (per curiam) (Ten Commandments fail purpose prong because "plainly" religious). 106. Goldstein v. Fire Dep't, 559 F. Supp. 1389 (S.D.N.Y. 1983). 107. Hall v. Bradshaw, 630 F.2d 1018 (4th Cir. 1980). 108. Johnson v. Board of County Comm'rs, 528 F. Supp. 919, 924-25 (D.N.M. 1981) (iling that county seal, historical and political symbol, survives Lenton review as not religious). aff'd sub nom. Friedman v. Board of Comm'rs, No. 82-1064 (10th Cir. Dec. 27, 1984). 1986] OF CROSSES AND CRECHES 495 doctrine in these cases, they have reached disparate results because no consensus existed as to when a religious symbol becomes "secularized." 1 0 9 III. LYNCH V. DONNELLY: THE Two PLASTIC REINDEER RULE The Supreme Court's decision in Lynch v. Donnelly110 exacerbated the two areas of confusion wrought by the Court's previous opinions: which establishment clause analysis is proper and what constitutes a religious symbol. In Lynch, the Court reversed the lower courts' rulings that a publicly sponsored creche display violated the Lemon test.1 1 ' The creche display at issue was owned and operated by the city of Pawtucket, Rhode Island, and located in a privately owned park. 1 12 Plastic figures of Santa Claus, reindeer, and other seasonal characters surrounded the nativity scene113 A. The Court's Establishment Clause Analysis The Court first employed an historical analysis to justify the constitutionality of the creche display. 114 Rather than describing a his- tory of religious persecution as it had done in pre-Lemon cases," 15 the Court chose instead to recite historical instances of state-recognized, religious practices. 116 The Court, for example, emphasized official references to God in our motto and currency, the nationwide celebration of Thanksgiving, and the display of religious art by the 109. Compare Stone v. Graham, 449 U.S. 39, 41-42 (1980) (per curiam) (ruling that Ten Commandments were religious despite their influence on western law) with Anderson v. Salt Lake City Corp., 475 F.2d 29, 34 (10th Cir.) (finding that Ten Commandments had lost their religious nature based on their influence on western law), cert. denied, 414 U.S. 879 (1973). The creche cases exhibit the same inconsistencies concerning the courts' determination of the secular or sectarian nature of symbols. See supra notes 100-03 and accompanying text. 110. 465 U.S. 668 (1984). 111. Id. at 687. The United States District Court for the District of Rhode Island ruled that the creche display violated all three prongs of the Lemon test. Donnelly v. Lynch, 525 F. Supp. 1050, 1169-80 (D.R.I. 1981), aff'd, 691 F.2d 1029 (1st Cir. 1982), reu'd, 465 U.S. 668 (1984). The United States Court of Appeals for the First Circuit ruled that the creche display violated the purpose prong of the Lemon test and declined to rule on the other prongs of the test. Donnelly v. Lynch, 691 F.2d 1029, 1035 (1st Cir. 1982), rev'd, 465 U.S. 668 (1984). 112. Lynch v. Donnelly, 465 U.S. 668, 671 (1984). 113. Id. The Supreme Court apparently determined that the plastic figures surrounding the creche diminished the manger scene's religiosity. Id. at 681. Commentators refer to the Lynch decision as "The Two Plastic Reindeer Rule" in reference to the Supreme Court's focus on the seasonal figures surrounding the Pawtucket creche. 114. Id. at 674-78. 115. See, e.g., Engel v. Vitale, 370 U.S. 421, 425-29 (1962) (noting history of religious persecution in context of decision on governmentally composed prayer); McGowan v. Maryland, 366 U.S. 420, 431-34 (1961) (referring to history of religious persecution in context of decision on Sunday closing laws); Everson v. Board of Educ., 330 U.S. 1, 8-13 (1947) (tracing history of religious persecution that influenced adoption of first amendment). 116. Lynch v. Donnelly, 465 U.S. 668, 679-80 (1984). 496 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 National Gallery in Washington, D.C.' 17 The Court concluded that the government's historical acknowledgment of religion supported the acceptance of a religious display in this instance. IIs After applying an historical analysis to examine the constitutionality of the creche display, the Court insisted that it was unwilling to be bound by the Lemon standard. 119 The Court nevertheless applied the tripartite Lemon test. 120 The Court first applied the purpose prong of Lemon, and concluded that the government's motivation behind displaying the creche need not be exclusively secular. 12 1 The Court further ruled that the government need only posit a single secular purpose to survive the purpose prong. 12 2 Consequently, as Justice Brennan observed in his dissenting opinion, even if the government's action was motivated primarily by "religious stirrings," the action would survive constitutional review if any secular motivation could be found.' 23 The Court explained that the nonreligious purpose of depicting the holiday's origins was a permissible secular purpose and, therefore, passed the first prong of the 12 4 tripartite test. The Court next addressed the effect prong of Lemon, analyzing the 117. Id These references may be distinguished from the creche display at issue in Lynch. Official references to God are nondenominational and, although they may be perceived as religious, they benefit no specific religion. A creche display, however, only can "benefit" Christians. Similarly, Thanksgiving is a holiday that recognizes God in an equally nondenominational way. Although Thanksgiving may establish religion over nonreligion, it does not establish Christianity over all religions or over nonreligion. The displays of art are similarly distinguishable. Religious art has a dual nature: religious and aesthetic. See infra notes 218-20 (discussing dual nature of religious objects). The aesthetic and cultural aspect of art is displayed at the National Gallery. It is difficult to perceive how a creche, made up of cheap plastic figures, could display similar secular aspects. Professor Pfeffer argues that the Court's reliance on references to governmental recognition of God to support government actions that ostensibly support religion glosses over substantive issues with circular reasoning. L. PFEFFER, supra note 23, at 239-40. He contends that official references to God are constitutional only because they have not been judged unconstitutional. Id Thus, a practice of unproved constitutionality is used to prove the constitutionality of another practice. Id. 118. Lynch v. Donnelly, 465 U.S. 668, 676-77 (1984). 119. Id. at 639. 120. Id. at 680-85. For an interesting analysis of the Supreme Court's application of the tripartite test in Lynch see VAN ALSTYNE, supra note 23, at 787-84 (arguing that Lynch's application of tripartite review constitutes new test). 121. Lynch v. Donnelly, 465 U.S. 668, 681 n.6 (1984). 122. Id. 123. Id. at 699-700 (Brennan,J., dissenting) (arguing that secular motivations behind display did not diminish fact that creche was included for religious purpose). 124. Id at 685. The Court argued that, because Christ's birth is recognized as a national holiday, the origins of that birth may be depicted. Id. at 680. Under this transparent analysis, however, any time the government chooses a religion for preferential treatment, it can circumvent the establishment clause and openly display religious symbols merely by reciting the symbol's history. Justice Brennan argued in dissent that "a narrower sectarian purpose lay behind the decision to include a nativity scene" in the holiday display. Id. at 700 (BrennanJ., dissenting). 19861 OF CROSSES AND CRECHES 497 religious effect of the nativity scene by placing it on a continuum with other controversies involving governmental support for reli1 26 gion. 125 The Court listed effects found permissible in the past, and concluded that the effect of the creche must exceed those historically permissible effects to be unconstitutional.127 Under this analysis, the Court decided that the creche had only an incidental effect,' 2 8 no worse than the government aid to religion found ac29 ceptable in the past.' In applying the entanglement prong of the Lemon test, neither the lower courts 130 nor the Supreme Court' 3 ' found any administrative entanglement. Although the district court did find that the lawsuit had created an atmosphere of political divisiveness, "of anger, hostility, name calling and political maneuvering," 13 2 the Supreme Court held that the mere "appearance" of political divisiveness created by a lawsuit did not demonstrate excessive political entanglement.' 3 3 Moreover, the Court ruled that divisiveness alone cannot 13 4 invalidate an otherwise permissible act. B. The Court's Analysis of Religious Symbols The Court in Lynch also was faced with the problem of determin125. Ia at 681-83. 126. Id. Although the Court did make superficial reference to such cases, it is unclear why the Court did not measure the effect on a continuum using effects found impermissible in the past. Id. at 682-83. See AMERICAN JEWISH COMMISSION ON LAW & SOCIAL ACrION REPORT 2425 (Mar. 1984) (noting that Court did not list prayer decisions that had impermissible religious effects). The Court could have judged the creche's religious effect against those effects that the Court has held unconstitutional. See, e.g., Larkin v. Grendel's Den, Inc., 459 U.S. 116, 127 (1982) (ruling that churches' veto power over liquor licensing is unconstitutional effect); Wolman v. Walter, 433 U.S. 229, 248-55 (1977) (holding that state provision of instructional materials and services to private schools impermissibly advances religion); Meek v. Pittenger, 421 U.S. 349, 372 (1975) (holding state provision of auxiliary services to parochial schools unconstitutional benefit to religion); Sloan v. Lemon, 413 U.S. 825, 835 (1973) (ruling that state reimbursement for tuition costs of private schools has unconstitutional effect). 127. Lynch v. Donnelly, 465 U.S. 668, 681 (1984). 128. Id. at 682-83. The Court provided no criteria for judging when one effect is more religious than another. Without any guidelines, courts may be free to build on the list of permissible effects. There is a danger, moreover, that, by substantially adding to the list, almost any religious aid could become constitutionally acceptable. 129. L at 683. 130. See Donnelly v. Lynch, 691 F.2d 1029, 1035 (1st Cir. 1982) (deciding case on effect prong and declining to rule on entanglement), rev'd, 465 U.S. 668 (1984); Donnelly v. Lynch, 525 F. Supp. 1150, 1178-80 (D.R.I. 1981) (recognizing element of political entanglement), aft'd, 691 F.2d 1029 (Ist Cir. 1982), rev'd, 465 U.S. 668 (1984). 131. See Lynch v. Donnelly, 465 U.S. 668, 683-85 (1984). 132. Donnelly v. Lynch, 525 F. Supp. 1150, 1180 (D.R.I. 1981) (characterizing political reaction as horrifying), aff'd, 691 F.2d 1029 (Ist Cir. 1982), rev'd, 465 U.S. 668 (1984). 133. Lynch v. Donnelly, 465 U.S. 668, 684-85 (1984). 134. Id. In dicta, the Court stated that inquiries into political divisiveness are required only when direct subsidies are given to religious institutions. Id. at 684. 498 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 ing whether an object is a religious symbol.' 35 The majority noted that, although the creche itself could be viewed as a religious symbol,' 3 6 when viewed in the Christmas context, it served merely as an 1 37 historical symbol of a particular religious event. IV. LYNCH'S APPLICATION: JUDICIAL DISARRAY The Supreme Court's decision in Lynch v. Donnelly' 38 has created confusion and conflict among the federal courts attempting to apply the establishment clause doctrine in religious symbol cases.' 3 9 Local governments, confronted by such judicial disaccord, have been left with little guidance in determining whether their actions impermissibly aid or accommodate religion. 140 Such uncertainty has engendered widespread controversy and debate, often along religious lines.141 A. Creches Since Lynch, two federal courts have addressed the issue of whether a creche, standing alone, is a religious symbol. Each applied the Lynch decision; each reached a different result. In McCreary v. Stone, 14 2 the Village of Scarsdale, New York, refused to allow a 135. Id at 685. 136. Id at 686. The Court also noted that although it had religious connotations, the creche engendered a friendly community spirit and helped bring tourists into the city. Id. Viewed from the perspective of minority religions, it is unclear whether the creche will create "good will" or attract a minority religion's practitioners to the site of the display. 137. Id. It is unclear how the seasonal context makes the creche less religious. On the contrary, it would seem that the celebration of Christ's birth invested the display with rcli- gious meaning. See supra note 103 (arguing that Christmas context does not make manger scene less religious). 138. 465 U.S. 668 (1984). 139. See, e.g., McCreary v. Stone, 739 F.2d 716, 725-30 (2d Cir. 1984) (referring to Supreme Court decision in Lynch to define creche without seasonal figures as nonreligious), aff'd mem. by an equally divided court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985); American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 233-38 (S.D. Tex. 1984) (applying several tests to find crosses and Star of David violative of establishment clause); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1339-40 (E.D. Mich. 1984) (using Supreme Court decision in Lynch to characterize creche without seasonal figures as religious). 140. See supra notes 18-20 and accompanying text (highlighting cases resulting from government's placement of symbols on property). 141. See supra note 21 (political debate highlighted in national newspapers). The Court observed in Lemon that "political division along religious lines was one of the principal evils against which the First Amendment was intended to protect." Lemon v. Kurtzman, 403 U.S. 602, 622 (1971). See also Committee for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 796 (1973) (stating that competition among religious sects for political recognition causes civil strife). 142. 739 F.2d 716 (2d Cir. 1984), aff'd mem. by an equally divided rowt sub now. Board or Trustees v. McCreary, 105 S. Ct. 1859 (1985). The Supreme Court affirmed the Second Circuit decision by memorandum in a four to four vote. Justice Powell taking no part in the result. In so doing, the Court left unanswered the questions raised by Lynch and its progeny. 1986] OF CROSSES AND CRECHES 499 private group to erect a nativity scene in a public park. 143 The proposed creche was to stand alone, without a Santa Claus, reindeer, or other seasonal figures. In American Civil Liberties Union v. City of Birmingham' 44 the City of Birmingham, Michigan, refused to dismantle city, that stood without the accompaa creche display, owned by the 45 niment of seasonal figures.' 1. The establishment clause analyses applied by the courts In Birmingham, the United States District Court for the Eastern District of Michigan distinguished Lynch on its facts and applied the traditional tripartite test, holding that the Birmingham creche display was unconstitutional. 46 In McCreary, however, the United States Court of Appeals for the Second Circuit upheld the creche clause analysis the Supreme display based on the establishment 47 Court had developed in Lynch. ' In McCreary, the Second Circuit accepted a hypothetical secular purpose without inquiring into the religious motivations behind the creche exhibit.' 48 The court judged the effect of the Scarsdale creche on a continuum: if the effect of the Lynch creche did not impermissibly advance religion, then the Scarsdale creche should not 143. Id. at 720. The park had been used to display manger scenes from 1956 until 1980. Traditionally, the display had consisted of nine carved wooden figures. Id. In 1981 and 1982 the Village denied the Scarsdale Creche Committee permission to place a creche in the park. Id. at 720-21. 144. 588 F. Supp. 1337 (E.D. Mich. 1984). 145. Id. at 1338. 146. Id. at 1339-41. The district court concluded that, because Lynch did not involve a nativity scene standing alone, the Supreme Court's analysis in Lynch was inapplicable to the Michigan creche. In Birmingham, the district court strictly applied each prong of the Lemon test. d at 1339-41. The court ruled that the city had no secular purposes for displaying the manger scene. Id. at 1339. Without using a continuum, the court held that the effect of the display was to advance, affirm, and validate the Christian religion. Id. Although the court found no administrative entanglement, it concluded that the exhibit might cause political divisiveness. Id at 1339-40. 147. McCreary v. Stone, 739 F.2d 716, 723-30 (2d Cir. 1984), aff'd mere. by an equally divided Court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985). The district court focused on the Supreme Court's use of the tripartite test in Lynch and the requirement, set forth in Widmar v. Vincent, 454 U.S. 2634 (1981), that religious groups have equal access to public forums. Id. at 273. 148. McCreary v. Stone, 739 F.2d 716, 725 (2d Cir. 1984), aff'd mem. by an equally divided Court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985). The court in McCreary followed the reasoning of Widmar v. Vincent, 454 U.S. 263 (1981) (providing all religious groups equal access to public forums), and hypothesized that allowing equal access to a religious display could constitute an acceptable purpose. McCreary v. Stone, 739 F.2d 716, 725 (2d Cir. 1984), aff'dmem. by an equally divided Court sub non,. Board ofTrustees v. McCreary, 105 S. Ct. 1859 (1985). It is unclear why the court referred to an acceptable hypothetical purpose rather than an unconstitutional hypothetical purpose. For example, the court in .fcCreany could have ruled that the motivation behind the creche display was a desire to promote Christianity. 500 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 be treated differently.' 49 Thus, the Circuit Court added another effect to the establishment clause continuum, without any supporting analysis. Finally, the court found no administrative entanglement1 50 and refused even to consider the issue of political entanglement. 5 ' 2. The courts' religious symbol analyses The two courts also disagreed on how a religious symbol becomes secularized. The Second Circuit concluded that Lynch clearly required religious figures to be viewed in their seasonal, rather than physical, context. 152 The Michigan district court, in contrast, ruled that the Lynch decision required the courts to examine the nature of the physical environment surrounding a creche display.' 53 Thus, the district court ruled that, absent the plastic figures, there was nothing to offset the purely religious nature of the Birmingham 154 display. As the divergent McCreary and Birmingham decisions indicate, the impact of Lynch on the traditional tripartite test is unclear. 155 If courts limit the Lynch interpretation of the Lemon doctrine to nativity scenes surrounded by seasonal figures, then the traditional tripartite 149. McCreary v. Stone, 739 F.2d 716, 725-26 (2d Cir. 1984), aff'd mem. by an equally di- vided court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985). 150. Id. at 725. 151. Id. at 726. The district court concluded that Lynch permitted an analysis of political divisiveness only when direct subsidies are given to religious institutions. Id. See supra note 134 (noting Lynch dicta on political entanglement). 152. McCreary v. Stone, 739 F.2d 716, 729 (2d Cir. 1984), a/f'd mem. by an equally divided Court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985). The Second Circuit also noted, in dicta, that the decorations in the Village, and the fact that the park had recently been the site of Christmas caroling, minimized any remaining vestiges of religiosity in the creche display. Id. at 729. 153. American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1340 (E.D. Mich. 1984). 154. Id. at 1339. 155. Compare McCreary v. Stone, 739 F.2d 716, 725-30 (2d Cir. 1984) (relying on Lynch analysis, rather than Lemon analysis), aff'd mew. by an equally divided Court sub now. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985) with American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 233-38 (S.D. Tex. 1984) (applying three different tests). Despite the Second Circuit's reliance on the Lynch tripartite analysis in McCreary, the Supreme Court's application of Lemon in Lynch may have little precedential value. First, the Court's use of the tripartite test in Lynch may only be dicta. The Court relied on both history and the context of the display to reach its conclusion. Lynch v. Donnelly, 465 U.S. 668, 673-80 (1984). The Court only used Lemon after concluding that it was not bound by the tripartite test. Id. at 679. Moreover, the Court found that Lemon is only "useful" in establishment clause analysis. Id Second, Justice O'Connor's concurrence in the Court's opinion does not mirror the Court's use of the tripartite test. For example, Justice O'Connor argued that the mere presence of some secular purposes does not automatically satisfy the purpose prong of the Lemon test, Id. at 691 (O'Connor, J., concurring). Moreover, Justice O'Connor did not judge the religious effect on a continuum, but concluded that an effect is impermissible if it communicates a message of government endorsement or disapproval of religion. Id. (O'Connor, J., concurring). Consequently, the Second Circuit's analysis in McCremy may be based on a plurality decision and, therefore, may be of limited precedential value. 1986] OF CROSSES AND CRECHES 501 approach still may be viable.' 56 If, however, as McCreary indicates, the loose Lynch analysis of the establishment clause is applicable equally to any religious symbol case, the Lynch decision may have a dramatic impact. B. Other Symbols The confusion engendered by Lynch on both establishment clause doctrine and the definition of religious symbols is also evident in 57 cases involving other "religious" symbols. In Fausto v. Diamond,1 the United States District Court for the District of Rhode Island was confronted with a memorial to an unknown child, replete with a plaque quoting Deuteronomy and depicting a mother and child.' 58 Similarly, the United States District Court for the Southern District of Texas, in American Civil Liberties Union v. Eckels, 159 had to decide the constitutionality of two crosses and a Star of David placed in a public park. 160 Interestingly, five weeks after the Eckels suit was 161 filed, the park was designated as a war memorial. 1. The establishment clause analyses applied by the courts In Fausto, the district court summarized the effect of Lynch on establishment clause doctrine,' 6 2 and then noted that the outlines of the establishment clause "have been scumbled once again."' 163 Applying Lynch's version of the tripartite test, the court upheld the constitutionality of the Rhode Island monument to the unknown child. 64 The court determined that the government's participation in the project had a secular purpose' 65 and was not excessively en156. See supra note 66 (discussing Court's'recent use of traditional doctrine in parochial school cases). 157. 589 F. Supp. 451 (D.R.I. 1984). 158. I at 462. The monument was a fountain renovated by church groups and located on government property. The district court gave a lengthy description of the religious nature of the proceedings involved in the renovation, but failed to decide whether the fountain itself was or was not a religious symbol. Id at 460-64. 159. 589 F. Supp. 222 (S.D. Tex. 1984). 160. Id at 234-35. The court noted that religious symbols are a powerful medium for communicating religious messages. Id 161. Id at 225. 162. Fausto v. Diamond, 589 F. Supp. 451, 465-67 (D.R.I. 1984). The court highlighted the Supreme Court's recognition, in Lynch, of church-state coexistence and the limitations that the court imposed on traditional doctrine. Id 163. Id at 469. 164. Id at 467-69. The district court interpreted Lynch as permitting satisfaction of the purpose prong of the Lemon test if the government's secular aims were legitimate. Id. at 467. The court further maintained that Lemon was of limited value in assessing religious effect in light of the new "objective standard" adopted by Lynch. Id at 468. Finally, the court noted that the definition of entanglement, a "question of kind and degree," now required a greater burden of proof than set forth in Lemon. Id at 468-69. 165. Id. at 467. The court recognized that, although there might be religious motivations THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 tangling.' 6 6 The court measured the religious effect of the monu- ment on a continuum, concluding that its religious effects were no worse than those found acceptable in Lynch. 167 Unlike the Supreme Court's approach in Lynch, however, the district court did not support its conclusion with a recitation of history. 168 In contrast, the district court in Eckels held unconstitutional the public display of crosses and a Star of David. 169 In reaching this conclusion, the court applied three different establishment clause tests because of "the iandom approach of the Supreme Court to its analysis of the Establishment Clause." 170 First, the Eckels court applied the Lemon tripartite test. 17 1 It concluded that there was no secular purpose behind the display. 172 Without citing Lynch or using a continuum, the court also held that the effect of the symbols was primarily religious. 173 The court ruled that the entanglement was not excessive. 174 Thus, the court found the display violated the es75 tablishment clause under the first two prongs of Lemon. 1 The court next applied the historical test the Supreme Court developed in Marsh v. Chambers.176 After analyzing the historical significance of crosses and Stars of David in American society, the court connected to the erection of the monument, Lynch had "sounded the death knell of any notion that the government's objectives must be wholly secular." Id. 166. Id. at 468-69. Despite evidence of administrative entanglement, the court concluded that "such a spotty record" of interaction between church and state was not enough to establish unconstitutionality under the Lynch standard. Id. at 469. Furthermore, the political divisiveness in the community was not fatal "when hefted on the Lynch scales." Id. 167. Id. at 468. The court specifically measured the monument's religious effect against the effect of the creche in Lynch. The court ruled, without discussion or support, that the monument at bar was no greater aid to the establishment of religion than was the creche. Id. 168. The Supreme Court relied in large measure on an historical analysis to justify its decision in Lynch. See Lynch v. Donnelly, 465 U.S. 668, 673-80 (1984) (discussing historical nature of creche display). Other federal courts have neglected to employ an historical analysis tojustify their decisions regarding the constitutionality of religious displays. See McCreary v. Stone, 739 F.2d 716, 718-27 (2d Cir. 1984) (reaching decision on constitutionality of creche display without using historical analysis), aff'd mem. by an equally divided Court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1339-41 (E.D. Mich. 1984) (reaching decision regarding constitutionality of creche display without reference to historical analysis). 169. American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 233-38 (S.D. Tex. 1984). 170. Id. at 233. 171. Id.at 233-36. The district court did not adopt any of Lynch's alterations to the purpose and effect prongs. Id 172. Id. at 233-34. 173. Id. at 234-35. 174. Id. at 235-37. Although there was evidence of political entanglement, the court concluded that it was insufficient to invalidate the display under the Lynch standard. Id. The lack of entanglement was not dispositive, however, because failure to meet any prong of the Lemon test is constitutionally fatal. See supra note 40 and accompanying text (discussing need to survive each prong of analysis). 175. American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 237 (S.D. Tex. 1984). 176. Id. at 237 (applying test developed in Marsh v. Chambers, 463 U.S. 783 (1983)). See supra notes 68-71 and accompanying text (discussing Supreme Court's analysis in Marsh). 1986] OF CROSSES AND CRECHES 503 found no historical acceptance of such symbols when used to commemorate war dead in public parks. 1 77 The display, therefore, also failed historical review. Finally, the district court applied the strict scrutiny test used by the Supreme Court in Larson v. Valente. 17 8 The court decided that the state's interest in paying tribute to its war dead was not sufficiently compelling to meet the standards of strict scrutiny. 179 Moreover, the state's display only recognized Jewish and Christian soldiers and, therefore, was not narrowly tailored to the articulated goal. 180 Consequently, the court ruled that the display failed the last of the three establishment clause tests.1 8 ' 2. The courts' religious symbol analyses In Fausto, the district court was unable to conclude that the monument to an unknown child was secular.' 8 2 Rather, the court held that the memorial transmitted "conflicting signals," and therefore was ambiguous.' 8 3 Nevertheless, the monument passed constitutional review under the court's Lemon analysis.' 8 4 The district court, however, failed to view the monument in its physical and transactional context: it was financed in part through church mailers that surhad described abortion as morally objectionable, set in a square 85 rounded by church property, and commemorated by prayer.' In Eckels, the Texas district court determined that the two crosses and a Star of David placed in a public park qualified as religious symbols.' 8 6 The court noted that religious symbols are a powerful medium for communicating religious messages. 8 7 The court further concluded that the placement of the religious symbols in an 177. American Civil Liberties Union v. Eckels,, 589 F. Supp. 222, 237 ( S.D. Tex. 1984). The district court maintained that to survive historical analysis, the public use of religious symbols must be interwoven "into the fabric ofsociety." Id. (quoting Marsh v. Chambers, 463 U.S. 783, 792 (1983)). 178. Id. at 237-38 (applying strict scrutiny developed in Larson v. Valente, 456 U.S. 228 (1982)). See supra notes 72-77 and accompanying text (discussing Supreme Court's analysis in Larson). 179. Id. at 238. 180. Id. A more narrowly tailored display would have recognized the war dead from all religions. Id. 181. 182. Id. Fausto v. Diamond, 589 F. Supp. 451 (D.R.I. 1984). 183. Id. at 462. 184. Id. at 470. See supra notes 162-68 and accompanying text (discussing courts' use of establishment clause analysis). 185. Id. at 455-56. The court was concerned because both the planning for and voting on the monument had been divided along religious lines. Id. at 463. The court nevertheless upheld the monument's display. Id. at 470. 186. American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 234 (S.D. Tex. 1984). 187. Id. at 234-35. 504 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 otherwise secular environment-designating them as a war memorial-did not dilute their religious nature.1 8 8 V. A. CRITICAL ANALYSIS The Supreme Court's Vascillating Establishment Clause Analyses The application of establishment clause doctrine to allegedly religious displays has been marked by a lack of uniformity. Absent a uniform analytical approach, courts have reached inconsistent results and provided little guidance to communities seeking to avoid establishment clause problems. The Supreme Court's use of numerous standards in establishment clause cases has contributed, in large part, to the confusion among. the lower courts. The Court has failed to make clear the circumstances under which an historical analysis,' 8 9 strict scrutiny, 90 the tripartite review, 19 ' or any combination of these tests,1 92 must be applied. Furthermore, the parameters of each of these analyses are not clear. If a court opts to conduct an historical review,' 93 it may select from among the three different historical approaches the Supreme Court has embraced within the past forty years. A court may choose to either analyze the history of religious persecution in America and abroad, 9 4 highlight instances of church-state coexistence, 195 or examine the history of the specific practice at issue.' 96 Depending on which historical perspective is chosen, two courts may reach con188. Id The court concluded that a religious means cannot be used to effectuate a secular purpose. Id. But see Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1016, 558 P.2d 338, 347 (1976) (cross, although religious, when included as part of war memorial does not violate establishment clause), cert. deniedsub nom. Lowe v. Eugene Sand & Gravel, Inc., 434 U.S. 876 (1977). 189. See Marsh v. Chambers, 463 U.S. 783, 785-96 (1983) (using historical analysis as establishment clause doctrine). 190. See Larson v. Valente, 456 U.S. 228, 244-46 (1982) (employing strict scrutiny to invalidate state statute under establishment clause). 191. See Lemon v. Kurtzman, 403 U.S. 602, 612-13 (1971) (consolidating three-part test for application in establishment clause cases). 192. See Lynch v. Donnelly, 465 U.S. 668, 678-85 (1984) (employing historical analysis and modified tripartite test); Widmar v. Vincent, 454 U.S. 263, 271-77 (1981) (using strict scrutiny analysis and Lemon test); American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 233-38 (S.D. Tex. 1984) (applying historical analysis, strict scrutiny review, and Lemon test). 193. Marsh v. Chambers, 463 U.S. 783, 785-96 (1983). In Marsh, ChiefJustice Burger, who also wrote the majority opinions in Lemon and Lynch, employed an historical analysis without adequately distinguishing earlier cases that had applied the Lemon tripartite test. Id. Consequently, it is unclear when courts should conduct an historical, rather than a tripartite, review. 194. See Everson v. Board of Educ., 330 U.S. 1, 8-12 (1947) (discussing persecution of religious groups in Great Britain and in United States and early movements toward reform). 195. See Lynch v. Donnelly, 465 U.S. 668, 674-78 (1984) (focusing on instances of official recognition of religion in American society). 196. See Marsh v. Chambers, 463 U.S. 783, 786-92 (1983) (detailing history of legislative religious invocations). 1986] OF CROSSES AND CRECHES 505 flicting results on the constitutionality of identical displays. Thus, an historical analysis approach may prove too malleable to be of value as a constitutional standard. 19 7 Moreover, the Court has failed to explain why the history of a practice has constitutional relevance. The Supreme Court has not yet applied strict scrutiny to religious symbol cases. 198 The Court has ruled, however, that a state action that denies religious speech must withstand strict scrutiny. 199 If religious symbols are forms of religious speech, 20 0 then strict scrutiny may bar local governments from prohibiting religious displays on public property. 20 ' If a state violates the first amendment by denying private groups the opportunity to display religious symbols, it would appear contradictory to then argue that the state also violates the first amendment by exhibiting the displays. Carried to its logical extension, this strict scrutiny analysis would compel governmental authorities to allow the public display of any religious symbols, notwithstanding the policies and precedents behind the establishment clause. 20 2 Moreover, such a compelled approach might well convey the impression that the government, in direct violation of 197. Because the study of history itself is not a "science" and is subject to the perspective of the historian, a court may select the view of history it finds most accomodating to its position. See R. COLLINGWOOD, THE IDEA OF HISTORY 218 (1946) (contending that usefulness of historical study is limited by historian's ability to "re-think" history); H. COMMAGER, THE STUDY OF HISTORY 53 (1966) (arguing that historical analysis necessarily is influenced by biases of historian); A. ROwSE, THE USE OF HISTORY 95-104 (new rev. ed. 1963) (discussing vulnerability of history to distortion in order to serve specific ends). See generally P. KAUPER, RELIGION AND THE CONSTITrrTION 47-53 (1964) (discussing limited effectiveness of historical analysis in construing religion clauses). 198. Lynch v. Donnelly, 465 U.S. 668, 687 n.13 (1984). Although the Court recognized the applicability of the strict scrutiny test to practices that are discriminatory on their face, the Court refused to apply the test to the public display of a creche. Id 199. See Widmar v. Vincent, 454 U.S. 263, 269-70 (1981) (establishing that state action involving discrimination against religious speech must withstand strict scrutiny test). 200. The Supreme Court has held that displays of symbols are protected forms of expression under the first amendment. See Clark v. Community for Creative Non-Violence, 468 U.S. 288 (1984) (stating that symbolic expression is protected speech under first amendment); see also Spence v. Washington, 418 U.S. 405, 415 (1974) (per curiam) (ruling that first amendment protects display of American flag with upside down peace symbol); Tinker v. Des Moines Community School Dist., 393 U.S. 503, 506 (1969) (holding that first amendment protects symbolic display of armband). Religious objects communicate ideas in the same manner as other symbols, and therefore the argument can be made that religious objects should receive first amendment protection. See supra notes 1-2 (discussing communicative properties of symbols). 201. See Widmar v. Vincent, 454 U.S. 263, 269-70 (1981) (holding that state's desire to avoid establishment clause problems was insufficient to justify abrogation of first amendment rights of speech and association). Government actions aimed at avoiding establishment clause problems are justified only on a showing that the action is necessary to serve a compelling state interest and is narrowly drawn to achieve that end. Carey v. Brown, 447 U.S. 455, 461, 464-65 (1980). 202. See supra note 23 and accompanying text (discussing policies behind establishment clause). 506 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 the establishment clause, supports the religions whose symbols are displayed.203 The Supreme Court's loose interpretation of the Lemon tripartite test 20 4 has contributed to the conflicting analyses and results among the courts. 20 5 It is unclear why religiously motivated governmental actions could ever withstand constitutional review. 20 6 Even a clearly religious governmental action is likely to have at least one secular purpose supporting it.207 Such governmental actions, however, lack the neutrality demanded by the establishment clause. 20 8 A government that chooses, for example, to display a cross supports religion over nonreligion, 20 9 and one religion over others.2 1 0 Merely because Americans "are a religious people" 2 11 does not justify governmental actions that favor the interests of a majority religion over those of minority faiths. 2 12 Religiously motivated state action, even when justified by a secular purpose, causes excessive interaction be203. See supra note 4 and accompanying text (discussing religious prohibitions under es- tablishment clause). 204. See Lynch v. Donnelly, 465 U.S. 668, 679-85 (1984) (loosely applying Lemon test and declaring unwillingness to be found by any single test). 205. See American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 233-38 (S.D. Tex. 1984) (applying various establishment clause tests); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1339-40 (E.D. Mich. 1984) (distinguishing Supreme Court discussion in Lynch and applying Lemon tripartite test). 206. One obvious explanation for permitting religiously motivated actions to survive constitutional scrutiny is that the Court's commitment to the Lemon test's requirement of a primarily secular purpose is, at best, superficial. See Lynch v. Donnelly, 465 U.S. 668, 696 (1984) (Brennan, J., dissenting). In its decision in Lynch, the Court offered no justification for permitting state actions that, although containing a secular purpose, were predominantly motivated by religious inclinations. Id. at 681 n.6. 207. See, e.g., Fausto v. Diamond, 389 F. Supp. 451, 460-64 (D.R.I. 1984) (despite highly religious nature of monument's renovation, court searched for and found sufficiently secular purpose to validate state involvement in renovation process); Meyer v. Oklahoma City, 496 P.2d 789, 792 (Okla. 1972) (finding valid secular purpose to support placement of enormous cross on public fairgrounds). Because governments in the United States are secular institutions, it seems likely that their actions will be at least marginally motivated by secular interests. 208. See supra note 33 and accompanying text (discussing mandates of establishment clause). 209. See Everson v. Board of Educ., 330 U.S. 1, 18 (1947) (stating that first amendment requires government to remain neutral in its relations between groups of religious believers and nonbelievers). 210. Id at 15 (concluding that first amendment prohibits government from aiding one religion, all religions, or preferring one religion over another). 211. Zorach v. Clauson, 343 U.S. 306, 313 (1952). 212. See School District of Abington Township v. Schempp, 374 U.S. 203, 214-15 (1963) (noting that America is religiously diverse, and that Constitution forbids preferential treatment among religions); Everson v. Board of Educ., 330 U.S. 1, 26-27 (1947) (Jackson, J., dissenting) (stating that first amendment was designed to excise religious controversy from public life by denying to any denomination any advantage over public policy or public purse); Gilfillan v. Philadelphia, 637 F.2d 924, 930 (3d Cir. 1980) (concluding that essential purpose of establishment clause is protection of religious pluralism), cert. denied, 451 U.S. 987 (1981); May v. Cooperman, 572 F. Supp. 1561, 1575 (D.NJ. 1983) (noting that first amendment prohibits governmental program that advances religion of some and inhibits religion of others). 1986] OF CROSSES AND CRECHES 507 2 13 tween church and state. Courts also have failed to articulate a workable standard for determining when a religious effect, the second prong of Lemon, is primary rather than incidental. 2 14 In applying the Lemon effect prong, courts have not accounted explicitly for the perspective of religious minorities in judging the effect of governmental aid to religion. 2 15 Consequently, courts have upheld governmental aid to the majority religion, viewed as preferential and16 oppressive by religious minori2 ties, with little analytical support. The Supreme Court's decision in Lynch compels an analysis of religious effects on a continuum. 21 7 This approach, however, does not account for the dual nature, and dual effects, of religious symbols. 2 18 A creche, for example, is both a significant religious symbol and a reminder of an historical event. 21 9 Yet, the Supreme Court has not examined this dual nature distinction. Before judging the creche's effect on a continuum, the majority in Lynch decided that the display was secular. 220 By making this initial, ad hoc determination, the Court in Lynch predetermined the ultimate result of its ef213. See generally 16 THE WRrITNGS OF THOMAS JEFFERSON 281 (Library ed. 1903) (arguing that church and state ought not to interact). If a government decision is religiously motivated, there is no "wall" between church and state. See supra note 24 (discussing "wall" of separation concept). 214. The Supreme Court further clouded application of the effect prong with its use of a continuum analysis in Lynch. See Lynch v. Donnelly, 465 U.S. 668, 681-83 (1984) (measuring effect of creche display on continuum). 215. See infra note 255 (discussing perspective of religious minority regarding effect of religious displays). 216. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 681-83 (1984) (conceding difficulty of measuring degree of benefits accruing from governmental programs and proceeding to conclude, without analysis, that Pawtucket creche display was no more intrusive than practices previously found to have acceptable religious effects); Fausto v. Diamond, 589 F. Supp. 451, 468 (D.R.I. 1984) (concluding, without analysis, that monument's religious effect was inconsequential); Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1016, 558 P.2d 338, 347 (1976) (undertaking superficial analysis of religious effect to rule that cross in war memorial loses religious effect), cert. denied sub nom. Lowe v. Eugene Sand & Gravel, Inc., 434 U.S. 876 (1977). See generally Van Alstyne, supra note 23 (arguing that christianization of American government is oppressive to religious minorities). 217. See Lynch v. Donnelly, 465 U.S. 668, 681-83 (1984) (using continuum to analyze effect of creche display). 218. See id. at 693 (Brennan, J., dissenting) (recognizing that religious symbols may have secular and religious effects); see also Fausto v. Diamond, 589 F. Supp. 451, 467-68 (D.R.I. 1984) (focusing on secular aspect of publicly maintained monument); American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 233-34 (S.D. Tex. 1984) (focusing on religious aspect of war memorial); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1339 (E.D. Mich. 1984) (centering analysis on religious aspect of creche after rejecting attempts to focus review on secular aspect of creche). The Supreme Court's reliance in Lynch on the presence of religious art in federal buildings and galleries to justify the secular nature of a creche epitomizes the dual nature of religious objects. Lynch v. Donnelly, 465 U.S. 668, 676-77 (1984). The Court referred to the dual nature of religious art to support its inquiry into the secular aspect of the creche display. Id. 219. Lynch v. Donnelly, 465 U.S. 668, 680 (1984) (recognizing dual nature of creche). 220. Id. at 683. 508 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 fect analysis. Applying this analysis, a court may place the secular aspect of any display on a continuum with other accepted government practices, and thereby circumvent a consideration of the reli- gious impact of the state action. As each new permissible effect is added to the continuum, however, the less force the effect prong of the tripartite test will have in establishment clause cases. Moreover, the very practice of applying a continuum to religion cases may be flawed analytically. The effect of a religious display in one community might not lend itself to comparison with religious aid in another 221 community. Although the Supreme Court has not altered substantially the ,traditional administrative entanglement test,222 it has not vigorously applied this portion of the Lemon tripartite analysis in the religious symbols context. 223 When governments use their administrative processes to allocate money or property for religious displays, the aid benefits institutions whose character and purpose 224 are largely religious. Consequently, the nature of such assistance 225 is public support for a symbolic religious display. Although the resultant church-state relations 2 26 may be relatively minor, 227 the ensuing ad221. See Lynch v. Donnelly, 465 U.S. 668, 681 (1984) (noting difficulty of comparing ben- efits conferred on religion by different governmental support). Justice Brennan pointed out in dissent that calm acceptance of a symbol in one community does not guarantee the same effect in another community. Id. at 703-04 (Brennan, J., dissenting). 222. Compare Lemon v. Kurtzman, 403 U.S. 602, 620 (1971) (concluding that administrative entanglement involves financial or administrative relationship between church and state) with Lynch v. Donnelly, 465 U.S. 668, 683 (1984) (accepting district court's application of Lemon test with respect to findings on administrative entanglement). 223. See Lynch v. Donnelly, 465 U.S. 668, 684 (1984) (stating that court need not thoroughly analyze entanglement issue); see also Fausto v. Diamond, 589 F. Supp. 451, 468-69 (D.R.I. 1984) (noting that despite strong showing of administrative entanglement, in absence of proof of ongoing, daily interaction between church and state, Supreme Court decisions require holding that entanglement is not excessive); Citizens Concerned for Separation of Church and State v. City of Denver, 526 F. Supp. 1310, 1315 (D. Colo. 1981) (finding no entanglement with secular administration of religious display when religious officials not involved); cf. Walz v. Tax Comm'n, 397 U.S. 664, 674 (1970) (failing to find entanglement where government granted tax exemptions to religious institutions and thereby initiated involvement between church and state). In analyzing specific governmental practices, courts rely on the proposition that "entanglement is a question of kind and degree." Lynch v. Donnelly, 465 U.S. 668, 684 (1984). See McCreary v. Stone, 739 F.2d 716, 725 (2d Cir. 1984) (noting that entanglement is question of kind and degree), aff'd mem. by an equally divided Court sub nom. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985). 224. See Lemon v. Kurtzman, 403 U.S. 602, 615 (1971) (establishing character and purpose as criteria for judging excessive entanglement). 225. Id. (establishing nature of aid as criterion to be considered in entanglement analysis). 226. IL (requiring courts to look to resulting relationship between church and state as part of entanglement analysis). 227. See Lynch v. Donnelly, 465 U.S. 668, 671 (1984) (listing nominal expenses connected with publicly-supported creche display). But cf. Gilfillan v. City of Philadelphia, 637 F.2d 924, 927 (3d Cir. 1980) (describing larger expenditures of government money in connection with Papal visit to Philadelphia), cert. denied, 451 U.S. 987 (1981). 1986] OF CROSSES AND CRECHES 509 ministrative relationship is symbolic of government support for religion. Complimenting its treatment of the administrative entanglement test, the Supreme Court essentially has discarded the political entanglement portion of the final prong in the Lemon tripartite analysis. 2 28 In so doing, the Court has discounted the political divisiveness and religious factionalism caused by displays of religious symbols on public property. 229 Of course, courts should not invalidate a state action simply because certain religious groups oppose or support a governmental practice. Courts should recognize, however, that when a state action stirs political debate along religious lines, the action's religious effect is more than incidental. In the Lynch case, for example, the district court cautioned that the traditionally quiet acceptance of the Pawtucket creche may have masked an unhealthy malaise within that Rhode Island community. 23 0 Thus, by emasculating the political entanglement analysis, the Supreme Court has neglected an essential policy behind the first amendment, 23 ' and eschewed an accurate measure of the religious impact of a state action. B. The Supreme Court's Failureto Define "Religious Symbol" The Supreme Court's failure to provide an analytical framework for determining whether a symbol is religious has contributed significantly to the current doctrinal conflict among the states and circuits. 2 32 A court's explicit or implicit 23 3 determination of whether a case involves a "religious symbol" influences both the type and the 228. See Lynch v. Donnelly, 465 U.S. 668, 684 (1984) (declining to hold that political divisiveness alone can invalidate otherwise permissible conduct). 229. See supra note 21 (highlighting controversy caused by religious displays). The volability connected to public perceptions of governmental endorsement of religion is partly a function of the heterogeneity of contemporary society. See School Dist. of Abington Township v. Schempp, 374 U.S. 203, 228 (1963) (Brennan,J., concurring) (observing that changes in makeup of society may contribute to antagonism toward practices previously deemed acceptable). 230. See Donnelly v. Lynch, 525 F. Supp. 1150, 1170 (D.R.I. 1981) (noting that silence of minority religions may have been "an unhealthy indication of the fear of angering the predominantly Christian majority"), af'd, 691 F.2d 1029 (1st Cir. 1982), rev'd, 465 U.S. 668 (1984). 231. See Lemon v. Kurtzman, 403 U.S. 602, 622 (1971) (observing that Framers wrote first amendment to avoid political division along religious lines). 232. See, e.g., McCreary v. Stone, 739 F.2d 716, 729 (2d Cir. 1984) (viewing creche in seasonal, as distinct from physical, context), aff'd mem. by an equally divided Court sub non. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985); American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 234 (S.D. Tex. 1984) (ruling that physical context does not alter religiosity of religious symbols erected as part of war memorial). 233. See supra note 88 and accompanying text (discussing courts' assumptions with respect to the nature of religious symbols). THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 scope of establishment clause analysis it applies.23 4 If a court determines, for example, that an object is secular, it likely will find that the object survives constitutional scrutiny, regardless of the type of 2 35 establishment clause analysis applied. Although the Supreme Court's ruling in Lynch compels an analysis of religious objects "in context," this standard has been of little assistance to courts struggling to define "religious symbol." The Court in Lynch failed to delineate the specific parameters of this "in context" review. 23 6 If the Supreme Court intends the lower courts to evaluate the religious nature of objects by viewing them in relation to their immediate surroundings, then a creche standing alone would be a religious symbol, but a creche surrounded by two plastic reindeer would be secular.2 3 7 Moreover, Lynch addressed only the issue of when admittedly religious objects become secularized by their context. The Lynch standard fails, however, to account for symbols that are not admittedly religious. Thus, local communities 234. See Lynch v. Donnelly, 465 U.S. 668, 680-85 (1984) (based on determination that creche was secular, Court applied limited version ofLemon); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1339 (E.D. Mich. 1984) (court's determination that creche was religious compelled conclusion that display failed tripartite review). Compare American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 234-38 (S.D. Tex. 1984) (declaring that conclusion that crosses and Stars of David are religious symbols influences holding that display is unconstitutional) with Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1010, 558 P.2d 338, 340 (1976) (declaring that conclusion that cross is secular mandates holding that display passes tripartite review), cert. denied, 434 U.S. 876 (1977). 235. See generally Lynch v. Donnelly, 465 U.S. 668, 680-85 (1984) (determining that, viewed in context, creche is secular and, therefore, passes historical and Lemon reviews); Anderson v. Salt Lake City Corp., 475 F.2d 29, 31-34 (10th Cir. 1973) (holding that Ten Commandments are historical foundation of modem law and, therefore, pass traditional Lemon review); Citizens Concerned for Separation of Church and State v. City of Denver, 526 F. Supp. 1310, 1311-15 (D. Colo. 1981) (holding that, viewed in context, creche is only nominally religious and, therefore, display passes Lemon test); Meyer v. Oklahoma, 496 P.2d 789, 791-93 (Okla. 1972) (determining that cross, in commercial context, was secular and, therefore, withstands scrutiny under state constitution). Conversely, ifa court finds an object to be patently religious, even in context, it is unlikely to hold the display constitutional. See, e.g., Stone v. Graham, 449 U.S. 39, 41-43 (1980) (per curiam) (holding that Ten Commandments are distinctly religious and, thus, cannot survive Lemon test); American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 233-38 (S.D. Tex. 1984) (finding cross and Star of David plainly religious symbols which, therefore, fail constitutional review). 236. Lynch v. Donnelly, 465 U.S. 668, 680 (1984). The Court referred to the broader context of the Christmas season. Id. Courts and commentators, however, have interpreted the Court's analysis to require review of displays in their immediate physical surroundings. See American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1339 (E.D. Mich. 1984) (basing decision on religious nature of creche by reference to absencc of other seasonal figures); see also Lynch v. Donnelly, 465 U.S. 668, 705-09 (1984) (Brennan, J., dissenting) (viewing majority's use of context as meaning physical surroundings); ANTi-DEFANiATION LEAGUE, LAW REPORT LiTIGATION DOCKET: 1984 17 n.5 (Spring 1984) (noting that Supreme Court decision in Lynch left unclear whether courts should review displays in their physical or seasonal context). 237. But see Donnelly v. Lynch, 525 F. Supp. 1150, 1169 (D.R.I. 1981) (observing that government cannot obscure its motives by surrounding creche with nonreligious objects), aft'd, 691 F.2d 1029 (1st Cir. 1982), rev'd, 465 U.S. 668 (1984). 1986] OF CROSSES AND CRECHES lack any guidelines for deciding whether the objects they wish to 2 38 display constitute "religious symbols." No matter how "religious symbol" ultimately is defined, the Lynch analysis allows courts to construe "context" so broadly as to allow for the display of virtually any religious object. 23 9 An example of this result is illustrative. Many Christians believe that Jesus Christ was an embodiment of God. 240 Congress, however, chose to recognize Christmas, the birth of Jesus Christ, as a secular national holiday. 24 1 Consequently, symbols of Christ's embodiment on earth could arguably be "secularized" by reference to the national secular holiday.2 42 Aside from unsupported judicial assertions, however, there is no support for the proposition that "context" reduces the religious nature of a symbol.2 4 3 On the contrary, the decisions of 238. It is unlikely that any government would display a religious symbol, knowing that it violated the establishment clause. Thus, cases arising in this area most likely involve governments that chose to display arguably religious symbols without knowing whether their displays violated the establishment clause. Compare Hall v. Bradshaw, 630 F.2d 1018, 1023 (4th Cir. 1980) (declaring prayer printed on back of state-sponsored map unconstitutional), cert. denied, 450 U.S. 965 (1981) with State ex rel Singleman v. Morrison, 57 So. 2d 238, 243-47 (La. Ct. App. 1952) (ruling that statue of Catholic saint erected on public property survives review under state constitution). See supra notes 106-09 and accompanying text (illustrating confusion among courts seeking to apply establishment clause). 239. See Meyer v. Oklahoma City, 496 P.2d 789, 792-93 (Okla. 1972) (proposing that commercial context obscures religious nature of 50 feet tall cross). 240. SeeJohn 1:1-18 (relating Christ's relationship to God and man); Colossians 2:9 (discussing God's presence in person of Christ). Conversely, an essential belief of many non-Christians is that Christ was not an embodiment of God. See Lynch v. Donnelly, 465 U.S. 668, 708 n.14 (1984) (Brennan, J., dissenting) (citing Martin Buber's explanation that, for Jews, God does not take human form). 241. 5 U.S.C. § 6103 (1982) (listing Christmas among national legal holidays). 242. Courts have not addressed the issue of whether association with other legal holidays, aside from Christmas, may secularize an otherwise religious object. Consequently, only Christian symbols could be designated as secular under a "secularized holiday" analysis, because only Christmas, a holiday of the majority religion, has received official governmental recognition. See id. (Christmas only religious national holiday). Thus, only Christian symbols can receive the "benefit" of being displayed without violating the establishment clause. Based on this analysis a cross, crucifix, or other icon could merely reflect the historical times of Christ, the basis for the secular, national holiday. Justice Brennan pointed out the faulty logic of the majority's syllogism: 1. Christmas is an acceptable national holiday; 2. The creche is a traditional element of Christmas; 3. Therefore the creche is acceptable under the establishment clause. Lynch v. Donnelly, 465 U.S. 668, 709 (1984) (Brennan, J., dissenting). Justice Brennan argued that the Court erred because the only constitutionally acceptable aspect of the nationally recognized Christmas holiday is the secular celebration. The holiday merely accomodates people who wish to spend a religious day with their families. The creche, however, serves only the religious aspect of the holiday, and is, therefore, impermissible. Id. 243. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 679-80 (1984) (asserting, without any explanation, that religious nature of creche depends on its display context); McCreary v. Stone, 739 F.2d 716, 729 (2d Cir. 1984) (concluding that creche standing alone is not religious because creche in Lynch was found not religious), aff'd mem. by an equally divided Court sub non. Board of Trustees v. McCreary, 105 S. Ct. 1859 (1985); see also Comment, supra note 46, at 353 (arguing that presence of Santa figure does not diminish religious nature of display); R. Teitel, The Impact of Lynch: The New Nativity Scene 12 (Sept. 1984) (unpublished paper THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 the Supreme Court demonstrate that religious practices in secular 2 44 settings violate the establishment clause. The Supreme Court has concluded that the American people would not accept a complete bar against all official recognitions of religion. 24 5 Consequently, courts have upheld governmental displays of religious symbols and, therefore, have failed to account for the essential policies behind the establishment clause. The cases finding such displays constitutional have involved almost exclusively Christian symbols.2 4 6 This preferential treatment among religions violates the spirit of neutrality that the first amendment was designed to preserve.2 4 7 None of the standards that the courts 248 utilizes considers the perspective of religious minorities. Although Christians may view an object as secular or neutral, Jews, Moslems, or Hindus may view the same object as oppressive and supportive of Christianity.2 4 9 Moreover, if the object is indeed a delivered at Conference on Church and State Separation at B'Nai B'rith, Washington, D.C., Nov. 4, 1984) (observing that Court muddled analysis of "context"). The majority in Lynch dismissed justice Brennan's charge that a secular context cannot diminish a creche's religious import with the statement "of course this is not true." Lynch v. Donnelly, 465 U.S. 668, 685 n.12 (1984). 244. See, e.g., Stone v. Graham, 449 U.S. 39,42 (1980) (per curiam) (declaring that posting Ten Commandments in secular school is unconstitutional); Meek v. Pittenger, 421 U.S. 349, 370-72 (1975) (ruling that providing instructional aids to sectarian school is unconstitutional); School Dist. of Abington Township v. Schempp, 374 U.S. 203, 223 (1963) (holding that bible reading and prayer recitation in secular school are unconstitutional activities). 245. Zorach v. CIauson, 343 U.S. 309, 312-13 (1952) (supporting proposition by reference to numerous situations in daily life that involve some interaction between church and state). 246. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 683 (1984) (creche); Meyer v. Oklahoma City, 496 P.2d 789, 790 (Okla. 1972) (cross). The only cases that have not involved exclusively religious symbols are Stone v. Graham, 449 U.S. 39, 39 (1980) (per curiam) (Ten Commandments); American Civil Liberties Union v. Eckels, 589 F. Supp. 222, 222 (S.D. Tex. 1984) (Star of David and crosses); Anderson v. Salt Lake City Corp., 475 F.2d 29, 30 (10th Cir. 1973) (Ten Commandments). The absence of cases involving non-Christian symbols could, however, indicate that there are few displays of such symbols. 247. See supra notes 31-33 and accompanying text (discussing policy behind first amendment); see also AMERICAN JEWISH CONGRESS, COMMISSION ON LAW AND SOCIAL ACTION REPORT 17 (Mar. 1984) (observing that Court's support of Christian symbols violates neutrality by creating impression that Christianity is state religion). 248. Although the courts do not explicitly state from whose perspective they judge the religiosity of a symbol, it is unlikely that the courts are using the minority's viewpoint. This may be inferred from the nature of symbols found acceptable by the courts. See Lynch v. Donnelly, 465 U.S. 668, 687 (1984) (upholding validity of creche display); Eugene Sand & Gravel, Inc. v. City of Eugene, 276 Or. 1007, 1016-19, 558 P.2d 338, 346-48 (1976) (upholding constitutionality of cross in war memorial), ceri. denied, 434 U.S. 876 (1977); Meyer v. Oklahoma City, 496 P.2d 789, 792 (Okla. 1972) (upholding validity of cross on public fairgrounds); see also supra notes 242, 246 (addressing limitations on secularization of religious symbols to symbols of majority religion). 249. See Lynch v. Donnelly, 465 U.S. 668, 701 (1984) (Brennan, J., dissenting) (contending that religious minority and atheists view Christian symbols differently from religious majority); Anti-Defamation League Press Release 13 (Mar. 5, 1984) (reporting that minority religions view creche as religious and Lynch decision as oppressive); see also infra note 255 (discussing perspective of religious minority toward creche displays). 1986] OF CROSSES AND CRECHES 513 "religious symbol," its display on public property may also tend to 250 cheapen its religious significance. VI. PROPOSED APPROACH Courts always could ensure religious neutrality by forbidding any governmental recognition of religion or a deity. 2 5 1 It is unlikely, however, that either the Supreme Court or the American people would ever accept such an absolutist approach to the first amendment. 2 52 Instead, the Supreme Court should adopt a two-part anal- ysis to determine whether an allegedly religious symbol can be placed on public property consistent with the Constitution: first, a court must decide whether the object to be displayed is religious; second, a court must determine whether the act of displaying the object violates the establishment clause. A. Definition of "Religious Symbol" In characterizing an object as a "religious symbol," courts should determine whether the object is religiously neutral. 25 3 In reaching this determination, courts should view the object from the perspective of religious minorities. Neutrality among religions would be ensured if a member of a minority religion would view the symbol displayed as secular or nondenominational. Under this standard, courts could be expected to view official references to God or holiday decorations as neutral symbols which could constitutionally be displayed on public property. In contrast, courts could also be expected to view crosses, creches, or Stars of David as sectarian, denominational symbols which could not be constitutionally displayed on public property. If an object is nonpreferential its display may 250. See Lynch v. Donnelly, 465 U.S. 668, 712 n.19 (1984) (Brennan, J., dissenting) (observing that many Christian commentators oppose creche displays as debasement of Christian beliefs). In his letter to Edward Livingston,James Madison wrote that religion only retains its purity by not mixing with the government. MADISON, supra note 23, at 309. Thus, creche displays present a paradox: if they are religious symbols, then their involvement with the state dilutes their religiosity; if they are secular symbols, then the miraculous nature of Christ's birth is trivialized. In either event, Christianity is not well served by publicly-supported creche displays. See R. Teitel, supra note 243, at 2. 251. -See generally Pfeffer, supra note 82, at 233 (noting that Framers opposed governmental invocation of deity). One commentator has explained that the Constitution makes no reference to God, and that Jefferson referred to God in the Declaration of Independence only because of political necessity. Id. at 217. See also id. at 238 (warning that references to deity are dangerous because they are often used to justify state meddling in religion). 252. See Lynch v. Donnelly, 465 U.S. 668, 674-78 (1984) (discussing role of religion in American life). 253. The possibility that this test would not isolate objects that advance the beliefs of all religions at the expense of secular interests should not discredit the test. This deficiency appears more theoretical than real as it is difficult to conceive of a governmental program that would deliver such a result. 514 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 nevertheless establish religion. Therefore courts should also apply a rigorous establishment clause analysis. B. Appropriate Establishment Clause Analysis Elimination of the ambiguous and contradictory results of establishment clause cases, especially in the area of religious symbols, requires a strict application of the Lemon test. As originally articulated by the Supreme Court, the Lemon tripartite test limits governmental support of religion over nonreligion or one religion over others. A strict application would provide a uniform method for resolving establishment clause issues raised by public displays of religious symbols as well as ambiguous or predominantly secular symbols.2 5 4 To ensure neutrality among religions, governmental actions must not be religiously motivated. Rather, state practices should have exclusively secular purposes. In order to guarantee that an effect is not primarily religious, courts must examine the relationship between the religious and secular effects of the governmental action. If a member of a religious minority would view the religious effect as more than incidental, the government practice must be barred. It is likely, for example, that ajew or Moslem would view Christian art in public galleries as producing only an incidental religious effect. Consequently, such a display would be constitutional under the effect prong. on the other hand, a Jew or Moslem would probably view a creche display as producing more than an incidentally religious effect. 25 5 254. It is conceivable, for example, that a court finding a violation of either the purpose prong or the entanglement prong of the tripartite test could find a secular object unconstitutional. Thus, although one court upheld the secular nature and constitutionality of a county seal embossed with the words "con esta vencemos" (with this we win), mountains, and a plain with grazing sheep, the court could have invalidated the seal on other grounds. SeeJohnson v. Board of County Comm'nrs, 528 F. Supp. 919, 920-25 (D.N.M. 1981), aff'd ub nora. Friedman v. Board of Comm'nrs, No. 82-1064, slip op. (10th Cir. Dec. 27, 1984). If the county had a religious purpose behind displaying the seal, even though the seal itself was secular, the display would violate the purpose prong. Based on the Supreme Court's ruling in Lynch, however, the court of appeals held that the district court need not search for religious purposes if the government offers a valid secular motivation. Friedman v. Board of Comm'nrs, No. 821064 (10th Cir. Dec. 27, 1984). Additionally, if church groups participated in the development of the logo, the county seal could violate the entanglement prong. This would be true even though the seal was not religious and had no religious effect. 255. Several authorities support the observation that a member of a minority religion would view a creche as religious. See Anti-Defamation League Press Release (Mar. 5, 198.1) (identifying creche as Christian symbol); cf. AMERICANJEWISH CONGRESS, supra note 247, at 24 (contending that Lynch decision allows government to support Christianity over other religions); R. Teitel, supra note 243, at 16 (suggesting that majority's decision in Lynch upholds Christian beliefs). One poll appears to indicate that a majority of society considers creche displays religious. Wash. Post, Nov. 30, 1984, at B5, col. 3 (80% of Metro riders surveyed in Washington, D.C. by Anti-Defamation League believed creche display was religious). Butfcf. CORNELISON, THE RELATION OF RELIGION TO CIVIL GOVERNMENT IN THE UNITED STATES OF 1986] OF CROSSES AND CRECHES 515 The effect of a symbolic display is likely to be more than incidentally religious if it generates political controversy along religious lines. 256 Courts should, therefore, incorporate the political entan- glement analysis into the effect prong of the Lemon test. If a creche display in a public park, for example, provokes emotional debate and lawsuits, courts may infer that the symbol has a primarily reli2 gious effect. 57 Courts should also apply the administrative entanglement analysis 258 more vigorously, focusing on the criteria established in Lemon. Although religious displays on public property often involve relatively minor amounts of money, 259 governments cannot, consistent with the establishment clause, expend public time and funds to exhibit sectarian symbols. CONCLUSION The first amendment precludes government from establishing religion. The Supreme Court developed the tripartite Lemon analysis to ensure that governmental actions and motivations remain separate from religion. Although the line separating church from state never has been clear, the Supreme Court's recent applications of establishment clause doctrine have further obscured this demarcation. The doctrinal analysis applied to publicly-sponsored religious displays has been especially confused and contradictory. Initially, most of the lower courts applied the traditional Lemon test. After the Supreme Court's ruling in Lynch v. Donnelly, however, the courts have applied varying tests with differing amounts of rigor. None of the courts has articulated an effective method for determining whether an object is a religious symbol. As a result, courts have upheld the constitutionality of publiclysponsored, religious displays despite the appearance of governmenAMERICA 331 (1895) (arguing that American government is Christian, and that "unbelievers" have no reason to complain). 256. See supra note 21 and accompanying text (highlighting instances of controversy caused by religious displays). 257. The question remains as to what level of emotional debate and litigation constitutes a threshold display of a primarily religious effect. See supra note 214 and accompanying text (noting courts' failure to articulate workable standard for determination of primary effect). 258. See supra notes 57-62 and accompanying text (outlining administrative entanglement elements). 259. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 671 (1984) (noting nominal costs associated with public maintenance of creche display); Fausto v. Diamond, 589 F. Supp. 451, 457 (D.R.I. 1984) (noting that dollar amount expended on monument was insubstantial); American Civil Liberties Union v. City of Birmingham, 588 F. Supp. 1337, 1338 (E.D. Mich. 1984) (observing that only small amount of public funds were expended on building, maintenance, and display of creche). 516 THE AMERICAN UNIVERSITY LAW REVIEW [Vol. 35:477 tal preference for one religion over another. These displays have spawned lawsuits and political divisions that have cut sharply along religious lines. Although judicial legitimation of religious displays on public property may not signal the demise of secular American government, it does signal closer ties between church and state. Such ties represent a weakening of the establishment clause. To counteract such a result, the courts must adopt a realistic and workable definition of what constitutes a "religious symbol" and strictly apply the standards annunciated in Lemon. JOSHUA D. ZARROW