Judge Cooper's opinion rests upon an interpretive paradigm that seems rather different from the typical opinion on these matters. This matter was decided after a full trial; no "construing all facts in favor of the nonmoving party" or "assuming, as we must, that all allegations in the complaint are true" here! That, however, is not all that Judge Cooper did that breaks with "legal tradition": He took notice of reality beyond the parties' attempts to confine it. The weakest part of the opinion—rhetorically, structurally, factually, and logically—is his examination of the first prong of the Lemon test. Although he didn't say it this bluntly, because he is at least marginally constrained by the need to maintain the dignity of the bench, the dozen or so pages devoted to that analysis boil down to this (as I might have written it in a particularly charitable mood):
The School Board was careful to avoid any overt statement of religious purpose. So long as a rational person who knew nothing of preexisting conditions, personalities, political pressures, election results, or virtually anything else of substance related to the matter's context could believe that the School Board's pointed silences and post hoc non-religious rationalizations for the Sticker indicate that there was, in this specific instance, a nonsectarian purpose, I am constrained to find that there was one. Adler v. Duval Cty. Sch. Bd., 206 F.3d 1070 (11th Cir. 2000), reinstated on remand, 250 F.3d 1330 (11th Cir. 2001); Bown v. Gwinnett Cty. Sch. Dist., 112 F.3d 1464, 1469 (11th Cir. 1997). The context indicating that this merely evades reality cannot be considered when analyzing the purpose behind actions of the political branches of government; we assume that all legislators, executives, and administrators have only constitutional motives, regardless of what we might infer of those motives from other actions, evidence, and statements outside the scope of the particular matter. Only a clearly "sham" secular purpose can overcome this presumption, and the evidence in this matter does not reach that exceptionally stringent standard. Edwards v. Aguillar, 482 U.S. 578, 586–87 (1987); Wallace v Jaffree, 472 U.S. 38, 75–76 (1985) (O'Connor, J., concurring) ("Even if the text and official history of a [statement] express no secular purpose, the [statement] should be held to have an improper purpose only if it is beyond purview that endorsement of religion or a religious belief 'was and is the… reason for the statement's existence.'") (emphasis added).
(Note, again, this is not an actual quotation, but a hypothetical passage.)
This does not mean that reality was completely on vacation in Selman. Bluntly, it is clear from the evidence that at minimum advancement of a religious point of view under political pressure was the sticker's purpose; Judge Cooper couldn't make that finding, though, because precedent restricts him from being properly skeptical of stated motivations. Instead, he reserved reality for the part of the case in which he was explicitly allowed to consider it: the result of the policy, however laudable (or at least not-unlaudable) its purpose may have been. In remarkably temperate language, Judge Cooper found that the sticker advanced the political agenda of Christian fundamentalists in Cobb County at the expense of those residents of Cobb County who do not share that set of beliefs; this is enough to invalidate the sticker, and the underlying policy. He did so principally by relying upon context, and not literal language and findings. The irony that his legal interpretive method itself contradicts the particular religious interpretive method at issue will no doubt escape virtually every media representative, and too many legal pundits and scholars. Most will instead argue on the wisdom of the policy, and the supposedly conflicting needs to educate children in both science and "proper" belief, when Judge Cooper properly saw that the question in this matter is quite different.
Too often, a legal proceeding is treated as if it is a laboratory experiment. That is, no outside variables are allowed to impinge on the experiment (although any competent scientist could tell you that no such experiment has ever been done; if nothing else, the quality and cleanliness of one's lab equipment matter). Judge Cooper recognized that in the "laboratory" of representative politics, there are always outside variables, and that they are particularly relevant when the argument is over what may be in the laboratory in the first place. The reflexive irony of this recognition is far too subtle for ordinary public discourse; which, I suppose, puts it well within the bounds of the blawgosphere.