In November 2004, while the appeals were pending, Oregon voters adopted Ballot Measure 36 (2004), a voter-initiated amendment to the Oregon Constitution aimed at defining marriage as a relationship between one man and one woman. That amendment, which became effective on December 2, 2004, provides:

"It is the policy of Oregon, and its political subdivisions, that only a marriage between one man and one woman shall be valid or legally recognized as a marriage."

This court solicited supplemental briefing before hearing oral arguments in these matters, asking the parties to address the effect (if any) of that new constitutional provision on the issues raised in these appeals. In that regard, we need not examine each and every issue tendered to us by the responding parties respecting Measure 36; indeed, most are not relevant to the present proceedings. One issue, however, is pertinent, because it affects the prospective ability of five of the plaintiff same-sex couples to pursue their claims that Article I, section 20, entitles them to obtain marriage licenses on the same terms as opposite-sex couples. That issue is whether Measure 36 is an operative statement of law or whether it is only a statement of aspirational principle that requires some further action to make it enforceable. If it is the former, then the measure forecloses the five same-sex couples from obtaining the marriage licenses that they seek. If it is the latter, then the measure may not foreclose their Article I, section 20 claim. We turn to that issue. (10)

In interpreting voter-initiated constitutional provisions, our goal is to discern the intent of the voters. Flavorland Foods v. Washington County Assessor, 334 Or 562, 567, 54 P3d 582 (2002). In doing so, the text of the constitutional provision itself provides the best evidence of the voters' intent. Martin v. City of Tigard, 335 Or 444, 451, 72 P3d 619 (2003). This court also considers the context of the provision, which includes other relevant constitutional provisions, case law from this court, and any relevant statutory framework in effect at the time that the voters adopted the provision. Id. If the voters' intent is clear from the text and context, then the court does not look further. If the provision's meaning remains ambiguous, however, the court will consider the history of the provision in an effort to resolve the matter. Flavorland Foods, 334 Or at 567.

In this case, the text of the new constitutional provision states that its substantive content is the "policy" of Oregon. The parties disagree about the legal effect of the term "policy" in the new provision: Is it an operative statement of law or just an aspirational principle that requires further action to establish an enforceable restriction?

In examining the text of a constitutional provision adopted by initiative or legislative referral, this court typically gives words of common usage their plain, natural, and ordinary meaning. Coultas v. City of Sutherlin, 318 Or 584, 588-89, 871 P2d 465 (1994). The dictionary definitions most applicable to the word "policy," as it is used in this context, are:

"a: a definite course or method of action selected (as by a government, institution, group, or individual) from among alternatives and in the light of given conditions to guide and usually determine present and future decisions. b(1): a specific decision or set of decisions designed to carry out such a chosen course of action (2): such a specific decision or set of decisions together with the related actions designed to implement them c: a projected program consisting of desired objectives and the means to achieve them[.]"

Webster's at 1754. Clearly, a "policy" may be something more than a set of intentions. Giving the word its plain and ordinary meaning, a "policy" can be a concrete course of action, the law necessary to implement it, or both.

The foregoing definitional review is not dispositive; the wording of Measure 36 still could be hortatory. However, two other considerations demonstrate that the amendment is intended to state present law.

First, Measure 36 states that "only a marriage between one man and one woman shall be valid or legally recognized." That is a statement not only of the policy itself, but also of particular consequences that are to occur as a result of that policy. Such wording is operational, not aspirational.

Second, Measure 36 lacks any wording directing the legislature to carry out the stated policy by appropriate legislation. If the measure were aspirational only, then we reasonably might expect to see such wording.

Based on the foregoing, we conclude that the use of the word "policy" in Measure 36 is intended to signal a presently enforceable tenet of Oregon constitutional law. And, with respect to the remaining text, there is no ambiguity regarding the measure's substantive effect. Today, marriage in Oregon -- an institution once limited to opposite-sex couples only by statute -- now is so limited by the state constitution as well. As the later-enacted (and more specific) constitutional provision, Measure 36 resolves any prospective claims that plaintiffs may have had under Article I, section 20, to obtain marriage licenses. The claims of the five same-sex couples that they are entitled as a matter of state law, now or hereafter, to obtain marriage licenses and to marry thus fail.

The parties also differ over the effect, if any, of the adoption of Measure 36 on the remaining issues in these appeals and, particularly, on the remedy that the trial court fashioned. In that regard, plaintiffs argue that, although the text of the measure prohibits same-sex marriage itself, it omits any reference to the benefits of marriage. Therefore, according to plaintiffs, Measure 36 does not speak to the issue whether Article I, section 20, prohibits using gender or sexual orientation as a basis for denying the benefits of marriage. Accordingly, plaintiffs urge this court to conclude that the voters did not intend to hinder this court from fashioning a remedy in these appeals that extends such benefits to same-sex couples. However, the issue of the availability of marriage benefits to same-sex couples is not properly before us. At trial, plaintiffs did not seek access to the benefits of marriage apart from, or as an alternative to, marriage itself. The trial court therefore improperly went beyond the pleadings in fashioning the particular remedy that it chose. We do not address that topic further.

Plaintiffs also raise issues concerning the effect of Measure 36 on the remaining same-sex couples, who received licenses and participated in marriage ceremonies before that measure became effective. They argue that the measure cannot be construed to affect the legal validity of those relationships because nothing in the text or context of the measure indicates an intent to either (1) retroactively invalidate the challenged marriage contracts; or (2) prospectively invalidate those contracts from the effective date of Measure 36. (11) That argument assumes that those marriages were legally valid before the adoption of Measure 36. However, we disagree with that premise. As we explain below, the county did not have authority to issue the licenses for the marriages in question....

County officials were entitled to have their doubts about the constitutionality of limiting marriage to opposite-sex couples. But, marriage and the laws governing it are matters of statewide, not local, concern. Thus, the remedy for such a perceived constitutional problem would be either to amend the statutes to meet constitutional requirements or to direct some other remedy on a statewide basis. Obviously, any such remedy must originate from a source with the authority to speak on that basis. The legislature has such authority and, in an appropriate adversary proceeding, the courts have it as well. But there is no source of law from which the county could claim such authority. To the contrary, the county's involvement in the license-issuing process is ministerial only.

These appeals do not require us to explore the full range of actions from which a governmental official might choose in vindicating that official's personal constitutional vision. Cooper illustrates one such way that that might occur, if the official has quasi-judicial authority. Another available choice, when an official is vested with discretion, is to choose not to act, such as when a prosecutor chooses not to prosecute a case under a statute of questionable constitutional validity. Yet a third choice, when an official has no discretion, might be to decline to perform a statutory duty and leave it to a party aggrieved by that action to seek a contested case decision or judicial intervention through mandamus or declaratory judgment proceedings. But none of those alternatives is analogous to what the county did here....

In summary, we conclude as follows. First, since the effective date of Measure 36, marriage in Oregon has been limited under the Oregon Constitution to opposite-sex couples. Second, Oregon statutory law in existence before the effective date of Measure 36 also limited, and continues to limit, the right to obtain marriage licenses to opposite-sex couples. Third, marriage licenses issued to same-sex couples in Multnomah County before that date were issued without authority and were void at the time that they were issued, and we therefore need not consider the independent effect, if any, of Measure 36 on those marriage licenses. In short, none of plaintiffs' claims properly before the court is well taken. Finally, the abstract question whether ORS chapter 106 confers marriage benefits in violation of Article I, section 20, of the Oregon Constitution is not properly before the court.

The judgment of the circuit court is reversed, and the case is remanded to the circuit court with instructions to dismiss the action.