The issue to be decided is the constitutional validity of a statute creating and defining the crime of “coercion.” Defendants were indicted under one subsection of the statute, ORS 163.275(1)(e), which makes it a crime to compel or induce another person to engage in conduct from which he has the legal right to abstain by causing him to fear the disclosure of discreditable assertions about some person.1 Defendants demurred on the ground that the terms of the statute are too vague for a penal law.

The trial court sustained the demurrers. On the state’s appeal, the Court of Appeals reversed, upholding the validity of the statute and the indictment by divided opinions first in a panel decision in State V. Robertson, 54 Or App 630, 635 P2d 1057 (1981) and State v. Young, 54 Or App 681, 635 P2d 681 (1981), and subsequently in banc, with three judges dissenting, in State v. Paige, 55 Or App 519, 638 P2d 1173 (1981). Having allowed review and heard argument in the first two cases, we take account also of the opinions of the entire court in Paige.

I. The state’s appeal.

There is a preliminary question whether the Court of Appeals obtained jurisdiction by the state’s appeal. The circuit court in each case entered an order sustaining defendant’s demurrer to the coercion count of the indictment, which was based “on the ground that the facts therein do not state a crime for the reason that the same and ORS 163.275 are unconstitutionally vague.” Finding no statute whose literal terms authorize an appeal from such an order, we asked the parties to submit additional memo-randa on the issue.

Since the Deady Code of 1864, the statutes have required that upon considering a demurrer in a criminal case, “the court shall give judgment, either allowing or disallowing it, and an entry to that effect shall be made in the journal.” ORS 135.660. For a century between that code and 1963, the state could appeal only from “a judgment for the defendant, on a demurrer to the indictment” or from an *405order arresting judgment. Act of Oct. 19, 1864, § 227, in Oregon Laws 1845-64, at 480. ORS 138.060 (1961 ed.) An order merely sustaining a demurrer was not appealable. State v. Cloran, 233 Or 400, 374 P2d 748 (1962), State v. Davis, 207 Or 525, 296 P2d 240 (1956). A 1963 amendment added an appeal from an order sustaining a plea of former conviction or acquittal; the existing reference to a “judgment for the defendant on a demurrer” was reenacted. 1963 Or Laws ch 385.

In 1968, the state attempted to appeal from a trial court order dismissing an indictment upon defendant’s motion. This court dismissed the appeal because the order was not a judgment on a demurrer and therefore not within the list of appealable orders. State v. Sieckmann, 251 Or 259, 445 P2d 599 (1968). Thereafter the Department of Justice obtained an amendment which allowed the state to appeal from an “order made prior to trial dismissing the indictment.” The amendment did not add these words to ORS 138.060; rather, they replaced the prior reference to an appeal from a judgment on a demurrer, removing the explicit basis for such an appeal. 1971 Or Laws ch 644. The statute was further amended in 1973 to add the words “setting aside” to “dismissing” and broadening “indictment” to “accusatory instrument,” so that the relevant provision now reads: “The state may take an appeal from the circuit court or the district court to the Court of Appeals from: (1) An order made prior to trial dismissing or setting aside the accusatory instrument.” ORS 138.060(1).

At the same time, the legislature enacted a new provision providing for dismissal of an accusatory instrument, ORS 135.470. The statutes now prescribe different grounds for demurring to an accusatory instrument, for “dismissing” such an instrument, and for “setting aside” an indictment.2 The demurrer in the present case properly *406invoked failure to state an offense, on grounds of constitutional defect. Under ORS 135.660 this would lead to a “judgment” that would have been appealable under ORS 138.060 as it stood until 1971, when the reference to appeals from judgments on demurrers was eliminated from ORS 138.060.

The state’s memorandum in response to this court’s questions concedes that the statutes now do not expressly allow the state to appeal from an order or judgment on a demurrer “if the present Oregon criminal procedure statutes are read literally, with an insistence upon precise and uniform terminology, and with an incomplete appreciation of the history behind them.” The state argues, however, that a proper appreciation of that history should persuade us not to limit ORS 138.060 literally to orders “dismissing” accusatory instruments. It contends that when the Department of Justice, after State v. Sieckmann, supra, induced the legislature to substitute “order. . . dismissing the indictment” for “judgment for the defendant on a demurrer,” the state did not mean to sacrifice its longstanding right to appeal from adverse judgments on demurrers but to expand it to include orders of dismissal, and also orders sustaining demurrers without requiring a judgment prescribed by ORS 135.660, superseding the rule of State v. Cloran, supra, and the decisions there cited. In short, the department asks us to effectuate the policy objectives shown by the legislative history of the 1971 and 1973 amendments.

Though the legislative history is sparse, it appears consistent with the department’s version of the intended goal of the amendments.3 When the 1971 amendment removed judgments on demurrers from ORS 138.060, any effect of rendering them unappealable was certainly *407unintended. In deciding that it had jurisdiction of this appeal, the Court of Appeals extended its earlier decision in State v. Thomas, 32 Or App 85, 573 P2d 1259, reconsidered in 34 Or App 187, 578 P2d 452 (1978), which it had initially decided the other way. In the second Thomas opinion, the court in banc expressed continued doubt whether an order sustaining a demurrer but not dismissing the accusatory instrument was appealable under the 1971 amendment, but the court concluded that at least the 1973 addition of the words “setting aside” to ORS 138.060 covered such an order if the 1971 reference to orders “dismissing” an indictment did not. As the department recognizes, the statutes can benefit from further revision. Nevertheless, we are not inclined to disturb the conclusion of the Court of Appeals that because the purpose of the amendments was to broaden the state’s ability to appeal orders that invalidate accusatory instruments, the court may take jurisdiction of such appeals from “orders” sustaining demurrers whether or not they are identified as judgments, as prescribed in ORS 135.660.

II. The issues.

Although the defendant’s demurrer alleged only that the crime charged in the indictment was “unconstitutionally vague,” the Court of Appeals construed a supporting trial memorandum as having attacked the statute also as unconstitutionally “overbroad.” As the court proceeded to decide the case on that basis, we accept its characterization of the memorandum. But we once again draw attention to the frequently misunderstood difference between the two constitutional claims. “Vagueness” and “overbreadth” of a law are not two alternative or cumulative epithets for the same shortcoming.

*408The rule against vague penal laws has been rested on various constitutional premises. In State v. Hodges, 254 Or 21, 457 P2d 491 (1969), this court suggested that abdication of the lawmakers’ responsibility to define a crime to prosecutors, judges, or jurors in case-by-case adjudication allowed those charged with enforcing the law to make the law after the event:

“A law that permits the judge and jury to punish or withhold punishment in their uncontrolled discretion is defective as much for its uncertainty of adjudication as for its failure to notify potential defendants of its scope and reach.

“. . . A vague statute lends itself to an unconstitutional delegation of legislative power to the judge and jury, and, by permitting the jury to decide what the law will be, it offends the principle, if not the rule, against ex post facto laws. See Oregon Constitution, Art I, §21.”

254 Or at 27, quoted in State v. Blair, 287 Or 519, 522-23 (1979).4 Perhaps the vice of the “uncontrolled discretion” mentioned in Hodges lies as much in inviting standardless and unequal application of penal laws, contrary to article I, section 20.5

*409Moreover, the unfairness of “failure to notify potential defendants of [the law’s] scope and reach” can constitute a denial of due process under the federal 14th amendment. See Lanzetta v. New Jersey, 306 US 451, 59 S Ct 618, 83 LEd 888 (1939). Recently in Village of Hoffman Estates v. Flipside, Hoffman Estates, Inc., 455 US 489, 102 S Ct 1186, .71 LEd2d 362 (1982), the Supreme Court of the United States reaffirmed the constitutional objections to vague laws in terms much like those in our own cases, quoting from Groyned v. City of Rockford, 408 US 104, 108, 92 S Ct 2294, 33 LEd2d 222 (1972):

“ ‘Vague laws offend several important values. First, because we assume that man is free to steer between lawful and unlawful conduct, we insist that laws give the person of ordinary intelligence a reasonable opportunity to know what is prohibited, so that he may act accordingly. Vague laws may trap the innocent by not providing fair warning. Second, if arbitrary and discriminatory enforcement is to be prevented, laws must provide explicit standards for those who apply them. A vague law impermissibly delegates basic policy matters to policemen, judges, and juries for resolution on an ad hoc and subjective basis, with the attendant dangers of arbitrary and discriminatory applications.’ (footnotes omitted).”

Village of Hoffman Estates, supra, 455 US at 498.6 In the Oregon Criminal Code of 1971, the legislature enacted a commitment to “give fair warning of the nature of the conduct declared to constitute an offense and of the sentences authorized upon conviction.” ORS *410161.025(l)(c). That commitment must be respected in the interpretation and application of the code.

“Overbreadth,” however, is a different claim. It should be kept in mind that the terms “overbroad” and “overbreadth” are not themselves terms of the state or federal constitutions, any more than the terms “vague” or “vagueness”; they are only lawyers’ phrases for shortcomings that are claimed to contravene other constitutional constraints. They have been used by different theorists and courts to mean different things and to carry different consequences. See Monaghan, Overbreadth, 1981 Sup Ct Rev 1 (1981). In principle, however, a claim of “over-breadth” asserts that the terms of a law exceed constitutional boundaries, purporting to reach conduct protected by guarantees such as, for instance, Oregon Constitution, article I, section 8 (freedom to speak and write) or section 27 (right to bear arms). As this court recently stated in a case under section 27:

“An ‘overbroad’ law, as that term has been developed by the United States Supreme Court, is not vague, or need not be. Its vice is not failure to communicate. Its vice may be clarity. For a law is overbroad to the extent that it announces a prohibition that reaches conduct which may not be prohibited. A legislature can make a law as ‘broad’ and inclusive as it chooses unless it reaches into constitutionally protected ground. The clearer an ‘overbroad’ statute is, the harder it is to confine it by interpretation within its constitutionally permissible reach.”

State v. Blocker, 291 Or 255, 261, 630 P2d 824, 827 (1981).

A claim of “overbreadth” cannot properly assert that the words of the law, read literally, are broader than the lawmaker intended, for that is an issue to be resolved by interpretation. A constitutional claim that the law as interpreted cannot be discerned from its terms once again is a claim of vagueness, not overbreadth. Again, the Supreme Court noted the tendency to confuse the arguments in Village of Hoffman Estates, supra, in which Flipside challenged an ordinance prohibiting the sale of paraphernalia designed for illegal drug use:

“Flipside also argues that the ordinance is ‘overbroad’ because it could extend to ‘innocent’ and ‘lawful’ uses of items as well as uses with illegal drugs . . . This argument *411seems to confuse vagueness and overbreadth doctrines. If Flipside is objecting that it cannot determine whether the ordinance regulates items with some lawful uses, then it is complaining of vagueness.”7

Village of Hoffman Estates, supra, 455 US at 497, n. 9. The Court of Appeals correctly understood the distinction.

When a statute is attacked as vague, for failing to define and communicate its coverage, the statute sometimes can be saved by a judicial interpretation that gives it the required definiteness. It is the court’s obligation to do so when this can be done without departing too far from what the legislature sought to accomplish or what the statute itself can convey to a reader. But when such a saving construction cannot be attributed to the legislature with reasonable fidelity to the legislature’s words and apparent intent, the statute is invalid as enacted, and it is immaterial whether the particular case in which it is challenged would be immune from a validly drawn law.8

*412A narrowing construction similarly may save a statute attacked as “overbroad,” unless the constitutional guarantee invoked against the statute forbade its very enactment as drafted. Article I, section 8, for instance, forbids lawmakers to pass any law “restraining the free expression of opinion, or restricting the right to speak, write, or print freely on any subject whatever,” beyond providing a remedy for any person injured by the “abuse” of this right.9 This forecloses the enactment of any law written in terms directed to the substance of any “opinion” or any “subject” of communication, unless the scope of the restraint is wholly confined within some historical exception that was well established when the first American guarantees of freedom of expression were adopted and that the guarantees then or in 1859 demonstrably were not intended to reach. Examples are perjury, solicitation or verbal assistance in crime, some forms of theft, forgery and fraud and their contemporary variants. See Greenawalt, Speech and Crime, 1980 Am B Found Res J 645, 648-70. Only if a law passes that test is it open to a narrowing construction to avoid “overbreadth” or to scrutiny of its application to particular facts. As the court recently said of another challenge under article I, section 8:

“This constitutional provision is a prohibition on the legislative branch. It prohibits the legislature from enacting laws restraining the free expression of opinion or restricting the right to speak freely on any subject. If a law concerning free speech on its face violates this prohibition, it is unconstitutional; it is not necessary to consider what the conduct is in the individual case. If the law is not unconstitutional on its face, it nevertheless might be applied in a manner that would violate Art I § 8.”

State v. Spencer, 289 Or 225, 228, 611 P2d 1147, 1148 (1980).10

*413The question is whether a saving construction is necessary, and if so, whether it is possible in the case of the coercion statute. The Court of Appeals took the defendants’ contentions in this case to raise the claim that by indiscriminately reaching expression designed to induce another’s conduct by means of threats or warnings, the coercion statute as written is a law restricting freedom of speech contrary to article I, section 8 of the Oregon Constitution and the federal first amendment. The court rejected this claim as well as that of vagueness. The issue, therefore, is whether the statute that the legislature passed can be understood in terms clear enough to avoid vagueness and also confined within constitutionally permissible limits. For the reasons that follow, we conclude that it cannot.

III. The statute.

ORS 163.275 provides:

“(1) A person commits the crime of coercion when he compels or induces another person to engage in conduct from which he has a legal right to abstain, or to abstain from engaging in conduct in which he has a legal right to engage, by means of instilling in him a fear that, if the demand is not complied with, the actor or another will:

“(a) Cause physical injury to some person; or

“(b) • Cause damage to property; or

“(c) Engage in other conduct constituting a crime; or

“(d) Accuse some person of a crime or cause criminal charges to be instituted against him; or

“(e) Expose a secret or publicize an asserted fact, whether true or false, tending to subject some person to hatred, contempt or ridicule; or

*414“(f) Cause or continue a strike, boycott or other collective action injurious to some person’s business, except that such a threat shall not be deemed coercive when the act or omission compelled is for the benefit of the group in whose interest the actor purports to act; or

“(g) Testify or provide information or withhold testimony or information with respect to another’s legal claim or defense; or

“(h) Use or abuse his position as a public servant by performing some act within or related to his official duties, or by failing or refusing to perform an official duty, in such manner as to affect some person adversely; or

“(i) Inflict any harm which would not benefit the actor.

“(2) Coercion is a Class C felony.”

We turn to an examination whether the statute on its face is a law whose enactment was forbidden by article I, section 8, or if not, whether it either is impermissibly vague or written so broadly as potentially to reach communications that must remain free.

The Court of Appeals divided on the question whether ORS 163.275 is a law “restricting the right to speak . . . freely on any subject whatever.” The majority of the Robertson panel held that it was not. The panel distinguished State v. Spencer, supra, on the ground that “speech itself was the direct object of [the] disorderly conduct provision,” whereas the coercion statute “is not directed at the act of communicating and in that respect is like any other criminal statute defining a crime, such as robbery, which may incidentally involve communication.” In the subsequent decision in banc in State v. Paige, supra, the dissenters disputed that view of the statute:

“Contrary to the view expressed in Robertson, the act forbidden by the statute in question here is clearly and solely speech. ‘Compelling’ or ‘inducing’ another person to engage in some conduct from which that person has a legal right to abstain, or to abstain from engaging in conduct in which that person has a legal right to engage, by means of instilling fear in that person that the one making the threat will expose a secret or publicize an asserted fact can only be accomplished by speech, be it telephonic, telegraphic, parol, or by sign language. It is the act of making the threat which is the gravamen of the offense.”

*41555 Or App at 524. The Paige majority disagreed with the dissent’s statement that “the act of making the threat. . . is the gravamen of the offense.” It noted that “the offense of coercion is not committed until the victim acts.” Because a threat of adverse consequences is not forbidden unless successful, the majority found that “[t]he making of the threat is not the basis of the offense, but only a manner of bringing it about.” It therefore concluded that the coercion statute proscribes “not speech but conduct — the act of overpowering another’s will.” 55 Or App at 522. The diverging premises touch some of the most difficult themes in the constitutional law of free speech.

In our view, it cannot be said that the coercion statute does not forbid speech. We agree that because the statute is directed in terms against the pursuit of a forbidden effect and not, like the disorderly conduct provision in Spencer, against forbidden speech as such, it is not a law whose enactment was for this reason alone wholly withdrawn from legislative authority by article I, section 8. But because speech is a statutory element in the definition of the offense, the statute is susceptible to attack for possible overbreadth.

Subsection (1) of ORS 163.275, which formulates the crime of “coercion,” has two major parts. The opening clause defines the essence of the crime as compelling or inducing otherwise voluntary conduct by means of fear. The remainder of the sentence qualifies the definition by specifying the kinds of consequences which may not be threatened in order to compel or induce the demanded conduct.

The elements stated in the opening clause are (1) that the accused makes a demand upon another person, (2) that the addressee otherwise is legally free to choose whether to engage in the demanded conduct, (3) that the accused puts the addressee in fear of one of the specified adverse actions by the accused or someone else if the demand is not met, and (4) that this fear compels or induces the addressee to comply with the demand. In view of the last element it is accurate, as the Paige majority said, that the offense is not committed until the victim acts. But this does not demonstrate that the statute is not directed against speech.

*416The offense presupposes a “demand” by the defendant, which means a communication in words or otherwise; and as a practical matter, if not as a logical necessity, the legislature no doubt also assumed that the impelling fear will be instilled by speech or its equivalent. If the victim acts in fearful anticipation without any demand, the definition of coercion is not met. The demand, the fear-instilling communication of a specified consequence of noncompliance, and the fear-induced compliance all are essential; the crime is not made out if the addressee, though frightened, does not comply or if he acts for reasons independent of the demand or the threat. The challenge under article I, section 8, therefore cannot be dismissed simply by saying that the statute forbids an “act” rather than “speech.” That distinction could be helpful if a law proscribed threatening gestures or other intimidating acts, or if it became necessary to differentiate between the communicative content and the noncommunicative means of proscribed expression. But speech often would be the offender’s only act in committing this crime. Indeed, the statute leaves it immaterial whether an accused or another person had either the intent or the capacity to carry out the threatened consequences; a successful bluff seems to suffice.

There is an intrinsic, though not insurmountable, paradox of overbreadth analysis. If a law is invalid whenever it literally could be applied to restrict free expression, there seemingly will never be occasion to hold that a valid law was unconstitutionally applied in a particular case. In practice the paradox causes less trouble when speech as such is not a specified element or intended target of the restrictive law. That an offense includes the use of words is not in itself fatal to the enactment of a prohibition in terms directed at causing harm rather than against words as such. Communication is an element in many traditional crimes. As stated above, article I, section 8, prohibits lawmakers from enacting restrictions that focus on the content of speech or writing, either because that content itself is deemed socially undesirable or offensive, or because it is thought to have adverse consequences. This is the principle applied in State v. Spencer, supra. It means that laws must focus on proscribing the pursuit or accomplishment of *417forbidden results rather than on the suppression of speech or writing either as an end in itself or as a means to some other legislative end. Spencer, in turn, quoted from State v. Blair, 287 Or 519, 523, 601 P2d 766, 768 (1979), which invalidated a section of the harrassment law for vagueness:

«* * * No doubt the difficulty is sharpened by the fact that ORS 166.065(1)(c) expressly makes the gravamen of the offense that the offender communicates rather than that he subjects the victim to some defined injury, of which the communication is only the means employed in the particular case. A statute that does not in its own terms forbid speech or other communication still would require sensitive confinement within constitutional limits, but it is less vulnerable to constitutional attack on its face.”

289 Or 225 at 229, 611 P2d at 1149. The statute invalidated in Spencer prohibited “abusive or obscene language, or ... an obscene gesture, in a public place” if intended to cause “public inconvenience, annoyance or alarm,” whether or not the language or gesture in fact did this. If that statute had been directed only against causing the forbidden effects, a person accused of causing such effects by language or gestures would be left to assert (apart from a vagueness claim) that the statute could not constitutionally be applied to his particular words or other expression, not that it was drawn and enacted contrary to article I, section 8.11

To this extent, the Court of Appeals correctly distinguished State v. Spencer. The coercion statute is written so as to focus on a forbidden effect — the effect of frightening another person into a nonobligatory and undesired course of conduct. But the statute continues by specifying that the compulsion or inducement may not be imposed in aid of a “demand,” and by means of specified verbal warnings or threats. We do not suggest that ORS 163.275 would be a better penal statute if it ended before the specification of means; as already stated, legislative attention to the reach of a penal law is a major object of the rule against vagueness. When the proscribed means *418include speech or writing, however, even a law written to focus on a forbidden effect, as ORS 163.275 is, must be scrutinized to determine whether it appears to reach privileged communication or whether it can be interpreted to avoid such “overbreadth.”

IV. The apparent reach of the coercion statute.

A notable characteristic of ORS 163.275 is that it requires neither the conduct demanded of the victim nor the action threatened in case of refusal to be wrongful. It therefore does not fit into the constitutional analysis of solicitation, conspiracy, advice, or other verbal participation in another’s criminal act, nor is it confined to threats of crimes or torts against an uncooperative victim, as in a robbery or rape. Only one item among the threatened consequences listed in ORS 163.275(1) assumes illegality, the catch-all item (c), “other conduct constituting a crime.” It is not contended that the word “other” in item (c) implies that the other threatened consequences listed in the law are covered only insofar as their execution would be illegal, and their character precludes such a construction of several of them. Some, indeed, would often be constitutionally privileged if carried out, such as the publication of derogatory facts under (e), or would actually be legal duties, such as giving testimony under (g) or performing one’s official duties under (h). Neither the action demanded nor the occurrence of the threatened consequences upon a refusal is the concern of this statute. The target of the law is the effective use of fear to induce compliance with a demand.

This spacious coverage offers tempting opportunities for an “overbreadth” attack on the statutory text, limited only by the imagination of court and counsel. This is particularly true of ORS 163.275(l)(e), supra, dealing with threats to expose or publicize some asserted fact that would tend to subject some person to hatred, contempt, or ridicule, under which the present charge was brought. Thus defendants posited situations in which one man tells another: “If you don’t quit making love to my wife, I’m going to tell your wife,” or someone proposes to disclose an airline pilot’s secret illness if he does not get medical attention, or a politician’s embarrassing past if he does not withdraw his candidacy from office. The dissenters in the *419Court of Appeals added illustrations in which a journalist advises a public official that he will disclose private facts showing an official’s financial interest in a pending measure if the official does not refrain from voting on the measure, or in which “one appellate judge might tell another, ‘Change your opinion, or I shall dissent and expose your complete ignorance of this area of the law.’ ” 55 Or App at 524 (Gillette, J., dissenting). Indeed, they might have added that a prosecutor’s plea-bargaining attempts to induce a defendant to plead guilty to one charge in order to avoid prosecution on other charges literally violates ORS 163.275(1) (d).

Some of the examples, particularly hypothetical statements addressed to minors or to family members, are offered to show that the legislature could not possibly have meant to outlaw them, but that is not to the point. The question of overbreadth is not even whether the statute covers situations in which the actual disclosure would be privileged expression, but whether the hypothetical demand backed by the threat of such a disclosure would be so privileged. The examples of such demands drawn from a political context and involving consequences directly relevant to the demanded act plainly would be an exercise of free speech or writing.12 The real disagreement between defendants and the prosecution, therefore, is whether ORS 163.275 can be given a principled interpretation that excludes its application to these and other instances of free expression.

The answer cannot be simple. The first difficulty lies in the search, not for an interpretation of ORS 163.275, but for the principles that the interpretation must preserve. Judicial and academic analyses of the principles governing freedom to make demands coupled with threats have been sparse and inconclusive, perhaps because these elements characterize the traditional domain of robbery and extortion of money by threats of bodily harm. Although the feared reaction of others underlies much inducement to *420desired conduct throughout social institutions, from family and schools to consumer advertising, a civilized legal system may choose to protect individual autonomy of choice against many kinds of compulsion, if its rules of civil or criminal liability remain within constitutional bounds. Yet some demands obviously are in the realm of free speech, as are warnings of the intended sanctions for noncompliance. It is easy to extend the political examples to others from academic, cultural, or religious settings and beyond.13 Verbal threats to take or initiate even unlawful, violent action are not beyond first amendment protection by their content alone, divorced from any imminent realization.14

Principles for a free speech analysis of demands combined with threats might be sought in various elements of the combination, such as (1) the lawfulness or unlawfulness of the demanded conduct; (2) the lawful, unlawful, or constitutionally privileged nature of the threatened conduct; (3) the objective and the motive of the person making the demand; (4) the relationship between the parties to the demand or between the parties to the threatened consequences; (5) the relevance of the threatened consequences to noncompliance with the demand. Additional variables might be (6) the means of expression employed in the demand or the threat, (7) the likelihood and imminence of the threatened acts, or (8) other distinctions in the social setting or function of the demand. We turn to an examination whether one or a combination of these variables explains such precedents as exist on the constitutional question.

*421V. Extortion, intimidation, and coercion in the courts.

The classic form of protection against being compelled by threats to give up what one is under no obligation to yield is found in the long-established laws against blackmail15 or similar forms of extortion. On historic grounds alone, we have no doubt that these or their contemporary equivalents survived the proscription of speech-restraining laws in article I, section 8. A central characteristic of these classic crimes, however, was that the unlawful threats were used to obtain money or other things of value. Extortion developed as an extension of robbery by including threats other than threats of physical injury and is generally classified as one of the crimes against property. See Perkins, Criminal Law 372-75 (2nd ed 1969).16 It continues to be so classified in the Model Penal Code and in the Oregon Criminal Code, which treats “theft by extortion” as one form of theft. ORS 164.075.

Theft by extortion, as there defined, itself represents a significant extension of robbery by threat insofar as it includes threats of future actions that would not be unlawful or tortious if carried out. The kinds of threatened consequences required for theft by extortion, listed in ORS 164.075(l)(a)-(i), are identical to those for coercion listed in ORS 163.275(l)(a)-(i), above. The two offenses differ only in that theft by extortion consists in unlawfully compelling another to part with property. Conceivably it is not beyond challenge in some hypothetical applications. But assuming *422for present purposes that the verbal element common to the two offenses causes no “overbreadth” in theft by extortion, as long as the statute is properly applied, does the same verbal element potentially reach instances of privileged speech when it is extended beyond compelling the unwilling delivery of property to other forms of unwilling behavior? That depends on whether there are situations in which one is constitutionally privileged to use threats to persuade another to change his course of conduct, though not to obtain property for oneself or another.

Courts have been able to avoid this question when the charge before the court was extortion in the narrower sense stated above. In Carricarte v. State, 384 So2d 1261, cert den, 449 US 874 (Fla 1980), a Florida court was faced with an “overbreadth” claim when a lawyer was charged with extortion for telling a real estate developer that if the developer would not employ him, he would attempt to stop the development, organize and represent opponents of the project, and cause newspaper articles to be published against it. Beginning with premises much like those we have set out above, the court sustained the charge over defendant’s objection that he would have been constitutionally privileged in carrying out the alleged threats.17 The objection was held to fail because the charge included the elements of “malice,” a term not further explained, and “intent to acquire pecuniary gain.” Id. The court concluded that “malicious” threats to injure another, when made with the intent to “extort” money or to compel another’s conduct “against his will” would always be “communications undeserving of First Amendment protection.” 384 So2d at 1262-63. Similarly, the Louisiana Supreme Court rejected *423an overbreadth attack on an extortion statute when a police officer demanded sexual intercourse under threat of arresting the victim. The court construed the statute to cover “what is commonly known as blackmail,” prohibiting the use of threats to cause the victim to part with his property or otherwise act to the advantage of the threatener, and it concluded that when so construed, the extortion statute did not infringe upon or inhibit any constitutionally protected expression. State v. Felton, 339 So2d 797, 800 (La 1976).

Thus these decisions sustained extortion statutes on the basis that the forbidden threats were used simply to gain something of personal benefit to the threatener in a bilateral transaction. In each case, however, another element could be relevant to the result. In the Florida case, the lawyer’s demand for employment was independently unlawful as a forbidden solicitation of legal business. Carricarte v. State, supra, 384 So2d at 1263-64. In State v. Felton, supra, the police officer’s threat to arrest the victims if they did not comply with his demands, but not otherwise, no doubt was an unlawful abuse of his authority analogous to the original common law crime of official extortion, supra note 16. This leaves some doubt whether the analysis in these cases would have been the same if the threateners’ conduct had not been unlawful apart from their use of threats to obtain personal objectives.

Lacking the element of a self-seeking objective that characterizes blackmail or extortion statutes, on the other hand, a federal district court relied only on this second element, the independent unlawfulness of coercive acts,18 in rejecting an “overbreadth” challenge to the New York coercion statute on which ORS 163.275 is modeled, and which does not require such an objective. The court declined to enjoin a state prosecution for coercion at the suit of persons who had been charged with preventing the *424work of school personnel by means of “intimidation, physical force and interference and by means of any independently unlawful act.” Bishop v. Golden, 302 F Supp 502, 505 (EDNY 1969). Although the challenge was for overbreadth, the court did not look beyond the charge and discuss whether the statute as written might extend to demands that are neither accompanied by nor threaten “independently unlawful” conduct.

These and other opinions characteristically start from the conduct charged in the particular case rather than from hypothetical applications that are presented to show the overbroad reach of the challenged statute. Potential overbreadth is faced more squarely when a statute is attacked outside the context of a specific prosecution, so that there is no concrete charge under the statute to be tested for constitutionality. This is apparent in another case that produced a far more searching examination of the constitutional status of “intimidation,” though also conducted under a federal court’s view of the federal first amendment without reference to a state guarantee such as article I, section 8.

Landry v. Daley, 280 F Supp 938 (ND Ill 1968) was a class action brought on behalf of “the class of all negroes in the City of Chicago” as well as individual and organizational plaintiffs which attacked, among other statutes, an Illinois “Intimidation Statute” under which no plaintiff then faced prosecution. The statute declared one guilty of “intimidation” when

“with intent to cause another to perform or to omit the performance of any act, he communicates to another a threat to perform without lawful authority”

one of various acts somewhat similar to those listed in ORS 163.275. Ill Rev Stat ch 38, § 12-6 (1979). Plaintiffs claimed that this statute could be construed so as to punish them for threatening to engage in “direct action and protest activities” against racial discrimination. 280 F Supp at 960.

First, the court assumed, with some misgivings, that “threats” are not constitutionally protected expression if the character, intent, and circumstances of the threat are *425narrowly circumscribed.19 This meant at least that the context and circumstances reasonably cause the addressee to fear that the threatener will perform the threatened act. More important, the court construed the words “without lawful authority” as confining the paragraphs listing the forbidden threats to cover only unlawful acts, and even then it invalidated one such a paragraph for covering minor offenses. Thus the court construed the proscription of threats “to inflict physical harm” on a person or property or to “subject any person to physical confinement or restraint” as excluding such threats when their execution would not be unlawful. A paragraph referring to threats to “accuse any person of an offense” was construed to mean only a formal accusation of an innocent person made “with malice” and without reasonable grounds before a court or magistrate having jurisdiction to inquire into the allegation, in other words, when execution of such a threat would constitute malicious prosecution. Another paragraph forbidding threats to “expose any person to hatred, contempt or ridicule” was limited to mean actionable defamation or invasion of privacy. A paragraph on taking or withholding official action was sustained as referring to a threatened abuse of authority. A threat to “bring about or continue a strike, boycott or other collective action” was construed, again by virtue of the statutory phrase “without lawful authority,” to mean only unlawful collective action.

*426In sum, the Landry court proceeded on the premise that the Illinois law could validly prohibit “threats” if it was construed to require that a threat be realistic in circumstances and intent and that the threatened action would be unlawful if executed. Even this was qualified, however. The paragraph of the Illinois law that referred expressly to threats “to commit any criminal offense” proved to be the one provision that the court invalidated as overbroad. The court did so on the ground that although the state might have valid reasons to penalize even petty offenses if actually committed, these reasons would not always justify punishing a threat to act in the forbidden manner, for instance to block a street, to carry a concealed weapon, or to engage in disorderly conduct. 280 F Supp at 964. Thus the Landry opinion concluded that the statute validly prosribed only “threats” of independently wrongful conduct, and then only if limited so as to exclude minor or relatively inconsequential wrongdoing. This conclusion implicitly rejects a broad state power to restrict threatening speech simply in order to prevent the immediate evil of the coercive effect, a protective aim to which the wrongfulness of the threatened conduct seems irrelevant.20

It must be remembered that the Illinois statute did not require the element of selfish gain found in the extortion statutes discussed above. But the Landry court clearly was troubled by one application of the statute to threats of quite serious illegal action for material gains, the case of a threatened unlawful strike. It escaped its doubts only by *427once again evading the method of overbreadth analysis.21 The court’s caution is understandable, because in relation to labor, civil rights, and other organizational activities the potential reach of laws against threats or other coercive expression raises constitutional issues that have resisted clear or cogent analysis under the federal first amendment.

In one recent prosecution, Moore v. State, 519 SW2d 604 (Tenn 1975), defendant was charged with threats amounting to extortion when he demanded that a food store contribute to his organization, the Black Panther Party, assertedly for charitable programs. When the store, manager put off a decision, defendant was heard to say to a companion: “Well, I guess we’ll have to close them up.” Thereafter the store was picketed, with consequent loss of business. A Tennessee court rejected defendant’s objection that the picketing was privileged under the first amendment on the grounds that defendant intended “to pressure the store into contributing money to his cause” and that the store should be free not to contribute, “secure in the knowledge that no retribution will be forthcoming under the guise of constitutional protection.” 519 SW2d 606-07. To show the limits of picketing for an unlawful purpose, the court cited Hughes v. Superior Court, 339 US 460, 70 S Ct 718, 94 LEd 985 (1950). This is doubtful authority, because in Hughes, as in its predecessor, Giboney v. Empire Storage & Ice Co., 336 US 490, 69 S Ct 684, 93 L Ed 834 (1949), the picketing demanded unlawful action by the picketed enterprise, unlike the donations requested of the food store in Moore. On the other hand, the Supreme Court also has sustained a prohibition against picketing “to *428threaten, coerce, or restrain” a person who is not a party to a labor dispute. NLRB v. Retail Store Employees, 447 US 607, 100 S Ct 2372, 65 LEd2d 377 (1980).

Moore’s conviction was sustained in subsequent federal habeas corpus proceedings. The district court denied relief because Moore was “not convicted for his picketing activities but for threatening to picket for the purpose of extorting a payment.” Moore v. Newell, 401 F Supp 1018, 1021 (1975). The court thought it immaterial whether petitioner sought the payment to enrich himself or for charitable purposes.22 The Court of Appeals affirmed, apparently on the theory that a constitutional privilege of picketing exists for publicizing “a dispute or grievance with a business,” but that a merchant’s refusal to contribute to a charitable cause was not such a grievance.23 Moore v. Newell, 548 F2d 671, 672 (6th Cir 1977). This implies either that a privilege not only to picket, but to threaten to do so, arises precisely from a demand for selfish material gain, for instance a demand for higher wages or contributions to a health or pension fund, or perhaps that the privilege is limited to such demands on behalf of employees or others in dealings with the enterprises as such. But there is no special right to “speak or write freely” on subjects limited to labor relations.24 In Organization for a Better Austin v. Keefe, 402 US 415, 91 S Ct 1575, 29 LEd2d 1 (1971), the Supreme Court overturned an injunction against the distribution of leaflets accusing a real estate broker of “blockbusting” tactics, which the organization employed as a means to coerce the broker into signing an agreement to change his practices. Implicitly the demand on the broker *429coupled with the threat of disclosing his alleged practices were considered privileged along with the actual distribution of the leaflets, for the Court wrote:

“[T]he Appellate Court was apparently of the view that petitioners’ purpose in distributing their literature was not to inform the public, but to ‘force’ respondent to sign a no-solicitation agreement. The claim that the expressions were intended to exercise a coercive impact on respondent does not remove them from the reach of the First Amendment. Petitioners plainly intended to influence respondent’s conduct by their activities; this is not fundamentally different from the function of a newspaper. . . .”

402 US at 419.

Similar issues have arisen in applying state laws against coercive action to organized boycotts in the context of race relations. Since this case was argued, the United States Supreme Court has reversed a civil judgment for damages awarded to retail merchants for loss of business resulting from a consumer boycott organized in support of a list of demands for actions by various community institutions. National Association for the Advancement of Colored People v. Claiborne Hardware Co., 458 US 886, 102 S Ct 3409, 73 LEd2d 1215 (1982). The boycott was preceded by a demand on the merchants for support backed by the threat of a “selective buying campaign.” The Mississippi Supreme Court rejected the defendants’ constitutional claims on the grounds, first, that the demands and the boycott did not relate to a direct grievance against the merchants but demanded their support for political actions by the community, and second, that defendants violated a criminal law by threatening others with bodily harm to intimidate or coerce them to participate in the boycott. N.A.A.C.P. v. Claiborne Hardware Co., 393 So2d 1290 (Miss 1981).25 The United *430States Supreme Court held that damages must be limited to those attributable to violence. The verbal elements of the boycott did not lose their privileged character because the organizers “sought to persuade others to join the boycott through social pressure and the ‘threat’ of social ostracism” such as publicizing names of nonparticipants at meetings and in a newspaper. “Speech does not lose its protected character ... simply because it may embarrass others or coerce them into action.” 458 US at 910. The Court continued by quoting the passage from Organization for a Better Austin v. Keefe set forth above, saying of that case: “In dissolving the prior restraint, the Court recognized that ‘offensive’ and ‘coercive’ speech was nevertheless protected by the First Amendment.” 458 US at 911.

Our review of the foregoing cases is relevant, not because anything like civil rights boycotts or labor picketing is involved here, but because they have tested the courts’ ability to explain the validity or invalidity of laws that forbid verbal demands coupled with verbal threats. It cannot be said that any single explanation has emerged from the decisions.

To return to the variables we identified at the outset, the fact that the threatener’s demand is for money or other selfish gain is a central element in robbery by threat, blackmail and extortion. But it alone is not conclusive, for peaceful informational picketing and by implication the threat to picket are free speech when designed to gain financial payments or other legally permissible benefits.

Nor does any other element alone seem conclusive. In the Supreme Court’s picketing decisions, perhaps picketing as the particular form of communication is the distinguishing element; it is doubtful that the unlawfulness of | the objective would equally permit the suppression of pamphlets or other verbal forms of stating the demand and the threat to stop work or patronage. Compare N.A.A.C.P. v. Claiborne Co., supra; Organization for a Better Austin v. *431Keefe, supra. But a demand that another engage in illegal acts can come close to solicitation to crime and be forbidden within analogous constitutional limits. See generally, Greenawalt, Speech and Crime, supra, 655-670, 689-690, 742-751, and compare Emerson, The System of Free Expression 401-412 (1970).

The Illinois Intimidation Statute was saved in Landry v. Daley, supra, on the premise that it forbade only threats of unlawful action under circumstances creating a realistic fear of such action; but again, the court found the element of unlawfulness alone insufficient with respect to threats of relatively minor offenses, and it balked at extending this premise to threats of an unlawful strike. Its hesitancy may suggest a notion that labor relations constitute a special and potentially decisive element in the analysis of demands and threats. There is no apparent reason so to limit a privilege. The drafters of the American Law Institute’s Model Penal Code believed that “some categories of threats. . . should obviously be privileged.”26 Moreover, the Landry court’s premise seems contrary to the interpretation given the Illinois statute by the state’s own courts. See People v. Hubble, supra, n. 20. Whatever difference labor or race relations or other special universes of discourse may make in the federal case law, article I, section 8, extends freedom to speak, write, or print “on any subject whatever.”

VI. Application of Article I, section 8, to ORS 163.275.

In sum, our review of the cases that have tested laws against extortion, intimidation, or coercion under the *432first amendment yields no principled guidance on freedom of expression to state verbal demands coupled with verbal threats. A statute or other rule does not escape potential overbreadth solely because it deals with demands for property or benefits, or because it is limited to threats involving unlawful conduct, or because it is directed against the use of intimidating words or means of expression such as picketing. Each element narrows the potential infringement of free expression, but each element, standing alone, encompasses free expression under some circumstances.

The state argues that ORS 163.275(l)(e), the specific provision under which the present charge is brought, can be saved if it is interpreted to forbid only threats to make defamatory disclosures if the threatener’s demand is not met. The state points to statements before the Criminal Law Revision Commission to show that the commission was aware of the Illinois intimidation statute and apparently of Landry v. Daley, supra, and it argues that despite the absence in ORS 163.275 of the words “without lawful authority” upon which the Landry court sustained the statute, the Oregon law could and should be given an equally limited interpretation. Perhaps so; but the argument misses the mark. The question is not whether the threatened disclosure in case of noncompliance sometimes would be privileged. As we have said, the statute is not concerned with the actual occurrence of the threatened consequences listed in ORS 163.275(l)(a)-(i); several items in that list are not only legally privileged but legal duties. When the prohibition concerns the use of verbal threats of disclosure or other consequences in order to induce another to comply with a verbal demand, we do not share the Landry court’s conviction that freedom of expression hinges on the legality of the threatened action.27 The question therefore is whether this prohibition reaches instances of free speech so as to render the statute overbroad.

*433To recapitulate, we believe that the constitutional right to speak, write, or print freely on any subject whatever guaranteed in article I, section 8, was not meant to immunize the use of words in some respects relevant to ORS 163.275. As we have said, one of these is the use of words in the course of what indisputably would have been a conventional crime when Oregon’s Bill of Rights was adopted in 1859, or in the course of similar kinds of conventional crimes that lawmakers may from time to time enact.28 This includes the use of words in the course of soliciting, attempting, carrying out, or concealing other crimes. Therefore article I, section 8 would not foreclose a statute, otherwise in valid form, that made it criminal to compel another to commit an offense by threats or other verbal means under circumstances in which the demand is meant to be followed and the compulsion is realistically plausible. In such a statute, the focus is on the actual or probable commission of the compelled offense, not on protecting the addressee from hearing the speaker’s threats.29 ORS 163.275, however, is not such a statute. It is not concerned with the performance of the compelled act, which the statute does not require to be unlawful.

As we also have said, blackmail or analogous forms of extortion by threats are among the conventional crimes that survive article I, section 8, despite being committed by verbal means. The legislature, of course, may revise these crimes and extend their principles to contemporary circumstances or sensibilities. If it was unlawful to defraud people by crude face-to-face lies, for instance, free speech allows the legislature some leeway to extend the fraud principle to *434sophisticated lies communicated by contemporary means. Constitutional interpretation of broad clauses locks neither the powers of lawmakers nor the guarantees of civil liberties into their exact historic forms in the 18th and 19th centuries, as long as the extension remains true to the initial principle. See, e.g., State v. Kessler, 289 Or 359, 614 P2d 94 (1980) (definition of “arms” within Or Const art I, § 27); State ex rel Russell v. Jones, 293 Or 312, 647 P2d 904 (1982) (right to counsel); Sterling v. Cupp, 290 Or 611, 625 P2d 123 (1981) (“unnecessary rigor” under art I, § 13). When extending an old crime to wider “subjects” of speech or writing, however, there is need for care that the extension does not leave its historical analogue behind and, perhaps inadvertently, reach instances of privileged expression.

Blackmail and extortion were characterized by the use of threats for the objective of obtaining property from a victim in a private, bilateral transaction akin to robbery, in which the parties’ relationship neither gives the threatener a legitimate basis for pressing his demand nor obliges the addressee to hear and consider it apart from the threat. Constitutional freedom of speech did not immunize the private communication of demands and threats in such a setting. That principle has survived.

Legislatures may extend the principle beyond the extortion of property to other objects of extortion. The extortion of sexual conduct is an obvious example. See State v. Felton, supra. That is the actual charge in the present case, in which defendants are accused of coercing the victim into sexual conduct by threatening to “expose a secret and publicize an asserted fact [not further described in the indictment] which would tend to subject [her] to hatred, contempt, and ridicule.” No doubt this could be prohibited, if the issue before us were the validity of the charge rather than the validity of the statute under which it is brought. But as a statute broadens the potential objects of forbidden demands, the relationship between the parties becomes more important.

There are few relationships in which a person is constitutionally privileged to insist, by private demands, *435coupled with private threats of otherwise unrelated sanctions, that another turn over property when he is under no legal duty to do so, and the legislature took pains to exempt threats of the kind of collective action in support of group demands upon a business that have troubled courts in the strike, boycott, and picketing cases. ORS 163.275(1)(f), supra. There may be more relationships, within and outside a family, in which it is an exercise of free speech to demand that a person refrain from turning over property, or refrain from particular sexual conduct, or face consequences of the kind listed in ORS 163.275.30 The right of free expression is as important to many people in their personal and institutional relationships as it is in the narrower “civil liberties” related to politics, and nothing in article I, section 8, suggests that it is limited to the latter.

The variety of arguably privileged expression proliferates when the statute is broadened to cover all demands “to engage in conduct from which [the addressee] has a legal right to abstain, or to abstain from engaging in conduct in which he has a legal right to engage.” It is this extension that encompasses most of the hypothetical examples drawn from politics, journalism, family or academic life cited by defendants and the dissenters in the Court of Appeals to demonstrate the overbreadth of ORS 163.275. Moreover, apart from reaching such relationships, the statute makes no distinction whether the coercive demands and threats are addressed by one person to another in a private confrontation or correspondence or in a more or less public setting designed to inform and perhaps involve others in the issues posed by the demand and the potential sanction. Yet such a setting often will involve protected communication with this wider audience.

Accordingly, we cannot escape the conclusion that ORS 163.275 as written reaches areas of. constitutionally *436privileged expression and thus is invalid unless its coverage is narrowed to exclude these areas. We also conclude that in the case of this statute the needed narrowing cannot be accomplished by judicial interpretation. The preceding paragraph has identified the sources of the statute’s over-breadth in extending the principle of extortion to cover all demands to take or to refrain from nonobligatory action coupled with threats of adverse consequences in case of noncompliance, whether these demands and threats are stated privately or publicly and irrespective of the relationship of the parties themselves or with the public. The Model Penal Code undertook to narrow its version of “criminal coercion” by excluding situations in which the purpose of the threatened disclosure or official sanction is limited to compelling another to behave in ways reasonably related to the circumstances of the demand and the threat. See the affirmative defense in section 212.5, supra note 26. We do not mean to endorse that particular solution, which antedates much of the modern development of overbreadth analysis of statutes restricting speech. In any event, the Oregon Criminal Code did not follow the Model Penal Code with respect to privileged threats of disclosure but only created a narrower defense limited to threats of prosecution made to secure restitution.31

We cannot substitute a wider set of exclusions for those knowingly chosen by the drafters of ORS 163.275, even assuming that such wider exclusions rather than narrower and more precise affirmative coverage are the chosen means to confine the statute within constitutional bounds. Nor can the statute be saved simply by adding or implying a limitation that it does not apply in any situation in which its application would restrict the accused’s “right to speak, write, or print freely on any subject whatever” within the guarantee of article I, section 8. Such an implied limitation not only trades overbreadth for vagueness; it abandons *437scrutiny of the statute altogether for case-by-case adjudication, contrary to the command that no law restricting this right “shall be passed.”32 It is, therefore, in the first instance a legislative responsibility to narrow and clarify the coverage of a statute so as to eliminate most apparent applications to free speech or writing, leaving only marginal and unforeseeable instances of unconstitutional applications to judicial exclusion.

The circuit court did not err in sustaining defendants’ demurrer. The decision of the Court of Appeals to the contrary is reversed, and the case is remanded to the circuit court for entry of a judgment pursuant to ORS 135.660,

concurring.

I concur with the majority’s analysis insofar as it applies to ORS 167.275(l)(e). As will appear below, I am not convinced that ORS 167.275 is constitutionally defective in all its parts, and therefore write separately to set forth the bases for my conclusions. One reason for the problem in this case is the awkwardly written statute, ORS 163.275. Using the subpart under which the charge was brought for illustration, it appears that the statute has three elements:

*4381.The threat: A threat by the defendant to expose or publicize a secret tending to subject some person to contempt.

2. The demand: A demand by the defendant that the one threatened engage in conduct from which he or she has a legal right to abstain or the threat will be carried out.

3. The action: The one threatened is instilled with fear and engages in the demanded conduct.

The statutory element involved in this case is a threat to expose a secret or publicize a fact, whether true or false, tending to subject the threatened person to hatred, contempt or ridicule, ORS 163.275(l)(e). Unlike the statute involved in Landry v. Daley, 280 F Supp 938 (ND Ill 1968), the statute at bar condemns speech which is otherwise lawful.

The majority opinion appears to invalidate the entire statute because its reach into the right of free expression is over broad, so much so that “narrowing cannot be accomplished by judicial interpretation.” I agree that the (l)(e) statutory incursions into protected speech are so substantial that there exists no satisfactory way of judb dally severing constitutional applications of (l)(e) from unconstitutional applications. The alternative of creating constitutional boundaries on a case-by-case basis is not a realistic one, for “the contours of regulation would have to be hammered out case-by-case — and tested only by those hardy enough to risk criminal prosecution to determine the proper scope of regulation.”1

Other subdivisions of ORS 167.275 may be immune from the overbreadth conclusions that fatally fetter subsection (l)(e). Chaplinsky v. New Hampshire, 315 US 568, 62 S Ct 766, 86 L Ed 1031 (1942), suggests that other subsections of the statute may be valid, for the Supreme Court there upheld a statute which forbade “offensive, derisive or annoying word to any other person who is lawfully in any * * * public place.” Although later federal decisions *439appear to restrict the Chaplinsky holding,2 as recently as July 2, 1982, the Supreme Court stated:

“* * * ]\fQ federal rule of law restricts a State from imposing tort liability for business losses that are caused by violence and by threats of violence. * * *” National Association for the Advancement of Colored People v. Claiborne Hardware Co., 458 US 886, 102 S Ct 3409, 3428, 73 L Ed 2d 1215 (1982).

Other sections of ORS 167.275 may not be constitutionally infirm for overbreadth, or it may be possible for a court to adopt a construction of one or more of the other subsections which will save the statute.3 The extent to which Article I, section 8, permits the imposition of criminal liability under other subsections of the statute is not before us. Therefore, though I concur that (l)(e) is invalid for the reasons set forth in the majority opinion, I do not agree that the other subsections are necessarily invalid.

Tanzer, J., joins in this opinion.