dissenting.
The central development that emerges from the aftermath of Roth v. United States, 354 U. S. 476, is that no stable approach to the obscenity problem has yet been devised by this Court. Two Justices believe that the First and Fourteenth Amendments absolutely protect obscene and nonobscene material alike. Another Justice believes that neither the States nor the Federal Government may suppress any material save for “hard-core pornography.” Roth in 1957 stressed prurience and *456utter lack of redeeming social importance;1 as Roth has been expounded in this case, in Ginzburg v. United States, post, p. 463, and in Mishkin v. New York, post, p. 502, it has undergone significant transformation. The concept of “pandering,” emphasized by the separate opinion of The Chief Justice in Roth, now emerges as an uncertain gloss or interpretive aid, and the further requisite of “patent offensiveness” has been made explicit as a result of intervening decisions. Given this tangled state of affairs, I feel free to adhere to the principles first set forth in my separate opinion in Roth, 354 U. S., at 496, which I continue to believe represent the soundest constitutional solution to this intractable problem.
My premise is that in the area of obscenity the Constitution does not bind the States and the Federal Government in precisely the same fashion. This approach is plainly consistent with the language of the First and Fourteenth Amendments and, in my opinion, more responsive to the proper functioning of a federal system of government in this area. See my opinion in Roth, 354 U. S., at 505-506. I believe it is also consistent with past decisions of this Court. Although some 40 years have passed since the Court first indicated that the Fourteenth Amendment protects “free speech,” see Gitlow v. New York, 268 U. S. 652; Fiske v. Kansas, 274 U. S. 380, the decisions have never declared that every utterance the Federal Government may not reach or every regulatory scheme it may not enact is also beyond the power of the State. The very criteria used in opinions to delimit the protection of free speech — the gravity of the evil being regulated, see Schneider v. State, 308 U. S. 147; how “clear and present” is the danger, Schenck v. *457United States, 249 U. S. 47, 52 (Holmes, J.); the magnitude of “such invasion of free speech as is necessary to avoid the danger,” United States v. Dennis, 183 F. 2d 201, 212 (L. Hand, J.) — may and do depend on the particular context in which power is exercised. When, for example, the Court in Beauharnais v. Illinois, 343 U. S. 250, upheld a criminal group-libel law because of the “social interest in order and morality,” 343 U. S., at 257, it was acknowledging the responsibility and capacity of the States in such public-welfare matters and not committing itself to uphold any similar federal statute applying to such communications as Congress might otherwise regulate under the commerce power. See also Kovacs v. Cooper, 336 U. S. 77.
Federal suppression of allegedly obscene matter should, in my view, be constitutionally limited to that often described as “hard-core pornography.” To be sure, that rubric is not a self-executing standard, but it does describe something that most judges and others will “know . . . when [they] see it” (Stewart, J., in Jacobellis v. Ohio, 378 U. S. 184, 197) and that leaves the smallest room.for disagreement between those of varying tastes. To me it is plain, for instance, that Fanny Hill does not fall within this class and could not be barred from the federal mails. If further articulation is meaningful, I would characterize as “hard-core” that prurient material that is patently offensive or whose indecency is self-demonstrating and I would describe it substantially as does Mr. Justice Stewart's opinion in Ginzburg, post, p. 499. The Federal Government may be conceded a limited interest in excluding from the mails such gross pornography, almost universally condemned in this country.2 But I believe the dangers of national *458censorship and the existence of primary responsibility at the state level amply justify drawing the line at this point.
State obscenity laws present problems of quite a different order. The varying conditions across the country, the range of views on the need and reasons for curbing obscenity, and the traditions of local self-government in matters of public welfare all favor a far more flexible attitude in defining the bounds for the States. From my standpoint, the Fourteenth Amendment requires of a State only that it apply criteria rationally related to the accepted notion of obscenity and that it reach results not wholly out of step with current American standards. As to criteria, it should be adequate if the court or jury considers such elements as offensiveness, pruriency, social value, and the like. The latitude which I believe the States deserve cautions against any federally imposed formula listing the exclusive ingredients of obscenity and fixing their proportions. This approach concededly lacks precision, but imprecision is characteristic of mediating constitutional standards; 3 voluntariness of a confession, clear and present danger, and probable cause are only the most ready illustrations. In time and with more litigated examples, predictability increases, but there is no shortcut to satisfactory solutions in this field, and there is no advantage in supposing otherwise.
I believe the tests set out in the prevailing opinion, judged by their application in this case, offer only an *459illusion of certainty and risk confusion and prejudice. The opinion declares that a book cannot be banned unless it is “utterly without redeeming social value” (ante, p. 418). To establish social value in the present case, a number of acknowledged experts in the field of literature testified that Fanny Hill held a respectable place in serious writing, and unless such largely uncontradicted testimony is accepted as decisive it is very hard to see that the “utterly without redeeming social value” test has any meaning at all. Yet the prevailing opinion, while denying that social value may be “weighed against” or “canceled by” prurience or offensiveness (ante, p. 419), terminates this case unwilling to give a conclusive decision on the status of Fanny Hill under the Constitution.4 Apparently, the Court believes that the social value of the book may be negated if proof of pandering is present. Using this inherently vague “pandering” notion to offset “social value” wipes out any certainty the latter term might be given by reliance on experts, and admits into the case highly prejudicial evidence without appropriate restrictions. See my dissenting opinion in Ginzburg, post, p. 493. I think it more satisfactory to acknowledge that on this record the book has been shown to have some quantum of social value, that it may at the same time be deemed offensive and salacious, and that the State’s decision to weigh these elements and to ban this particular work does not exceed constitutional limits.
A final aspect of the obscenity problem is the role this Court is to play in administering its standards, a matter *460that engendered justified concern at the oral argument of the cases now decided. Short of saying that no material relating to sex may be banned, or that all of it may be, I do not see how this Court can escape the task of reviewing obscenity decisions on a case-by-case basis. The views of literary or other experts could be made controlling, but those experts had their say in Fanny Hill and apparently the majority is no more willing than I to say that Massachusetts must abide by their verdict. Yet I venture to say that the Court’s burden of decision would be ameliorated under the constitutional principles that I have advocated. “Hard-core pornography” for judging federal cases is one of the more tangible concepts in the field. As to the States, the due latitude my approach would leave them ensures that only the unusual case would require plenary review and correction by this Court.
There is plenty of room, I know, for disagreement in this area of constitutional law. Some will think that what I propose may encourage States to go too far in this field. Others will consider that the Court’s present course unduly restricts state experimentation with the still elusive problem of obscenity. For myself, I believe it is the part of wisdom for those of us who happen currently to possess the “final word” to leave room for such experimentation, which indeed is the underlying genius of our federal system.
On the premises set forth in this opinion, supplementing what I have earlier said in my opinions in Roth, supra, Manual Enterprises, Inc. v. Day, 370 U. S. 478, and Jacobellis v. Ohio, 378 U. S., at 203, I would affirm the judgment of the Massachusetts Supreme Judicial Court.
Given my view of the applicable constitutional standards, I find no occasion to consider the place of “redeeming social importance” in the majority opinion in Roth, an issue which further divides the present Court.
This interest may be viewed from different- angles. Compelling the Post Office to aid actively in disseminating this most obnoxious material may simply appear too offensive in itself. Or, *458more concretely, use of the mails may facilitate or insulate distribution so greatly that federal inaction amounts to thwarting state regulation.
The deterrent effect of vagueness for that critical class of books near the law’s borderline could in the past be ameliorated by devices like the Massachusetts in rern procedure used in this case. Of course, the Court’s newly adopted “panderer” test, turning as it does on the motives and actions of the particular defendant, seriously undercuts the effort to give any seller a yes or no answer on a book in advance of his own criminal prosecution.
As I understand the prevailing opinion, its rationale is that the state court may not condemn Fanny Hill as obscene after finding the book to have a modicum of social value; the opinion does note that proof of pandering “might justify the conclusion” that the book wholly lacks social value (ante, p. 420). Given its premise for reversal, the opinion has “no occasion to assess” for itself the pruri-ency, offensiveness, or lack of social value of the book (ante, p. 420).