with whom Mr. Justice Brennan and Mr. Justice Marshall join, dissenting.
The Court’s crabbed view of the First Amendment reflects a disturbing insensitivity to the critical role of an independent press in our society. The question whether a reporter has a constitutional right to a confidential relationship with his source is of first impression here, but the principles that should guide our decision are as basic as any to be found in the Constitution. While Mr. Justice Powell’s enigmatic concurring opinion gives some hope of a mpre flexible view in the future, the Court in these cases holds that a newsman has no First Amendment right to protect his sources when called before a grand jury. The Court thus invites state and federal authorities to undermine the historic independence of the press by attempting to annex the journalistic profession as an investigative arm of government. Not only will this decision impair performance of the press’ constitutionally protected functions, but it will, I am convinced, in the long run harm rather than help the administration of justice.
I respectfully dissent.
I
The reporter’s constitutional right to a confidential relationship with his source stems from the broad societal interest in a full and free flow of information to the public. It is this basic concern that underlies the Con*726stitution’s protection of a free press, Grosjean v. American Press Co., 297 U. S. 233, 250; New York Times Co. v. Sullivan, 376 U. S. 254, 269,1 because the guarantee is “not for the benefit of the press so much as for the benefit of all of us.” Time, Inc. v. Hill, 385 U. S. 374, 389.2
Enlightened choice by an informed citizenry is the basic ideal upon which an open society is premised,3 and a free press' is thus indispensable to a free society. Not only does the press enhance personal self-fulfillment *727by providing the people with the widest possible range of fact and opinion, but it also is an incontestable precondition of self-government. The press “has been a mighty catalyst in awakening public interest in governmental affairs, exposing corruption among public officers and employees and generally informing the citizenry of public events and occurrences . . . Estes v. Texas, 381 U. S. 532, 539; Mills v. Alabama, 384 U. S. 214, 219; Grosjean, supra, at 250. As private and public aggregations of power burgeon in size and the pressures for conformity necessarily mount, there is obviously a continuing need for an independent press to disseminate a robust variety of information and opinion through reportage, investigation, and criticism, if we are to preserve our constitutional tradition of maximizing freedom of choice by encouraging diversity of expression.
A
In keeping with this tradition, we have held that the right to publish is central to the First Amendment and basic to the existence of constitutional democracy. Grosjean, supra, at 250; New York Times, supra, at 270.
A corollary of the right to publish must be the right to gather news. The full flow of information to the public protected by the free-press guarantee would be severely curtailed if no protection whatever were afforded to the process by which news is assembled and disseminated. We have, therefore, recognized that there is a right to publish without prior governmental approval, Near v. Minnesota, 283 U. S. 697; New York Times Co. v. United States, 403 U. S. 713, a right to distribute information, see, e. g., Lovell v. Griffin, 303 U. S. 444, 452; Marsh v. Alabama, 326 U. S. 501; Martin v. City of Struthers, 319 U. S. 141; Grosjean, supra, and a right to receive printed matter, Lamont v. Postmaster General, 381 U. S. 301.
*728No less important to the news dissemination process is the gathering of information. News must not be unnecessarily cut off at its source, for without freedom to acquire information the right to publish would be im-permissibly compromised. Accordingly, a right to gather news, of some dimensions, must exist. Zemel v. Rusk, 381 U. S. 1.4 Note, The Right of the Press to Gather Information, 71 Col. L. Rev. 838 (1971). As Madison wrote: “A popular Government, without popular information, or the means of acquiring it, is but a Prologue to a Farce or a Tragedy; or, perhaps both.” 9 Writings of James Madison 103 (G. Hunt ed. 1910).
B
The right to gather news implies, in turn, a right to a confidential relationship between a reporter and his source. This proposition follows as a matter of simple logic once three factual predicates are recognized: (1) newsmen require informants to gather news; (2) confidentiality — the promise or understanding that names or certain aspects of communications will be kept off the record — is essential to the creation and maintenance of a news-gathering relationship with informants; and (3) an unbridled subpoena power — the absence of a constitutional right protecting, in any way, a confidential relationship from compulsory process — will either deter sources from divulging information or deter reporters from gathering and publishing information.
*729It is obvious that informants are necessary to the news-gathering process as we know it today. If it is to perform its constitutional mission, the press must do far more than merely print public statements or publish prepared handouts. Familiarity with the people and circumstances involved in the myriad background activities that result in the final product called "news” is vital to complete and responsible journalism, unless the press is to be a captive mouthpiece of "newsmakers.” 5
It is equally obvious that the promise of confidentiality may be a necessary prerequisite to a productive relationship between a newsman and his informants. An officeholder may fear his superior; a member of the bureaucracy, his associates; a dissident, the scorn of majority opinion. All may have information valuable to the public discourse, yet each may be willing to relate that information only in confidence to a reporter whom he trusts, either because of excessive caution or because of a reasonable fear of reprisals or censure for unorthodox *730views. The First Amendment concern must not be with the motives of any particular news source, but rather with the conditions in which informants of all shades of the spectrum may make information available through the press to the public. Cf. Talley v. California, 362 U. S. 60, 65; Bates v. Little Rock, 361 U. S. 516; NAACP v. Alabama, 357 U. S. 449.6
In Caldwell, the District Court found that “confidential relationships . . . are commonly developed, and maintained by professional journalists, and are indispensable to their work of gathering, analyzing and publishing the news.” 7 Commentators and individual reporters have repeatedly noted the importance of confidentiality.8 *731And surveys among reporters and editors indicate that the promise of nondisclosure is necessary for many types of news gathering.9
Finally, and most important, when governmental officials possess an unchecked power to compel newsmen to disclose information received in confidence, sources will clearly be deterred from giving information, and reporters will clearly be deterred from publishing it, because uncertainty about exercise of the power will lead to “self-censorship.” Smith v. California, 361 U. S. 147, 149-154; New York Times Co. v. Sullivan, 376 U. S., at 279. The uncertainty arises, of course, because the judiciary has traditionally imposed virtually no limitations on the grand jury’s broad investigatory powers. See Antell, The Modern Grand Jury: Benighted Super-government, 51 A. B. A. J. 153 (1965). See also Part II, infra.
After today's decision, the potential informant can never be sure that his identity or off-the-record communications will not subsequently be revealed through the compelled testimony of a newsman. A public-spirited person inside government, who is not implicated in any crime, will now be fearful of revealing corruption or other governmental wrongdoing, because he will now know he can subsequently be identified by use of compulsory process. The potential source must, therefore, choose between risking exposure by giving information or avoiding the risk by remaining silent.
The reporter must speculate about whether contact with a controversial source or publication of controversial material will lead to a subpoena. In the event of a *732subpoena, under today’s decision, the newsman will know that he must choose between being punished for contempt if he refuses to testify, or violating his profession’s ethics10 and impairing his resourcefulness as a reporter if he discloses confidential information.11
Again, the commonsense understanding that such deterrence will occur is buttressed by concrete evidence. The existence of deterrent effects through fear and self-censorship was impressively developed in the District Court in Caldwell12 Individual reporters13 and commentators14 have noted such effects. Surveys have verified that an unbridled subpoena power will substan*733tially impair the flow of news to the public, especially in sensitive areas involving governmental officials, financial affairs, political figures, dissidents, or minority groups that require in-depth, investigative reporting.15 And the Justice Department has recognized that “compulsory process in some circumstances may have a limiting effect on the exercise of First Amendment rights.” 16 No evidence contradicting the existence of such deterrent effects was offered at the trials or in the briefs here by the petitioner in Caldwell or by the respondents in Branz-burg and Pappas.
The impairment of the flow of news cannot, of course, be proved with scientific precision, as the Court seems to demand. Obviously, not every news-gathering relationship requires confidentiality. And it is difficult to pinpoint precisely how many relationships do require a promise or understanding of nondisclosure. But we have never before demanded that First Amendment rights rest on elaborate empirical studies demonstrating beyond any conceivable doubt that deterrent effects exist; we have never before required proof of the exact number of people potentially affected by governmental action, who would actually be dissuaded from engaging in First Amendment activity.
Rather, on the basis of common sense and available information, we have asked, often implicitly, (1) whether there was a rational connection between the cause (the governmental action) and the effect (the deterrence or *734impairment of First Amendment activity), and (2) whether the effect would occur with some regularity, i. e., would not be de minimis. See, e. g., Grosjean v. American Press Co., 297 U. S., at 244-245 ; Burstyn, Inc. v. Wilson, 343 U. S. 495, 503; Sweezy v. New Hampshire, 354 U. S. 234, 248 (plurality opinion); NAACP v. Alabama, 357 U. S., at 461-466; Smith v. California, 361 U. S., at 150-154; Bates v. Little Rock, 361 U. S., at 523-524; Talley v. California, 362 U. S., at 64-65; Shelton v. Tucker, 364 U. S. 479, 485-486; Cramp v. Board of Public Instruction, 368 U. S. 278, 286; NAACP v. Button, 371 U. S. 415, 431-438; Gibson v. Florida Legislative Investigation Committee, 372 U. S. 539, 555-557; New York Times Co. v. Sullivan, 376 U. S., at 277-278; Freedman v. Maryland, 380 U. S. 51, 59; DeGregory v. New Hampshire Attorney General, 383 U. S. 825; Elfbrandt v. Russell, 384 U. S. 11, 16-19. And, in making this determination, we have shown a special solicitude towards the “indispensable liberties” protected by the First Amendment, NAACP v. Alabama, supra, at 461; Bantam Books, Inc. v. Sullivan, 372 U. S. 58, 66, for “[f]reedoms such as these are protected not only against heavy-handed frontal attack, but also from being stifled by more subtle governmental interference.” Bates, supra, at 523.17 Once this threshold inquiry has been satisfied, we have then examined the competing interests in determining whether *735there is an unconstitutional infringement of First Amendment freedoms.
For example, in NAACP v. Alabama, supra, we found that compelled disclosure of the names of those in Alabama who belonged to the NAACP “is likely to affect adversely the ability [of the NAACP] and its members to pursue their . . . beliefs which they admittedly have the right to advocate, in that it may induce members to withdraw from the Association and dissuade others from joining it because of fear of exposure of their beliefs shown through their associations and of the consequences of this exposure.” Id., at 462-463. In Talley, supra, we held invalid a city ordinance that forbade circulation of any handbill that did not have the distributor's name on it, for there was “no doubt that such an identification requirement would tend to restrict freedom to distribute information and thereby freedom of expression.” Id., at 64. And in Burstyn, Inc., supra, we found deterrence of First Amendment activity inherent in a censor’s power to exercise unbridled discretion under an overbroad statute. Id., at 503.
Surely the analogous claim of deterrence here is as securely grounded in evidence and common sense as the claims in the cases cited above, although the Court calls the claim “speculative.” See ante, at 694. The deterrence may not occur in every confidential relationship between a reporter and his source.18 But it will cer*736tainly occur in certain types of relationships involving sensitive and controversial matters. And such relationships are vital to the free flow of information.
To require any greater burden of proof is to shirk our duty to protect values securely embedded in the Constitution. We cannot await an unequivocal — and therefore unattainable — imprimatur from empirical studies.19 We can and must accept the evidence developed in the record, and elsewhere, that overwhelmingly supports the premise that deterrence will occur with regularity in important types of news-gathering relationships.20
Thus, we cannot escape the conclusion that when neither the reporter nor his source can rely on the shield of confidentiality against unrestrained use of the grand jury’s subpoena power, valuable information will not be published and the public dialogue will inevitably be impoverished.
II
Posed against the First Amendment’s protection of the newsman’s confidential relationships in these cases is society’s interest in the use of the grand jury to ad*737minister justice fairly and effectively. The grand jury serves two important functions: “to examine into the commission of crimes” and “to stand between the prosecutor and the accused, and to determine whether the charge was founded upon credible testimony or was dictated by malice or personal ill will.” Hale v. Henkel, 201 U. S. 43, 59. And to perform these functions the grand jury must have available to it every man’s relevant evidence. See Blair v. United States, 250 U. S. 273, 281; Blackmer v. United States, 284 U. S. 421, 438.
Yet the longstanding rule making every person’s evidence available to the grand jury is not absolute. The rule has been limited by the Fifth Amendment,21 the Fourth Amendment,22 and the evidentiary privileges of the common law.23 So it was that in Blair, supra, after recognizing that the right against compulsory self-incrimination prohibited certain inquiries, the Court noted that “some confidential matters are shielded from considerations of policy, and perhaps in other cases for special reasons a witness may be excused from telling all that he knows.” Id., at 281 (emphasis supplied). And in United States v. Bryan, 339 U. S. 323, the Court observed that any exemption from the duty to testify before the grand jury “presupposes a very real interest to be protected.” Id., at 332.
Such an interest must surely be the First Amendment protection of a confidential relationship that I have discussed above in Part I. As noted there, this protection does not exist for the purely private interests of the *738newsman or his informant, nor even, at bottom, for the First Amendment interests of either partner in the news-gathering relationship.24 Rather, it functions to insure nothing less than democratic decisionmaking through the free flow of information to the public, and it serves, thereby, to honor the “profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open.” New York Times Co. v. Sullivan, 376 U. S., at 270.
In striking the proper balance between the public interest in the efficient administration of justice and the First Amendment guarantee of the fullest flow of information, we must begin with the basic proposition that because of their “delicate and vulnerable” nature, NAACP v. Button, 371 U. S., at 433, and their transcendent importance for the just functioning of our society, First Amendment rights require special safeguards.
A
This Court has erected such safeguards when government, by legislative investigation or other investigative means, has attempted to pierce the shield of privacy inherent in freedom of association.25 In no previous case have we considered the extent to which the First Amendment limits the grand jury subpoena power. But the *739Court has said that “[t]he Bill of Rights is applicable to investigations as to all forms of governmental action. Witnesses cannot be compelled to give evidence against themselves. They cannot be subjected to unreasonable search and seizure. Nor can the First Amendment freedoms of speech, press ... or political belief and association be abridged.” Watkins v. United States, 354 U. S. 178, 188. And in Sweezy v. New Hampshire it was stated: “It is particularly important that the exercise of the power of compulsory process be carefully circumscribed when the investigative process tends to impinge upon such highly sensitive areas as freedom of speech or press, freedom of political association, and freedom of communication of ideas.” 354 U. S., at 245 (plurality opinion).
The established method of “carefully” circumscribing-investigative powers is to place a heavy burden of justification on government officials when First Amendment rights are impaired. The decisions of this Court have “consistently held that only a compelling state interest in the regulation of a subject within the State’s constitutional power to regulate can justify limiting First Amendment freedoms.” NAACP v. Button, 371 U. S., at 438. And “it is an essential prerequisite to the validity of an investigation which intrudes into the area of constitutionally protected rights of speech, press, association and petition that the State convincingly show a substantial relation between the information sought and a subject of overriding and compelling state interest.” Gibson v. Florida Legislative Investigation Committee, 372 U. S., at 546 (emphasis supplied). See also DeGregory v. Attorney General of New Hampshire, 383 U. S. 825; NAACP v. Alabama, 357 U. S. 449; Sweezy, supra; Watkins, supra.
Thus, when an investigation impinges on First Amendment rights, the government must not only show that *740the inquiry is of “compelling and overriding importance” but it must also “convincingly” demonstrate that the investigation is “substantially related” to the information sought.
Governmental officials must, therefore, demonstrate that the information sought is clearly relevant to a precisely defined subject of governmental inquiry. Watkins, supra; Sweezy, supra.26 They must demonstrate that it is reasonable to think the witness in question has that information. Sweezy, supra; Gibson, supra,27 And they must show that there is not any means of obtaining the information less destructive of First Amendment liberties. Shelton v. Tucker, 364 U. S., at 488; Louisiana ex rel. Gremillion v. NAACP, 366 U. S. 293, 296-297.28
These requirements, which we have recognized in decisions involving legislative and executive investigations, serve established policies reflected in numerous First *741Amendment decisions arising in other contexts. The requirements militate against vague investigations that, like vague laws, create uncertainty and needlessly discourage First Amendment activity.29 They also insure that a legitimate governmental purpose will not be pursued by means that “broadly stifle fundamental personal liberties when the end can be more narrowly achieved.” Shelton, supra, at 488.30 As we said in Gibson, supra, “Of course, a legislative investigation — as any investigation — must proceed ‘step by step,’ . . . but step by step or in totality, an adequate foundation for inquiry must be laid before proceeding in such a manner as will substantially intrude upon and severely curtail or inhibit constitutionally protected activities or seriously interfere with similarly protected associational rights.” 372 U. S., at 557.
I believe the safeguards developed in our decisions involving governmental investigations must apply to the grand jury inquiries in these cases. Surely the function of the grand jury to aid in the enforcement of the law is no more important than the function of the legislature, and its committees, to make the law. We have long recognized the value of the role played by legislative investigations, see, e. g., United States v. Rumely, *742345 U. S. 41, 43; Barenblatt v. United States, 360 U. S. 109, 111-112, for the “power of the Congress to conduct investigations is inherent . . . [encompassing] surveys of defects in our social, economic or political system for the purpose of enabling the Congress to remedy them.” Watkins, supra, at 187. Similarly, the associational rights of private individuals, which have been the prime focus of our First Amendment decisions in the investigative sphere, are hardly more important than the First Amendment rights of mass circulation newspapers and electronic media to disseminate ideas and information, and of the general public to receive them. Moreover, the vices of vagueness and overbreadth that legislative investigations may manifest are also exhibited by grand jury inquiries, since grand jury investigations are not limited in scope to specific criminal acts, see, e. g., Wilson v. United States, 221 U. S. 361, Hendricks v. United States, 223 U. S. 178, 184, United States v. Johnson, 319 U. S. 503, and since standards of materiality and relevance are greatly relaxed. Holt v. United States, 218 U. S. 245; Costello v. United States, 350 U. S. 359. See generally Note, The Grand Jury as an Investigatory Body, 74 Harv. L. Rev. 590, 591-592 (1961).31 For, as the United States notes in its brief in Caldwell, the *743grand jury “need establish no factual basis for commencing an investigation, and can pursue rumors which further investigation may prove groundless.”
Accordingly, when a reporter is asked to appear before a grand jury and reveal confidences, I would hold that the government must (1) show that there is probable cause to believe that the newsman has information that is clearly relevant to a specific probable violation of law;32 (2) demonstrate that the information sought cannot be obtained by alternative means less destructive of First Amendment rights; and (3) demonstrate a compelling and overriding interest in the information.33
This is not to say that a grand jury could not issue a subpoena until such a showing were made, and it is not to say that a newsman would be in any way privileged to ignore any subpoena that was issued. Obviously, before the government’s burden to make such a showing were triggered, the reporter would have to move to quash the subpoena, asserting the basis on which he considered the particular relationship a confidential one.
*744B
The crux of the Court’s rejection of any newsman’s privilege is its observation that only “where news sources themselves are implicated in crime or possess information relevant to the grand jury’s task need they or the reporter be concerned about grand jury subpoenas.” See ante, at 691 (emphasis supplied). But this is a most misleading construct. For it is obviously not true that the only persons about whom reporters will be forced to testify will be those “confidential informants involved in actual criminal conduct” and those having “information suggesting illegal conduct by others.” See ante, at 691, 693. As noted above, given the grand jury’s extraordinarily broad investigative powers and the weak standards of relevance and materiality that apply during such inquiries, reporters, if they have no testimonial privilege, will be called to give information about informants who have neither committed crimes nor have information about crime. It is to avoid deterrence of such sources and thus to prevent needless injury to First Amendment values that I think the government must be required to show probable cause that the newsman has information that is clearly relevant to a specific probable violation of criminal law.34
*745Similarly, a reporter may have information from a confidential source that is “related” to the commission of crime, but the government may be able to obtain an indictment or otherwise achieve its purposes by subpoenaing persons other than the reporter. It is an obvious but important truism that when government aims have been fully served, there can be no legitimate reason to disrupt a confidential relationship between a reporter and his source. To do so would not aid the administration of justice and would only impair the flow of information to the public. Thus, it is to avoid deterrence of such sources that I think the government must show that there are no alternative means for the grand jury to obtain the information sought.
Both the “probable cause” and “alternative means” requirements would thus serve the vital function of mediating between the public interest in the administration of justice and the constitutional protection of the full flow of information. These requirements would avoid a direct conflict between these competing concerns, and they would generally provide adequate protection for newsmen. See Part III, m/m.35 No doubt the courts would be required to make some delicate judgments in working out this accommodation. But that, after all, *746is the function of courts of law. Better such judgments, however difficult, than the simplistic and stultifying absolutism adopted by the Court in denying any force to the First Amendment in these cases.36
The error in the Court’s absolute rejection of First Amendment interests in these cases seems to me to be most profound. For in the name of advancing the administration of justice, the Court’s decision, I think, will only impair the achievement of that goal. People entrusted with law enforcement responsibility, no less than private citizens, need general information relating to controversial social problems. Obviously, press reports have great value to government, even when the newsman cannot be compelled to testify before a grand jury. The sad paradox of the Court’s position is that when a grand jury may exercise an unbridled subpoena power, and sources involved in sensitive matters become fearful of disclosing information, the newsman will not only cease to be a useful grand jury witness; he will cease to investigate and publish information about issues of public import. I cannot subscribe to such an anomalous result, for, in my view, the interests protected by the First Amendment are not antagonistic to the administration of justice. Rather, they can, in the long run, only be complementary, and for that reason must be given great “breathing space.” NAACP v. Button, 371 U. S., at 433.
Ill
In deciding what protection should be given to information a reporter receives in confidence from a news source, the Court of Appeals for the Ninth Circuit affirmed the holding of the District Court that the grand *747jury power of testimonial compulsion must not be exercised in a manner likely to impair First Amendment interests “until there has been a clear showing of a compelling and overriding national interest that cannot be served by any alternative means.” Caldwell v. United States, 434 F. 2d 1081, 1086. It approved the request of respondent Caldwell for specification by the government of the “subject, direction or scope of the Grand Jury inquiry.” Id., at 1085. And it held that in the circumstances of this case Caldwell need not divulge confidential information.
I think this decision was correct. On the record before us the United States has not met the burden that I think the appropriate newsman’s privilege should require.
In affidavits before the District Court, the United States said it was investigating possible violations of 18 U. S. C. § 871 (threats against the President), 18 U. S. C. § 1751 (assassination, attempts to assassinate, conspiracy to assassinate the President), 18 U. S. C. § 231 (civil disorders), 18 U. S. C. § 2101 (interstate travel to incite a riot), 18 U. S. C. § 1341 (mail fraud and swindles) and other crimes that were not specified. But, with one exception, there has been no factual showing in this case of the probable commission of, or of attempts to commit, any crimes.37 The single exception relates to the allegation that a Black Panther Party leader, David Hilliard, violated 18 U. S. C. § 871 during the course of a speech in November 1969. But Caldwell was subpoenaed two months after an indictment was returned against Hil-liard, and that charge could not, subsequent to the indictment, be investigated by a grand jury. See In re National Window Glass Workers, 287 F. 219; United *748States v. Dardi, 330 F. 2d 316, 336.38 Furthermore, the record before us does not show that Caldwell probably had any information about the violation of any other federal criminal laws,39 or that alternative *749means of obtaining the desired information were pursued.40
In the Caldwell case, the Court of Appeals further found that Caldwell’s confidential relationship with the leaders of the Black Panther Party would be impaired if he appeared before the grand jury at all to answer questions, even though not privileged. Caldwell v. United States, 434 F. 2d, at 1088. On the particular facts before it,41 the court concluded that the very *750appearance by Caldwell before the grand jury would jeopardize his relationship with his sources, leading to a severance of the news-gathering relationship and impairment of the flow of news to the public:42
“Appellant asserted in affidavit that there is nothing to which he could testify (beyond that which he has already made public and for which, therefore, his appearance is unnecessary) that is not protected by the District Court’s order. If this is true — and the Government apparently has not believed it necessary to dispute it — appellant’s response to the subpoena would be a barren perform-*751anee — one of no benefit to the Grand Jury. To destroy appellant’s capacity as news gatherer for such a return hardly makes sense. Since the cost to the public of excusing his attendance is so slight, it may be said that there is here no public interest of real substance in competition with the First Amendment freedoms that are jeopardized.
“If any competing public interest is ever to arise in a case such as this (where First Amendment liberties are threatened by mere appearance at a Grand Jury investigation) it will be on an occasion in which the witness, armed with his privilege, can still serve a useful purpose before the Grand Jury. Considering the scope of the privilege embodied in the protective order, these occasions would seem to be unusual. It is not asking too much of the Government to show that such an occasion is presented here.” Id., at 1089.
I think this ruling was also correct in fight of the particularized circumstances of the Caldwell case. Obviously, only in very rare circumstances would a confidential relationship between a reporter and his source be so sensitive that mere appearance before the grand jury by the newsman would substantially irhpair his news-gathering function. But in this case, the reporter made out a prima facie case that the flow of news to the public would be curtailed. And he stated, without contradiction, that the only nonconfidential material about which he could testify was already printed in his newspaper articles.43 Since the United States has not attempted to *752refute this assertion, the appearance of Caldwell would, on these facts, indeed be a “barren performance.” But this aspect of the Caldwell judgment I would confine to its own facts. As the Court of Appeals appropriately observed: “[T]he rule of this case is a narrow one. . . .” Caldwell, supra, at 1090.
Accordingly, I would affirm the judgment of the Court of Appeals in No. 70-57, United States v. Caldwell.44 In the other two cases before us, No. 70-85, Branzburg v. Hayes and Meigs, and No. 70-94, In re Pappas, I would vacate the judgments and remand the cases for further proceedings not inconsistent with the views I have expressed in this opinion.
We have often described the process of informing the public as the core purpose of the constitutional guarantee of free speech and a free press. See, e. g., Stromberg v. California, 283 U. S. 359, 369; De Jonge v. Oregon, 299 U. S. 353, 365; Smith v. California, 361 U. S. 147, 153.
As I see it, a reporter’s right to protect his source is bottomed on the constitutional guarantee of a full flow of information to the public. A newsman’s personal First Amendment rights or the associational rights of the newsman and the source are subsumed under that broad societal interest protected by the First Amendment. Obviously, we are not here concerned with the parochial personal concerns of particular newsmen or informants.
“The newsman-informer relationship is different from . . . other relationships whose confidentiality is protected by statute, such as the attorney-client and physician-patient relationships. In the case of other statutory privileges, the right of nondisclosure is granted to the person making the communication in order that he will be encouraged by strong assurances of confidentiality to seek such relationships which contribute to his personal well-being. The judgment is made that the interests of society will be served when individuals consult physicians and lawyers; the public interest is thus advanced by creating a zone of privacy that the individual can control. However, in the case of the reporter-informer relationship, society’s interest is not in the welfare of the informant per se, but rather in creating conditions in which information possessed by news sources can reach public attention.” Note, 80 Yale L. J. 317, 343 (1970) (footnotes omitted) (hereinafter Yale Note).
See generally Z. Chafee, Free Speech in the United States (1941) ; A. Meikeljohn, Free Speech and Its Relation to Self-Government (1948); T. Emerson, Toward a General Theory of the First Amendment (1963).
In Zemel v. Rusk, 381 U. S. 1, we held that the Secretary of State’s denial of a passport for travel to Cuba did not violate a citizen’s First Amendment rights. The rule was justified by the “weightiest considerations of national security” and we concluded that the “right to speak and publish does not carry with it the unrestrained right to gather information.” Id., at 16-17 (emphasis supplied). The necessary implication is that some right to gather information does exist.
In Caldwell v. United States, 434 F. 2d 1081, the Government claimed that Caldwell did not have to maintain a confidential relationship with members of the Black Panther Party and provide independent reporting of their activities, since the Party and its leaders could issue statements on their own. But, as the Court of Appeals for the Ninth Circuit correctly observed:
“[I]t is not enough that Black Panther press releases and public addresses by Panther leaders may continue unabated in the wake of subpoenas such as the one here in question. It is not enough that the public’s knowledge of groups such as the Black Panthers should be confined to their deliberate public pronouncements or distant news accounts of their occasional dramatic forays into the public view.
“The need for an untrammeled press takes on special urgency in times of widespread protest and dissent. In such times the First Amendment protections exist to maintain communication with dissenting groups and to provide the public with a wide range of information about the nature of protest and heterodoxy.” Citing Associated Press v. United States, 326 U. S. 1, 20; Thornhill v. Alabama, 310 U. S. 88, 102. Id., at 1084-1085.
As we observed in Talley v. California, 362 U. S. 60, “Anonymous pamphlets, leaflets, brochures and even books have played ai important role in the progress of mankind. . . . Before the Revolutionary War colonial patriots frequently had to conceal their authorship or distribution of literature that easily could have brought down on them prosecutions by English-controlled courts. . . . Even the Federalist Papers, written in favor of the adoption of our Constitution, were published under fictitious names. It is plain that anonymity has sometimes been assumed for the most constructive purposes.” Id., at 64-65. And in Lamont v. Postmaster General, 381 U. S.-'301, we recognized the importance to First Amendment values of the right to receive information anonymously.
Application of Caldwell, 311 F. Supp. 358, 361.
See, e. g., F. Chalmers, A Gentleman of the Press: The Biography of Colonel John Bayne MacLean 74r-75 (1969); H. Klurfeld, Behind the Lines: The World of Drew Pearson 50, 52-55 (1968); A. Krock, Memoirs: Sixty Years on the Firing Line 181, 184-185 (1968); E. Larsen, First with the Truth 22-23 (1968); R. Ottley, The Lonely Warrior — The Life and Times of Robert S. Abbott 143-145 (1955); C. Sulzberger, A Long Row of Candles; Memoirs and Diaries 241 (1969).
As Walter Cronkite, a network television reporter, said in an affidavit in Caldwell: “In doing my work, I (and those who assist me) depend constantly on information, ideas, leads and opinions received in confidence. Such material is essential in digging out newsworthy facts and, equally important, in assessing the importance and analyzing the significance of public events.” App. 52.
See Guest & Stanzler, The Constitutional Argument for Newsmen Concealing Their Sources, 64 Nw. U. L. Rev. 18 (1969); V. Blasi, Press Subpoenas: An Empirical and Legal Analysis, Study Report of the Reporters’ Committee on Freedom of the Press 20-29 (hereinafter Blasi).
The American Newspaper Guild has adopted the following rule as part of the newsman’s code of ethics: “[Newspapermen shall refuse to reveal confidences or disclose sources of confidential information in court or before other judicial or investigating bodies.” G. Bird & F. Merwin, The Press and Society 592 (1971).
Obviously, if a newsman does not honor a confidence he will have difficulty establishing other confidential relationships necessary for obtaining information in the future. See Siebert & Ryniker, Press Winning Fight to Guard Sources, Editor & Publisher, Sept. 1, 1934, pp. 9, 36-37.
The court found that “compelled disclosure of information received by a journalist within the scope of . . . confidential relationships jeopardizes those relationships and thereby impairs the journalist’s ability to gather, analyze and publish the news.” Application of Caldwell, 311 F. Supp., at 361.
See n. 8, supra.
Recent commentary is nearly unanimous in urging either an absolute or qualified newsman’s privilege. See, e. g., Goldstein, Newsmen and Their Confidential Sources, New Republic, Mar. 21, 1970, pp. 13-14; Yale Note, supra, n. 2; Comment, 46 N. Y. U. L. Rev. 617 (1971); Nelson, The Newsmen’s Privilege Against Disclosure of Confidential Sources and Information, 24 Vand. L. Rev. 667 (1971); Note, The Right of the Press to Gather Information, 71 Col. L. Rev. 838 (1971); Comment, 4 U. Mich. J. L. Ref. 85 (1970); Comment, 6 Harv. Civ. Rights-Civ. Lib. L. Rev. 119 (1970); Comment, The Newsman’s Privilege: Government Investigations, Criminal Prosecutions and Private Litigation, 58 Calif. L. Rev. 1198 (1970). But see the Court’s opinion, ante, *733at 690 n. 29. And see generally articles collected in Yale Note, supra, n. 2.
Recent decisions are in conflict both as to the importance of the deterrent effects and, a jortiori, as to the existence of a constitutional right to a confidential reporter-source relationship. See the Court’s opinion, ante, at 686, and cases collected in Yale Note, at 318 nn. 6-7.
See Blasi 6-71; Guest & Stanzler, supra, n. 9, at 43-50.
Department of Justice Memo. No. 692 (Sept. 2, 1970).
Although, as the Court points out, we have held that the press is not free from the requirements of the National Labor Relations Act, the Fair Labor Standards Act, the antitrust laws, or nondiscriminatory taxation, ante, at 683, these decisions were concerned “only with restraints on certain business or commercial practices” of the press. Citizen Publishing Co. v. United States, 394 U. S. 131, 139. And due weight was given to First Amendment interests. For example, “The First Amendment, far from providing an argument against application of the Sherman Act . . . provides powerful reasons to the contrary.” Associated Press v. United States, 326 U. S., at 20.
The fact that some informants will not be deterred from giving information by the prospect of the unbridled exercise of the subpoena power only means that there will not always be a conflict between the grand jury’s inquiry and the protection of First Amendment activities. But even if the percentage of such informants is relatively large compared to the total “universe” of potential informants, there will remain a large number of people in “absolute” terms who will be deterred, and the flow of news through mass circulation newspapers and electronic media will inevitably be impaired.
Empirical studies, after all, can only provide facts. It is the duty of courts to give legal significance to facts; and it is the special duty of this Court to understand the constitutional significance of facts. We must often proceed in a state of less than perfect knowledge, either because the facts are murky or the methodology used in obtaining the facts is open to question. It is then that we must look to the Constitution for the values that inform our presumptions. And the importance to our society of the full flow of information to the public has buttressed this Court’s historic presumption in favor of First Amendment values.
See, e. g., the uncontradicted evidence presented in affidavits from newsmen in Caldwell, Appendix to No. 70-57, pp. 22-61 (statements from Gerald Fraser, Thomas Johnson, John Kifner, Timothy Knight, Nicholas Proffitt, Anthony Ripley, Wallace Turner, Gilbert Noble, Anthony Lukas, Martin Arnold, David Burnham, Jon Lowell, Frank Morgan, Min Yee, Walter Cronkite, Eric Sevareid, Mike Wallace, Dan Rather, Marvin Kalb).
See Blau v. United States, 340 U. S. 159; Quinn v. United States, 349 U. S. 155; Curdo v. United States, 354 U. S. 118; Malloy v. Hogan, 378 U. S. 1.
See Silverthorne Lumber Co. v. United States, 251 U. S. 385.
See Committee on Rules of Practice and Procedure of Judicial Conference of the United States, Revised Draft of Proposed Rules of Evidence for the United States Courts and Magistrates (1971); 8 J. Wigmore, Evidence §§2290-2391 (McNaughton rev. 1961).
Although there is a longstanding presumption against creation of common-law testimonial privileges, United States v. Bryan, 339 U. S. 323, these privileges are grounded in an “individual interest which has been found ... to outweigh the public interest in the search for truth” rather than in the broad public concerns that inform the First Amendment. Id., at 331.
The protection of information from compelled disclosure for broad purposes of' public policy has been recognized in decisions involving police informers, see Roviaro v. United States, 353 U. S. 53, United States v. Ventresca, 380 U. S. 102, 108, Aguilar v. Texas, 378 U. S. 108, 114, McCray v. Illinois, 386 U. S. 300, and military and state secrets, United States v. Reynolds, 345 U. S. 1.
As we said in Watkins v. United States, 354 U. S. 178, “[W]hen First Amendment rights are threatened, the delegation of power to the [legislative] committee must be clearly revealed in its charter.” “It is the responsibility of the Congress ... to insure that compulsory process is used only in furtherance of a legislative purpose. That requires that the instructions to an investigating committee spell out the group’s jurisdiction and purpose with sufficient particularity. . . . The more vague the committee’s charter is, the greater becomes the possibility that the committee’s specific actions are not in conformity with the will of the parent House of Congress.” Id., at 198, 201.
We noted in Sweezy v. New Hampshire, 354 U. S. 234:
“The State Supreme Court itself recognized that there was a weakness in its conclusion that the menace of forcible overthrow of the government justified sacrificing constitutional rights. There was a missing link in the chain of reasoning. The syllogism was not complete. There was nothing to connect the questioning of petitioner with this fundamental interest of the State.” Id., at 251 (emphasis supplied).
See generally Note, Less Drastic Means and the First Amendment, 78 Yale L. J. 464 (1969).
See Watkins, supra, at 208-209. See generally Baggett v. Bullitt, 377 U. S. 360, 372; Speiser v. Randall, 357 U. S. 513, 526; Ashton v. Kentucky, 384 U. S. 195, 200-201; Dombrowski v. Pfister, 380 U. S. 479, 486; Smith v. California, 361 U. S., at 150-152; Winters v. New York, 333 U. S. 507; Stromberg v. California, 283 U. S., at 369. See also Note, The Chilling Effect in Constitutional Law, 69 Col. L. Rev. 808 (1969).
See generally Zwickler v. Koota, 389 U. S. 241, 249-250, and cases cited therein; Coates v. Cincinnati, 402 U. S. 611, 616; Cantwell v. Connecticut, 310 U. S. 296, 307; De Jonge v. Oregon, 299 U. S., at 364-365; Schneider v. State, 308 U. S. 147, 164; Cox v. Louisiana, 379 U. S. 559, 562-564. Cf. NAACP v. Button, 371 U. S. 415, 438. See also Note, The First Amendment Overbreadth Doctrine, 83 Harv. L. Rev. 844 (1970).
In addition, witnesses customarily are not allowed to object to questions on the grounds of materiality or relevance, since the scope of the grand jury inquiry is deemed to be of no concern to the witness. Carter v. United States, 417 F. 2d 384, cert. denied, 399 U. S. 935. Nor is counsel permitted to be present to aid a witness. See In re Groban, 352 U. S. 330.
See generally Younger, The Grand Jury Under Attack, pt. 3, 46 J. Crim. L. C. & P. S. 214 (1955); Recent Cases, 104 U. Pa. L. Rev. 429 (1955); Watts, Grand Jury: Sleeping Watchdog or Expensive Antique, 37 N. C. L. Rev. 290 (1959); Whyte, Is the Grand Jury Necessary?, 45 Va. L. Rev. 461 (1959); Note, 2 Col. J. Law & Soc. Prob. 47, 58 (1966); Antell, The Modern Grand Jury: Benighted Supergovernment, 51 A. B. A. J. 153 (1965); Orfield, The Federal Grand Jury, 22 F. R. D. 343.
The standard of proof employed by most grand juries, federal and State, is simply “probable cause” to believe that the accused has committed a crime. See Note, 1963 Wash. U. L. Q. 102; L. Hall et al., Modern Criminal Procedure 793-794 (1969). Generally speaking, it is extremely difficult to challenge indictments on the ground that they are not supported by adequate or competent evidence. Cf. Costello v. United States, 350 U. S. 359; Beck v. Washington, 369 U. S. 541.
Cf. Garland v. Torre, 259 F. 2d 545. The Court of Appeals for the Second Circuit declined to provide a testimonial privilege to a newsman called to testify at a civil trial. But the court recognized a newsman’s First Amendment right to a confidential relationship with his source and concluded: “It is to be noted that we are not dealing here with the use of the judicial process to force a wholesale disclosure of a newspaper’s confidential sources of news, nor with a case where the identity of the news source is of doubtful relevance or materiality. . . . The question asked . . . went to the heart of the plaintiff’s claim.” Id., at 549-550 (citations omitted).
If this requirement is not met, then the government will basically be allowed to undertake a “fishing expedition” at the expense of the press. Such general, exploratory investigations will be most damaging to confidential news-gathering relationships, since they will create great uncertainty in both reporters and their sources. The Court sanctions such explorations, by refusing to apply a meaningful “probable cause” requirement. See ante, at 701-702. As the Court states, a grand jury investigation “may be triggered by tips, rumors, evidence proffered by the prosecutor, or the personal knowledge of the grand jurors.” Ante, at 701. It thereby invites government to try to annex the press as an investigative arm, since any time government wants to probe the relationships between the *745newsman and his source, it can, on virtually any pretext, convene a grand jury and compel the journalist to testify..
The Court fails to recognize that under the guise of “investigating crime” vindictive prosecutors can, using the broad powers of the grand jury which are, in effect, immune from judicial supervision, explore the newsman’s sources at will, with no serious law enforcement purpose. The secrecy of grand jury proceedings affords little consolation to a news source; the prosecutor obviously will, in most cases, have knowledge of testimony given by grand jury witnesses.
We need not, therefore, reach the question of whether government’s interest in these cases is “overriding and compelling.” I do not, however, believe, as the Court does, that all grand jury investigations automatically would override the newsman’s testimonial privilege.
The disclaimers in Mr. Justice Powell’s concurring opinion leave room for the hope that in some future case the Court may take a less absolute position in this area.
See Blasi 61 et seq.
After Caldwell was first subpoenaed to appear before the grand jury, the Government did undertake, by affidavits, to “set forth facts indicating the general nature of the grand jury’s investigation [and] witness Earl Caldwell’s possession of information relevant to this general inquiry.” In detailing the basis for the belief that a crime had probably been committed, the Government simply asserted that certain actions had previously been taken by other grand juries, and by Government counsel, with respect to certain members of the Black Panther Party (i. e., immunity grants for certain Black Panthers were sought; the Government moved to compel party members to testify before grand juries; and contempt citations were sought when party members refused to testify). No facts were asserted suggesting the actual commission of crime. The exception, as noted, involved David Hilliard’s speech and its republication in the party newspaper, the Black Panther, for which Hilliard had been indicted before Caldwell was subpoenaed.
In its affidavits, the Government placed primary rebanee on certain articles published by Caldwell in the New York Times during 1969 (on June 15, July 20, July 22, July 27, and Dec. 14). On Dec. 14, 1969, Caldwell wrote:
“ 'We are special,’ Mr. Hilliard said recently 'We advocate the very direct overthrow of the Government by way of force and violence. By picking up guns and moving against it because we recognize it as being oppressive and in recognizing that we know that the only solution to it is armed struggle.’
“In their role as the vanguard in a revolutionary struggle, the Panthers have picked up guns.
“Last week two of their leaders were killed during the police raid on one of their offices in Chicago. And in Los Angeles a few days earlier, three officers and three Panthers were wounded in a similar shooting incident. In these and in some other raids, the police have found caches of weapons, including high-powered rifles.” App. in No. 70-57, p. 13.
In my view, this should be read as indicating that Caldwell had interviewed Panther leaders. It does not indicate that he probably had knowledge of the crimes being investigated by the Government. And, to repeat, to the extent it does relate to Hilliard’s threat, an *749indictment had already been brought in that matter. The other articles merely demonstrate that Black Panther Party leaders had told Caldwell their ideological beliefs — beliefs that were readily available to the Government through other sources, like the party newspaper.
The Government did not attempt to show that means less impinging upon First Amendment interests had been pursued.
In an affidavit filed with the District Court, Caldwell stated:
“I began covering and writing articles about the Black Panthers almost from the time of their inception, and I myself found that in those first months . . . they were very brief and reluctant to discuss any substantive matter with me. However, as they realized I could be trusted and that my sole purpose was to collect my information and present it objectively in the newspaper and that I had no other motive, I found that not only were the party leaders available for in-depth interviews but also the rank and file members were cooperative in aiding me in the newspaper stories that I wanted to do. During the time that I have been covering the party, I have noticed other newspapermen representing legitimate organizations in the news media being turned away because they were not known and trusted by the party leadership.
“As a result of the relationship that I have developed, I have been able to write lengthy stories about the Panthers that have appeared in The New York Times and have been of such a nature that other reporters who have not known the Panthers have not been able to write. Many of these stories have appeared in up to 50 or 60 other newspapers around the country.
“The Black Panther Party’s method of operation with regard to members of the press is significantly different from that of other organizations. For instance, press credentials are not recognized as being of any significance. In addition, interviews are not normally designated as being ‘backgrounders’ or ‘off the record’ or ‘for *750publication’ or 'on the record.’ Because no substantive interviews are given until a relationship of trust and confidence is developed between the Black Panther Party members and a reporter, statements are rarely made to such reporters on an expressed ‘on’ or ‘off’ the record basis. Instead, an understanding is developed over a period of time between the Black Panther Party members and the reporter as to matters which the Black Panther Party wishes to disclose for publications and those matters which are given in confidence. . . . Indeed, if I am forced to appear in secret grand jury proceedings, my appearance alone would be interpreted by the Black Panthers and other dissident groups as a possible disclosure of confidences and trusts and would similarly destroy my effectiveness as a newspaperman.”
The Government did not contradict this affidavit.
“Militant groups might very understandably fear that, under the pressure of examination before a Grand Jury, the witness may fail to protect their confidences .... The Government characterizes this anticipated loss of communication as Black Panther reprisal.... But it is not an extortionate threat we face. It is human reaction as reasonable to expect as that a client will leave his lawyer when his confidence is shaken. ... As the Government points out, loss of such a sensitive news source can also result from its reaction to indiscreet or unfavorable reporting or from a reporter’s association with Government agents or persons disapproved of by the news source. Loss in such a case, however, results from an exercise of the choice and prerogative of a free press. It is not the result of Government compulsion.” Caldwell v. United States, 434 F 2d at 1088.
Caldwell stated in his affidavit filed with the District Court, see n. 40, supra:
“It would be virtually impossible for me to recall whether any particular matter disclosed to me by members of the Black Panther Party since January 1, 1969, was based on an understanding that it would or would not be confidential. Generally, those matters which were made on a nonconfidential or ‘for publication’ basis have been *752published in articles I have written in The New York Times; conversely, any matters which I have not thus far disclosed in published articles would have been given to me based on the understanding that they were confidential and would not be published.”
The District Court reserved jurisdiction to modify its order on a showing of a governmental interest which cannot be served by means other than Caldwell’s grand jury testimony. The Government would thus have further opportunity in that court to meet the burden that, I think, protection of First Amendment rights requires.