dissenting.
Several years ago, Mr. Justice Stewart reminded us that “[t]he Constitution guarantees ... a society of free choice. Such a society presupposes the capacity of its members to choose.” Ginsberg v. New York, 390 U. S. 629, 649 (1968) (concurring in result). I would have thought that the capacity to choose necessarily depends upon knowledge that there is a choice to be made. But today the Court reaches the curious result that one can choose to relinquish a constitutional right— the right to be free of unreasonable searches — without knowing that he has the alternative of refusing to accede to a police request to search.1 I cannot agree, and therefore dissent.
*278I
1 believe that the Court misstates the true issue in this case. That issue is not, as the Court suggests, whether the police overbore Alcala’s will in eliciting his consent, but rather, whether a simple statement of assent to search, without more,2 should be sufficient to permit the police to search and thus act as a relinquishment of Alcala’s constitutional right to exclude the police.3 This Court has always scrutinized with great care claims that a person has forgone the opportunity to assert constitutional rights. See, e. g., Fuentes v. Shevin, 407 U. S. 67 (1972); D. H. Overmyer Co. v. Frick Co., 405 U. S. 174 (1972); Boykin v. Alabama, 395 U. S. 238 (1969); Carnley v. Cochran, 369 U. S. 506 (1962). I see no reason to give the claim that a person consented to a search any less rigorous scrutiny. Every case in this Court involving this kind of search has heretofore spoken *279of consent as a waiver.4 See, e. g., Amos v. United States, 255 TJ. S. 313, 317 (1921); Zap v. United States, 328 U. S. 624, 628 (1946); Johnson v. United States, 333 U. S. 10, 13 (1948).5 Perhaps one skilled in lin*280guistics or epistemology can disregard those comments, but I find them hard to ignore.
To begin, it is important to understand that the opinion of the Court is misleading in its treatment of the issue here in three ways. First, it derives its criterion for determining when a verbal statement of assent to search operates as a relinquishment of a person’s right to preclude entry from a justification of consent searches that is inconsistent with our treatment in earlier cases of exceptions to the requirements of the Fourth Amendment, and that is not responsive to the unique nature of the consent-search exception. Second, it applies a standard of voluntariness that was developed in a very different context, where the standard was based on policies different from those involved in this case. Third, it mischaracterizes our prior cases involving consent searches.
A
The Court assumes that the issue in this case is: what are the standards by which courts are to determine that consent is voluntarily given? It then imports into the law of search and seizure standards developed to decide entirely different questions about coerced confessions.6
The Fifth Amendment, in terms, provides that no person “shall be compelled in any criminal case to be a witness against himself.” Nor is the interest protected by the Due Process Clause of the Fourteenth Amendment any different. The inquiry in a case where a confession is challenged as having been elicited in an unconstitutional manner is, therefore, whether the behavior *281of the police amounted to compulsion of the defendant.7 Because of the nature of the right to be free of compulsion, it would be pointless to ask whether a defendant knew of it before he made a statement; no sane person would knowingly relinquish a right to be free of compulsion. Thus, the questions of compulsion and of violation of the right itself are inextricably intertwined. The cases involving coerced confessions, therefore, pass over the question of knowledge of that right as irrelevant, and turn directly to the question of compulsion.
Miranda v. Arizona, 384 U. S. 436 (1966), confirms this analysis. There the Court held that certain warnings must be given to suspects prior to their interrogation so that the inherently coercive nature of in-custody questioning would be diminished by the suspect’s knowledge that he could remain silent. But, although those warnings, of course, convey information about various rights of the accused, the information is intended only to protect the suspect against acceding to the other coercive aspects of police interrogation. While we would not ordinarily think that a suspect could waive his right to be free of coercion, for example, we do permit suspects to waive the rights they are informed of by police warnings, on the belief that such information in itself sufficiently decreases the chance that a statement would be elicited by compulsion. Id., at 475-476. Thus, nothing the defendant did in the cases involving coerced confessions was taken to operate as a relinquishment of his rights; certainly the fact that the defendant made *282a statement was never taken to be a relinquishment of the right to be free of coercion.8
B
In contrast, this case deals not with “coercion,” but with “consent,” a subtly different concept to which different standards have been applied in the past. Freedom from coercion is a substantive right, guaranteed by the Fifth and Fourteenth Amendments. Consent, however, is a mechanism by which substantive requirements, otherwise applicable, are avoided. In the context of the Fourth Amendment, the relevant substantive requirements are that searches be conducted only after evidence justifying them has been submitted to an impartial magistrate for a determination of probable cause. There are, of course, exceptions to these requirements based on a variety of exigent circumstances that make it impractical to invalidate a search simply because the police failed to get a warrant.9 But none of the exceptions *283relating to the overriding needs of law enforcement are applicable when a search is justified solely by consent. On the contrary, the needs of law enforcement are significantly more attenuated, for probable cause to search may be lacking but a search permitted if the subject’s consent has been obtained. Thus, consent searches are permitted, not because such an exception to the requirements of probable cause and warrant is essential to proper law enforcement, but because we permit our citizens to choose whether or not they wish to exercise their constitutional rights. Our prior decisions simply do not support the view that a meaningful choice has been made solely because no coercion was brought to bear on the subject.
For example, in Bumper v. North Carolina, 391 U. S. 543 (1968), four law enforcement officers went to the home of Bumper’s grandmother. They announced that they had a search warrant, and she permitted them to enter. Subsequently, the prosecutor chose not to rely on the warrant, but attempted to justify the search by the woman’s consent. We held that consent could not be established “by showing no more than acquiescence to a claim of lawful authority,” id., at 548-549. We did not there inquire into all the circumstances, but focused on a single fact, the claim of authority, even though the grandmother testified that no threats were made. Id., at 547 n. 8. It may be that, on the facts of that case, her consent was under all the circumstances involuntary, but it is plain that we did not apply the test adopted by the Court today. And, whatever the posture of the case when it reached this Court, it could *284not be said that the police in Bumper acted in a threatening or coercive manner, for they did have the warrant they said they had; the decision not to rely on it was made long after the search, when the case came into court.10
That case makes it clear that police officers may not courteously order the subject of a search simply to stand aside while the officers carry out a search they have settled on. Yet there would be no coercion or brutality in giving that order. No interests that the Court today recognizes would be damaged in such a search. Thus, all the police must do is conduct what will inevitably be a charade of asking for consent. If they display any firmness at all, a verbal expression of assent will undoubtedly be forthcoming. I cannot believe that the protections of the Constitution mean so little.
II
My approach to the case is straightforward and, to me, obviously required by the notion of consent as a relinquishment of Fourth Amendment rights. I am at a loss to understand why consent “cannot be taken literally to mean a 'knowing’ choice.” Ante, at 224. In fact, I have difficulty in comprehending how a decision made without knowledge of available alternatives can be treated as a choice at all.
If consent to search means that a person has chosen to forgo his right to exclude the police from the place they seek to search, it follows that his consent cannot *285be considered a meaningful choice unless he knew that he could in fact exclude the police. The Court appears, however, to reject even the modest proposition that, if the subject of a search convinces the trier of fact that he did not know of his right to refuse assent to a police request for permission to search, the search must be held unconstitutional. For it says only that “knowledge of the right to refuse consent is one factor to be taken into account.” Ante, at 227. I find this incomprehensible. I can think of no other situation in which we would say that a person agreed to some course of action if he convinced us that he did not know that there was some other course he might have pursued. I would therefore hold, at a minimum, that the prosecution may not rely on a purported consent to search if the subject of the search did not know that he could refuse to give consent. That, I think, is the import of Bumper v. North Carolina, supra. Where the police claim authority to search yet in fact lack such authority, the subject does not know that he may permissibly refuse them entry, and it is this lack of knowledge that invalidates the consent.
If one accepts this view, the question then is a simple one: must the Government show that the subject knew of his rights, or must the subject show that he lacked such knowledge?
I think that any fair allocation of the burden would require that it be placed on the prosecution. On- this question, the Court indulges in what might be called the “straw man” method of adjudication. The Court responds to this suggestion by overinflating the burden. And, when it is suggested that the prosecution’s burden of proof could be easily satisfied if the police informed the subject of his rights, the Court responds by refusing to require the police to make a “detailed” inquiry. Ante, at 245. If the Court candidly faced the real *286question of allocating the burden of proof, neither of these maneuvers would be available to it.
If the burden is placed on the defendant, all the subject can do is to testify that he did not know of his rights. And I doubt that many trial judges will find for the defendant simply on the basis of that testimony. Precisely because the evidence is very hard to come by, courts have traditionally been reluctant to require a party to prove negatives such as the lack of knowledge. See, e. g., 9 J. Wigmore, Evidence 274 (3d ed. 1940); F. James, Civil Procedure § 7.8 (1965); E. Morgan, Some Problems of Proof Under the Anglo-American System of Litigation 75-76 (1956).
In contrast, there are several ways by which the subject’s knowledge of his rights may be shown. The subject may affirmatively demonstrate such knowledge by his responses at the time the search took place, as in United States v. Curíale, 414 F. 2d 744 (CA2 1969). Where, as in this case, the person giving consent is someone other than the defendant, the prosecution may require him to testify under oath. Denials of knowledge may be disproved by establishing that the subject had, in the recent past, demonstrated his knowledge of his rights, for example, by refusing entry when it was requested by the police. The prior experience or training of the subject might in some cases support an inference that he knew of his right to exclude the police.
The burden on the prosecutor would disappear, of course, if the police, at the time they requested consent to search, also told the subject that he had a right to refuse consent and that his decision to refuse would be respected. The Court’s assertions to the contrary notwithstanding, there is nothing impractical about this method of satisfying the prosecution’s burden of proof.11 *287It must be emphasized that the decision about informing the subject of his rights would lie with the officers seeking consent. If they believed that providing such information would impede their investigation, they might simply ask for consent, taking the risk that at some later date the prosecutor would be unable to prove that the subject knew of his rights or that some other basis for the search existed.
The Court contends that if an officer paused to inform the subject of his rights, the informality of the exchange would be destroyed. I doubt that a simple statement by an officer of an individual’s right to refuse consent would do much to alter the informality of the exchange, except to alert the subject to a fact that he surely is entitled to know. It is not without significance that for many years the agents of the Federal Bureau of Investigation have routinely informed subjects of their right to refuse consent, when they request consent to search. Note, Consent Searches: A Reappraisal After Miranda v. Arizona, 67 Col. L. Rev. 130, 143 n. 75 (1967) (citing letter from J. Edgar Hoover). The reported cases in which the police have informed subjects of their right to refuse consent show, also, that the information can be given without disrupting the casual flow of events. See, e. g., United States v. Miller, 395 F. 2d 116 (CA7 1968). What evidence there is, then, rather strongly suggests that nothing disastrous would happen if the police, before requesting consent, informed the subject that he had *288a right to refuse consent and that his refusal would be respected.12
I must conclude, with some reluctance, that when the Court speaks of practicality, what it really is talking of is the continued ability of the police to capitalize on the ignorance of citizens so as to accomplish by subterfuge what they could not achieve by relying only on the knowing relinquishment of constitutional rights. Of course it would be “practical” for the police to ignore the commands of the Fourth Amendment, if by practicality we mean that more criminals will be apprehended, even though the constitutional rights of innocent people also go by the board. But such a practical advantage is achieved only at the cost of permitting the police to disregard the limitations that the Constitution places on their behavior, a cost that a constitutional democracy cannot long absorb.
I find nothing in the opinion of the Court to dispel my belief that, in such a case, as the Court of Appeals for *289the Ninth Circuit said, “[u]nder many circumstances a reasonable person might read an officer’s ‘May I’ as the courteous expression of a demand backed by force of law.” 448 F. 2d, at 701. Most cases, in my view, are akin to Bumper v. North Carolina, 391 U. S. 543 (1968) : consent is ordinarily given as acquiescence in an implicit claim of authority to search. Permitting searches in such circumstances, without any assurance at all that the subject of the search knew that, by his consent, he was relinquishing his constitutional rights, is something that I cannot believe is sanctioned by the Constitution.
W-i I — H 1 — 1
The proper resolution of this case turns, I believe, on a realistic assessment of the nature of the interchange between citizens and the police, and of the practical import of allocating the burden of proof in one way rather than another. The Court seeks to escape such assessments by escalating its rhetoric to unwarranted heights, but no matter how forceful the adjectives the Court uses, it cannot avoid being judged by how well its image of these interchanges accords with reality. Although the Court says without real elaboration that it “cannot agree,” ante, at 248, the holding today confines the protection of the Fourth Amendment against searches conducted without probable cause to the sophisticated, the knowledgeable, and, I might add, the few.13 In the final analysis, the Court now sanctions a game of blindman’s buff, in which the police always have the upper hand, for the sake of nothing more than the convenience of *290the police. But the guarantees of the Fourth Amendment were never intended to shrink before such an ephemeral and changeable interest. The Framers of the Fourth Amendment struck the balance against this sort of convenience and in favor of certain basic civil rights. It is not for this Court to restrike that balance because of its own views of the needs of law enforcement officers. I fear that that is the effect of the Court's decision today.
It is regrettable that the obsession with validating searches like that conducted in this case, so evident in the Court’s hyperbole, has obscured the Court’s vision of how the Fourth Amendment was designed to govern the relationship between police and citizen in our society. I believe that experience and careful reflection show how narrow and inaccurate that vision is, and I respectfully dissent.
The Court holds that Alcala’s consent to search was shown, in the state court proceedings, to be constitutionally valid as a relinquishment of his Fourth Amendment rights. In those proceedings, no evidence was adduced as to Alcala’s knowledge of his right to refuse assent. The Court of Appeals for the Ninth Circuit, whose judgment is today reversed, would have required petitioner to pro*278duce such evidence. As discussed infra, at 286, the Court of Appeals did not hold that the police must inform a subject of investigation of his right to refuse assent as an essential predicate to their effort to secure consent to search.
The Court concedes that the police lacked probable cause to search. Ante, at 227-228. At the time the search was conducted, there were three police vehicles near the car. 270 Cal. App. 2d 648, 651, 76 Cal. Rptr. 17, 19 (1969). Perhaps the police in fact had some reason, not disclosed in this record, to believe that a search would turn up incriminating evidence. But it is also possible that the late hour and the number of men in the car suggested to the first officer on the scene that it would be prudent to wait until other officers had arrived before investigating any further.
Because Bustamonte was charged with possessing stolen checks found in the search at which he was present, he has standing to object to the search even though he claims no possessory or proprietary interest in the car. Jones v. United States, 362 U. S. 257 (1960). Cf. People v. Ibarra, 60 Cal. 2d 460, 386 P. 2d 487 (1963); People v. Perez, 62 Cal. 2d 769, 401 P. 2d 934 (1965).
The Court reads Davis v. United States, 328 U. S. 582 (1946), as upholding a search like the one in this case on the basis of consent. But it was central to the reasoning of the Court in that case that the items seized were the property of the Government temporarily in Davis’ custody. See id., at 587-593. The agents of the Government were thus simply demanding that property to which they had a lawful claim be returned to them. Because of this, the Court held that “permissible limits of persuasion are not so narrow as where 'private papers are sought.” Id., at 593. The opinion of the Court therefore explicitly disclaimed stating a general rule for ordinary searches for evidence. That the distinction, for purposes of Fourth Amendment analysis, between mere evidence and contraband or instrumentalities has now been abolished, Warden v. Hayden, 387 U. S. 294 (1967), is no reason to disregard the fact that when Davis was decided, that distinction played an important role in shaping analysis.
In Zap v. United States, 328 U. S. 624, 628 (1946), the Court held that “when petitioner, in order to obtain the Government’s business, specifically agreed to permit inspection of his accounts and records, he voluntarily waived such claim to privacy which he otherwise might have had as respects business documents related to those contracts.” (Emphasis added.) Because Zap had signed a contract specifically providing that his records would be open at all time to the Government, he had indeed waived his right to keep those records private. Cf. United States v. Biswell, 406 U. S. 311 (1972).
Aside from Zap and Davis, supra, n. 4, I have found no cases decided by this Court explicitly upholding a search based on the consent of the defendant. It is hardly surprising, then, that “[t]he approach of the Court of Appeals for the Ninth Circuit finds no support in any of our decisions,” ante, at 229. But in nearly every case discussing the problem at length, the Court referred to consent as a waiver. And it mischaracterizes those cases to describe them as analyzing the totality of the circumstances, ante, at 243 n. 31. See infra, at 283-284.
That this application of the ''domino” method of adjudication is misguided is shown, I believe, by the fact that the phrase “voluntary consent” seems redundant in a way that the phrase “voluntary confession” does not.
The Court used the terms “voluntary” or “involuntary” in such cases as shorthand labels for an assessment of the police behavior in light of the particular characteristics of the individual defendant because behavior that might not be coercive of some individuals might nonetheless compel others to give incriminating statements. See, e. g., Haley v. Ohio, 332 U. S. 596, 599 (1948); Stein v. New York, 346 U. S. 156, 185 (1953); Fikes v. Alabama, 352 U. S. 191 (1957).
I, of course, agree with the Court’s analysis to the extent that it treats a verbal expression of assent as no true consent when it is elicited through compulsion. Ante, at 229. Since, in my view, it is just as unconstitutional to search after coercing consent as it is to search after uninformed consent, I agree with the rationale of Amos v. United States, 25S U. S. 313 (1921), Johnson v. United States, 333 TJ. S. 10 (1948), and Bumper v. North Carolina, 391 U. S. 543 (1968). That an alternative rationale might have been used in those cases seems to me irrelevant.
See, e. g., Coolidge v. New Hampshire, 403 U. S. 443 (1971); Chimel v. California, 395 U. S. 752 (1969); Warden v. Hayden, 387 U. S. 294 (1967).
In Chimel, we explained that searches incident to arrest were justified by the need to protect officers from attacks by the persons they have arrested, and by the need to assure that easily destructible evidence in the reach of the suspect will not be destroyed. 395 U. S., at 762-763. And in Coolidge, we said that searches of automobiles on the highway are justified because an alerted criminal might easily drive the evidence away while a warrant was sought. *283403 U. S., at 459-462. In neither situation is police convenience alone a sufficient reason for establishing an exception to the warrant requirement. Yet the Court today seems to say that convenience alone justifies consent searches.
The Court’s interpretation of Johnson v. United States, 333 U. S. 10 (1948), a similar case, is baffling. The Court in Johnson did not in fact analyze the totality of the circumstances, as the Court now argues, ante, at 243 n. 31; the single fact that the police claimed authority to search when in truth they lacked such authority conclusively established that no valid consent had been given.
The proposition rejected in the cases cited by the Court in nn. 13 and 14, was that, as in Miranda v. Arizona, 384 U. S. 436 (1966), *287a statement to the subject of his rights must be given as an indispensable prerequisite to a request for consent to search. This case does not require us to address that proposition, for all that is involved here is the contention that the prosecution could satisfy the burden of establishing the knowledge of the right to refuse consent by showing that the police advised the subject of a search, that is sought to be justified by consent, of that right.
The Court’s suggestion that it would be “unrealistic” to require the officers to make “the detailed type of examination” involved when a court considers whether a defendant has waived a trial right, ante, at 245, deserves little comment. The question before us relates to the inquiry to be made in court when the prosecution seeks to establish that consent was given. I therefore do not address the Court’s strained argument that one may waive constitutional rights without making a knowing and intentional choice so long as the rights do not relate to the fairness of a criminal trial. I would suggest, however, that that argument is fundamentally inconsistent with the law of unconstitutional conditions. See, e. g., Perry v. Sindermann, 408 U. S. 593 (1972); Shapiro v. Thompson, 394 U. S. 618 (1969); Sherbert v. Verner, 374 17. S. 398 (1963); Speiser v. Randall, 357 U. S. 513 (1958). The discussion of United States v. Wade, 388 U. S. 218 (1967), ante, at 239-240, also seems inconsistent with the opinion of Mr. Justice Stewart in Kirby v. Illinois, 406 U. S. 682 (1972). In any event, I do not understand how one can relinquish a right without knowing of its existence, and that is the only issue in this case.
The Court’s half-hearted defense, that lack of knowledge is to be “taken into account,” rings rather hollow, in light of the apparent import of the opinion that even a subject who proves his lack of knowledge may nonetheless have consented “voluntarily,” under the Court’s peculiar definition of voluntariness.