dissenting.
In Crawford v. Washington, 541 U. S. 36 (2004), we held that the Sixth Amendment’s Confrontation Clause bars ad*381mission against a criminal defendant of an un-cross-examined “testimonial” statement that an unavailable witness previously made out of court. Id., at 68. We simultaneously recognized an exception: that the defendant, by his own “wrongdoing,” can forfeit “on essentially equitable grounds” his Confrontation Clause right. Id., at 62. In Davis v. Washington, 547 U. S. 813 (2006), we again recognized this exception, stating that “one who obtains the absence of a witness by wrongdoing forfeits the constitutional right to confrontation.” Id., at 833.
This case involves a witness who, crying as she spoke, told a police officer how her former boyfriend (now, the defendant) had choked her, “opened a folding knife,” and “threatened to kill her.” Ante, at 357 (opinion of the' Court). Three weeks later, the defendant did kill her. At his murder trial, the defendant testified that he had acted in self-defense. To support that assertion, he described the victim as jealous, vindictive, aggressive, and violent. To rebut the defendant’s claim of self-defense and impeach his testimony, the State introduced into evidence the witness’ earlier un-crossexamined statements (as state hearsay law permits it to do) to help rebut the defendant’s claim of self-defense. It is important to underscore that this case is premised on the assumption, not challenged here, that the witness’ statements are testimonial for purposes of the Confrontation Clause. With that understanding, we ask whether the defendant, through his wrongdoing, has forfeited his Confrontation Clause right. The Court concludes that he may not have forfeited that right. In my view, however, he has.
I
Like the majority, I believe it important to recognize the relevant history, and I start where the majority starts, with Lord Morley’s Case, 6 How. St. Tr. 769 (H. L. 1666). In that case, the judges of the House of Lords wrote that a coroner’s out-of-court “examinations” of witnesses “might be read” in court if “the witnesses . . . were dead, or unable to travel.” *382Id., at 770. Additionally, they agreed, an examination “might be read” if the “witness who had been examined by the coroner, and was then absent, was detained by the means or procurement of the prisoner.” Id., at 770-771 (emphasis added). Later cases repeated this rule and followed it, admitting depositions where, e. g., “there ha[d] been evidence given of ill practice to take [the witness] out of the way,” Harrison’s Case, 12 How. St. Tr. 833, 868 (H. L. 1692), where “the prisoner ha[d], by fraudulent and indirect means, procured a person that hath given information against him to a proper magistrate, to withdraw himself,” Lord Fenwick’s Case, 13 How. St. Tr. 537, 594 (H. C. 1696), where the prisoner “had resorted to a contrivance to keep the witness out of the way,” Queen v. Scaife, 117 Q. B. 238, 242, 117 Eng. Rep. 1271, 1273 (Q. B. 1851), and so forth.
Nineteenth-century American case law on the subject said approximately the same thing. See Reynolds v. United States, 98 U. S. 145, 158 (1879). For example, an 1819 South Carolina case held that a witness’ prior formal examination could be admitted because “the witness had been kept away by the contrivance of the opposite party.” Drayton v. Wells, 10 S. C. L. 409, 411. An 1856 Georgia case, relying on Lord Morley’s Case, held that a similar “examination should be read” if the witness “was detained by the means or procurement of the prisoner.” Williams v. State, 19 Ga. 402, 403. And in 1878, this Court held that “if a witness is absent by [the defendant’s] own wrongful procurement, he cannot complain” about the admission of the witness’ prior testimonial statement. Reynolds, supra, at 158.
Reynolds stated that, “if [the defendant] voluntarily keeps the witnesses away, he cannot insist on” the “privilege of being confronted with the witnesses against him,” in part because of Lord Morley’s Case and in part because the rule of forfeiture “has its foundation in the maxim that no one shall be permitted to take advantage of his own wrong[,]... *383a maxim based on the principles of common honesty.” 98 U. S., at 158-159.
These sources make clear that “forfeiture by wrongdoing” satisfies Crawford’s requirement that the Confrontation Clause be “read as a reference to the right of confrontation at common law” and that “any exception” must be “established at the time of the founding.” 541 U. S., at 54. The remaining question concerns the precise metes and bounds of the forfeiture by wrongdoing exception. We ask how to apply that exception in the present case.
II
There are several strong reasons for concluding that the forfeiture by wrongdoing exception applies here — reasons rooted in common-law history, established principles of criminal law and evidence, and the need for a rule that can be applied without creating great practical difficulties and evidentiary anomalies.
First, the language that courts have used in setting forth the exception is broad enough to cover the wrongdoing at issue in the present case (murder) and much else besides. A witness whom a defendant murders is kept from testifying “by the means ... of the prisoner,” i. e., the defendant, Lord Morley’s Case, supra, at 771; murder is indeed an “ill practice” that leads to the witness' absence, Harrison’s Case, supra, at 868; one can fairly call a murder a “contrivance to keep the witness out of the way,” Queen v. Scaife, supra, at 242, 117 Eng. Rep., at 1273; murder, if not a “fraudulent and indirect means” of keeping the witness from testifying, is a far worse, direct one, Fenwick’s Case, supra, at 594; and when a witness is “absent” due to murder, the killer likely brought about that absence by his “own wrongful procurement,” Reynolds, supra, at 158. All of the relevant English and American cases use approximately similar language. See, e. g., 1 G. Gilbert, Law of Evidence 214-215 (1791) (ex-*384animations are “to be read on the Trial” where it can be proved that the witness is “kept back from appearing by the means and procurement of the prisoner”). And I have found no case that uses language that would not bring a murder and a subsequent trial for murder within its scope.
Second, an examination of the forfeiture rule’s basic purposes and objectives indicates that the rule applies here. At the time of the founding, a leading treatise writer described the forfeiture rule as designed to ensure that the prisoner “shall never be admitted to shelter himself by such evil Practices on the Witness, that being to give him Advantage of his own Wrong.” Ibid. This Court’s own leading case explained the exception as finding its “foundation in the maxim that no one shall be permitted to take advantage of his own wrong.” Reynolds, supra, at 159. What more “evil practice,” what greater “wrong,” than to murder the witness? And what greater evidentiary “advantage” could one derive from that wrong than thereby to prevent the witness from testifying, e. g., preventing the witness from describing a history of physical abuse that is not consistent with the defendant’s claim that he killed her in self-defense?
Third, related areas of the law motivated by similar equitable principles treat forfeiture or its equivalent similarly. The common law, for example, prohibits a life insurance beneficiary who murders an insured from recovering under the policy. See, e. g., New York Mut. Life Ins. Co. v. Armstrong, 117 U. S. 591, 600 (1886) (“It would be a reproach to the jurisprudence of the country, if one could recover insurance money payable on the death of a party whose life he had feloniously taken”). And it forbids recovery when the beneficiary “feloniously kills the insured, irrespective of the purpose.” National Life Ins. Co. v. Hood’s Adm’r, 264 Ky. 516, 518, 94 S. W. 2d 1022, 1023 (Ct. App. 1936) (emphasis added) (“no difference of opinion among the courts” on the matter). Similarly, a beneficiary of a will who murders the testator *385cannot inherit under the will. See 1 W. Page, Wills § 17.19, pp. 999-1001 (2003). And this is so “whether the crime was committed for that very purpose or with some other felonious design.” Van Alstyne v. Tuffy, 103 Misc. 455, 459, 169 N. Y. S. 173, 175 (1918); see also 1 Page, supra, § 17.19, at 1002 (“This common law doctrine applies alike whether the devisee is guilty of murder, or of manslaughter” (footnote omitted)); see generally H. Hart & A. Sacks, The Legal Process: Basic Problems in the Making and Application of Law 76-94 (W. Eskridge & P. Frickey eds. 1994) (discussing so-called “slayer’s rules”); Wade, Acquisition of Property by Willfully Killing Another—A Statutory Solution, 49 Harv. L. Rev. 715, 716 (1936) (“It must be recognized . . . that the adoption of some means to prevent a slayer from acquiring property as the result of the death of a man whom he has killed is desirable”).
Fourth, under the circumstances presented by this case, there is no difficulty demonstrating the defendant’s intent. This is because the defendant here knew that murdering his ex-girlfriend would keep her from testifying; and that knowledge is sufficient to show the intent that law ordinarily demands. As this Court put the matter more than a century ago: A “‘man who performs an act which it is known will produce a particular result is from our common experience presumed to have anticipated that result and to have intended it.’” Allen v. United States, 164 U. S. 492, 496 (1896); see United States v. Aguilar, 515 U. S. 593, 613 (1995) (Scalia, J., concurring in part and dissenting in part) (“[T]he jury is entitled to presume that a person intends the natural and probable consequences of his acts”); see also G. Williams, Criminal Law § 18, p. 38 (2d ed. 1961) (“There is one situation where a consequence is deemed to be intended though it is not desired. This is where it is foreseen as substantially certain”); ALI, Model Penal Code § 2.02(2)(b)(ii) (1962) (a person acts “knowingly” if “the element involves a result of *386his conduct” and “he is aware that it is practically certain that his conduct will cause such a result”); Restatement (Second) of Torts § 8A (1977) (“The word ‘intent’ is used throughout... to denote that the actor desires to cause consequences of his act, or that he believes that the consequences are substantially certain to result from it”).
With a few criminal law exceptions not here relevant, the law holds an individual responsible for consequences known likely to follow just as if that individual had intended to achieve them. A defendant, in a criminal or a civil case, for example, cannot escape criminal or civil liability for murdering an airline passenger by claiming that his purpose in blowing up the airplane was to kill only a single passenger for her life insurance, not the others on the same flight. See 1 W. LaFave, Substantive Criminal Law § 5.2(a), p. 341 (2d ed. 2003).
This principle applies here. Suppose that a husband, H, knows that after he assaulted his wife, W, she gave statements to the police. Based on the fact that W gave statements to the police, H also knows that it is possible he will be tried for assault. If H then kills W, H cannot avoid responsibility for intentionally preventing W from testifying, not even if H says he killed W because he was angry with her and not to keep her away from the assault trial. Of course, the trial here is not for assault; it is for murder. But I should think that this fact, because of the nature of the crime, would count as a stronger, not a weaker, reason for applying the forfeiture rule. Nor should it matter that H, at the time of the murder, may have believed an assault trial more likely to take place than a murder trial, for W’s unavailability to testify at any future trial was a certain consequence of the murder. And any reasonable person would have known it. Cf. United States v. Falstaff Brewing Cory., 410 U. S. 526, 570, n. 22 (1973) (Marshall, J., concurring in result) (“[P]erhaps the oldest rule of evidence—that a man is presumed to intend the natural and probable consequences *387of his acts — is based on the common law’s preference for objectively measurable data over subjective statements of opinion and intent”).
The majority tries to overcome this elementary legal logic by claiming that the “forfeiture rule” applies, not where the defendant intends to prevent the witness from testifying, but only where that is the defendant’s purpose, i. e., that the rule applies only where the defendant acts from a particular motive, a desire to keep the witness from trial. See ante, at 359, 360 (asserting that the terms used to describe the scope of the forfeiture rule “suggest that the exception applied only when the defendant engaged in conduct designed to prevent the witness from testifying” and that a “purpose-based definition . . . governed”). But the law does not often turn matters of responsibility upon motive, rather than intent. See supra, at 385-386. And there is no reason to believe that application of the rule of forfeiture constitutes an exception to this general legal principle.
Indeed, to turn application of the forfeiture rule upon proof of the defendant’s purpose (rather than intent), as the majority does, creates serious practical evidentiary problems. Consider H who assaults W, knows she has complained to the police, and then murders her. H knows that W will be unable to testify against him at any future trial. But who knows whether H’s knowledge played a major role, a middling role, a minor role, or no role at all, in H’s decision to kill W? Who knows precisely what passed through H’s mind at the critical moment? See, e. g., State v. Romero, 2007-NMSC-013, 156 P. 3d 694, 702-703 (finding it doubtful that evidence associated with the murder would support a finding that the purpose of the murder was to keep the victim’s earlier statements to police from the jury).
Moreover, the majority’s insistence upon a showing of purpose or motive cannot be squared with the exception’s basically ethical objective. If H, by killing W, is able to keep W’s testimony out of court, then he has successfully “take[n] *388advantage of his own wrong.” Reynolds, 98 U. S., at 159. And he does so whether he killed her for the purpose of keeping her from testifying, with certain knowledge that she will not be able to testify, or with a belief that rises to a reasonable level of probability. The inequity consists of his being able to use the killing to keep out of court her statements against him. That inequity exists whether the defendant’s state of mind is purposeful, intentional (i. e., with knowledge), or simply probabilistic.
Fifth, the majority’s approach both creates evidentiary anomalies and aggravates existing evidentiary incongruities. Contrast (1) the defendant who assaults his wife and subsequently threatens her with harm if she testifies, with (2) the defendant who assaults his wife and subsequently murders her in a fit of rage. Under the majority’s interpretation, the former (whose threats make clear that his purpose was to prevent his wife from testifying) cannot benefit from his wrong, but the latter (who has committed what is undoubtedly the greater wrong) can. This is anomalous, particularly in this context where an equitable rule applies.
Now consider a trial of H for the murder of W at which H claims self-defense. As the facts of this very case demonstrate, H may be allowed to testify at length and in damning detail about W’s behavior — what she said as well as what she did — both before and during the crime. See, e. g., Tr. 643-645 (Apr. 1, 2003). H may be able to introduce some of W’s statements (as he remembers them) under hearsay exceptions for excited utterances or present sense impressions or to show states of mind (here the victim’s statements were admitted through petitioner’s testimony to show her state of mind). W, who is dead, cannot reply. This incongruity arises in part from the nature of hearsay and the application of ordinary hearsay rules. But the majority would aggravate the incongruity by prohibiting admission of W’s out-of-court statements to the police (which contradict H’s account), even when they too fall within a hearsay exception, simply *389because there is no evidence that H was focused on his future trial when he killed her. There is no reason to do so.
Consider also that California’s hearsay rules authorize admission of the out-of-court statement of an unavailable declarant where the statement describes or explains the “infliction or threat of physical injury upon the declarant,” if the “statement” was “made at or near the time of the infliction or threat of physical injury.” Cal. Evid. Code Ann. § 1370 (West Supp. 2008). Where a victim’s statement is not “testimonial,” perhaps because she made it to a nurse, the statement could come into evidence under this Rule. But where the statement is made formally to a police officer, the majority’s rule would keep it out. Again this incongruity arises in part because of pre-existing confrontation-related rules. See Davis, 547 U. S., at 831, n. 5 (“[F]ormality is indeed essential to testimonial utterance”). But, again, the majority would aggravate the incongruity by prohibiting admission of W’s out-of-court statements to the police simply because there is no evidence that H was focused on his future trial when he killed her. Again, there is no reason to do so.
Sixth, to deny the majority’s interpretation is not to deny defendants evidentiary safeguards. It does, of course, in this particular area, deny defendants the right always to cross-examine. But the hearsay rule has always contained exceptions that permit the admission of evidence where the need is significant and where alternative safeguards of reliability exist. Those exceptions have evolved over time, see 2 K. Broun, McCormick on Evidence § 326 (6th ed. 2006) (discussing the development of the modern hearsay rule); Fed. Rule Evid. 102 (“These rules shall be construed to secure ... promotion of growth and development of the law of evidence”), often in a direction that permits admission of hearsay only where adequate alternative assurance of reliability exists, see, e. g., Rule 807 (the “Residual Exception”). Here, for example, the presence in court of a witness who took the declarant’s statement permits cross-examination of that *390witness as to just what the declarant said and as to the surrounding circumstances, while those circumstances themselves provide sufficient guarantees of accuracy to warrant admission under a State’s hearsay exception. See Cal. Evid. Code Ann. § 1370.
More importantly, to apply the forfeiture exception here simply lowers a constitutional barrier to admission of earlier testimonial statements; it does not require their admission. State hearsay rules remain in place; and those rules will determine when, whether, and how evidence of the kind at issue here will come into evidence. A State, for example, may enact a forfeiture rule as one of its hearsay exceptions, while simultaneously reading into that rule requirements limiting its application. See ante, at 367-368, n. 2. To lower the constitutional barrier to admission is to allow the States to do just that, i. e., to apply their evidentiary rules with flexibility and to revise their rules as experience suggests would be advisable. The majority’s rule, which requires exclusion, would deprive the States of this freedom and flexibility.
Ill
A
The majority tries to find support for its view in 17th-, 18th-, and 19th-century law of evidence. But a review of the cases set forth in Part I, supra, makes clear that no case limits forfeiture to instances where the defendant’s purpose or motivation is to keep the witness away. See supra, at 381-383. To the contrary, this Court stated in Reynolds that the “Constitution does not guarantee an accused person against the legitimate consequences of his own wrongful acts.” 98 U. S., at 158 (emphasis added). The words “legitimate consequences” do not mean “desired consequences” or refer to purpose or motive; in fact, the words “legitimate consequences” can encompass imputed consequences as well as intended consequences. And this Court’s statement in *391Reynolds that the rule “has its foundation in the maxim that no one shall be permitted to take advantage of his own wrong” suggests that forfeiture applies where the defendant benefits from a witness’ absence, regardless of the defendant’s specific purpose. Id., at 159.
Rather than limit forfeiture to instances where the defendant’s act has absence of the witness as its purpose, the relevant cases suggest that the forfeiture rule would apply where the witness’ absence was the known consequence of the defendant’s intentional wrongful act. Lord Morley’s Case and numerous others upon which the forfeiture rule is based say that a Marian deposition (i. e., a deposition taken by a coroner or magistrate pursuant to the Marian bail and commitment statutes) may be read to the jury if the witness who was absent was detained “by the means or procurement of the prisoner.” Lord Morley’s Case, 6 How. St. Tr., at 771. The phrase “by means of” focuses on what the defendant did, not his motive for (or purpose in) doing it. In Diaz v. United States, 223 U. S. 442 (1912), which followed Reynolds, this Court used the word “by” (the witness was absent “by the wrongful act of” the accused), a word that suggests causation, not motive or purpose. 223 U. S., at 452; see Eureka Lake & Yuba Canal Co. v. Superior Court of Yuba Cty., 116 U. S. 410, 418 (1886). And in Motes v. United States, 178 U. S. 458, 473-474 (1900), the Court spoke of absence “with the assent of” the defendant, a phrase perfectly consistent with an absence that is a consequence of, not the purpose of, what the assenting defendant hoped to accomplish.
Petitioner’s argument that the word “procurement” implies purpose or motive is unpersuasive. See Brief for Petitioner 26-28. Although a person may “procure” a result purposefully, a person may also “procure” a result by causing it, as the word “procure” can, and at common law did, mean “cause,” “bring about,” and “effect,” all words that say nothing about motive or purpose. 2 N. Webster, An American Dictionary of the English Language (1828); see also 2 C. *392Richardson, New Dictionary of the English Language 1514 (1839) (defining “procure” to mean “[t]o take care for; to take care or heed,... that any thing be done; to urge or endeavor, to manage or contrive that it be done; to acquire; to obtain”). The majority’s similar argument about the word “contrivance” fares no better. See ante, at 360 (citing, e.g., 1 J. Chitty, A Practical Treatise on the Criminal Law 81 (1816) (hereinafter Chitty) (“kept away by the means and contrivance of the prisoner”)). Even if a defendant had contrived, i. e., devised or planned, to murder a victim, thereby keeping her away, it does not mean that he did so with the purpose of keeping her away in mind. Regardless, the relevant phrase in Lord Morley’s Case is “by the means or procurement of” the defendant. 6 How. St. Tr., at 771 (emphasis added). And, as I have explained, an absence “by means of” the defendant’s actions may, or may not, refer to an absence that the defendant desired, as compared to an absence that the defendant caused.
The sole authority that expressly supports the majority’s interpretation is an 1858 treatise stating that depositions were admissible if the witness “had been kept out of the way by the prisoner, or by some one on the prisoner’s behalf, in order to prevent him from giving evidence against him.” E. Powell, Practice of the Law of Evidence 166. This treatise was written nearly 70 years after the founding; it does not explain the basis for this conclusion; and, above all, it concerns a complete exception to the hearsay rule. Were there no such limitation, all a murder victim’s hearsay statements, not simply the victim’s testimonial statements, could be introduced into evidence. Here we deal only with a constitutional bar to the admission of testimonial statements. And an exception from the general constitutional bar does not automatically admit the evidence. Rather, it leaves the State free to decide, via its own hearsay rules and hearsay exceptions, which such statements are sufficiently reliable to admit.
*393B
Given the absence of any evidence squarely requiring purpose rather than intent, what is the majority to say? The majority first tries to draw support from the absence of any murder case in which the victim’s Marian statement was read to the jury on the ground that the defendant had killed the victim. See ante, at 361-364. I know of no instance in which this Court has drawn a conclusion about the meaning of a common-law rule solely from the absence of cases showing the contrary — at least not where there are other plausible explanations for that absence. And there are such explanations here.
The most obvious reason why the majority cannot find an instance where a court applied the rule of forfeiture at a murder trial is that many (perhaps all) common-law courts thought the rule of forfeiture irrelevant in such cases. In a murder case, the relevant witness, the murder victim, was dead; and historical legal authorities tell us that, when a witness was dead, the common law admitted a Marian statement. See, e. g., Lord Morley’s Case, supra, at 770-777 (Marian depositions “might be read” if the witness was “dead or unable to travel”); King v. Woodcock, 1 Leach 500, 502, 168 Eng. Rep. 352, 353 (1789) (“[I]f the deponent should die between the time of examination and the trial of the prisoner, [the Marian deposition] may be, substituted in the room of that viva voce testimony which the deponent, if living, could alone have given, and is admitted of necessity as evidence of the fact”); J. Archbold, A Summary of the Law Relative to Pleading and Evidence in Criminal Cases 85 (1822) (where a witness was “dead,” “unable to travel,” or “kept away by the means or procurement of the prisoner,” Marian depositions “may be given in evidence against the prisoner”). Because the Marian statements of a deceased witness were admissible simply by virtue of the witness’ death, there would have been no need to argue for their admission pursuant to a forfeiture rule.
*394Historical authorities also tell us that a Marian statement could not be admitted unless it was a proper Marian deposition, meaning that the statement was given in the presence of the defendant thereby providing an opportunity to cross-examine the witness. And this was the case whether the witness’ unavailability was due to death or the “means or procurement” of the defendant. See, e. g., ibid. (Where a witness was “dead,” “unable to travel,” or “kept away by the means or procurement of the prisoner” depositions could be read but they “must have been taken in the presence of the prisoner, so that he might have had an opportunity of cross examining the witness” (emphasis added)); 2 W. Hawkins, Pleas of the Crown 605-606 (6th ed. 1787) (hereinafter Hawkins); Chitty 78-80; 2 J. Bishop, New Criminal Procedure §§ 1194-1195, pp. 1020-1022 (2d ed. 1913) (hereinafter Bishop); Lord Fenwick’s Case, 13 How. St. Tr., at 602. Thus, in a murder trial, where the witness was dead, either the Marian statement was proper and it came into evidence without the forfeiture exception; or it was improper and the forfeiture exception could not have helped it come in. Cf. King v. Dingier, 2 Leach 561, 563, 168 Eng. Rep. 383, 384 (1791) (a top barrister of the day argued successfully that “it is utterly impossible, unless the prisoner had been present [at the Marian deposition], that depositions thus taken can be read”). No wonder then that the majority cannot find a murder case that refers directly to the forfeiture exception. Common-law courts likely thought the forfeiture exception irrelevant in such a case.
The majority highlights two common-law murder cases that demonstrate this point—King v. Woodcock and King v. Dingler. See ante, at 362-363. As the majority explains, in each of these two cases, the defendant stood accused of killing his wife. In each case, the victim had given an account of the crime prior to her death. And in each case, the court refused to admit the statements (statements that might have been admitted simply by virtue of the fact that *395the witness had died) on the ground that they were not properly taken Marian statements, i. e., not made in the presence of the defendant. Because admission pursuant to the forfeiture rule also would have required the statements to have been properly taken, there would have been no reason to argue for their admission on that basis. Instead, in each case, the prosecution argued that the statement be admitted as a dying declaration. In Woodcock, depending on the account, the court either instructed the jury to consider whether the statements were made “under the apprehension of death,” or determined for itself that they were and admitted them into evidence. 1 Leach, at 504, 168 Eng. Rep., at 354; see 1 E. East, Pleas of the Crown 356 (1803) (reprinted 2004). In Dingier, because the Crown admitted that the statements were not made “under apprehension of immediate death,” the statements were excluded. 2 Leach, at 563, 168 Eng. Rep., at 384. The forfeiture rule thus had no place in Woodcock or Dingier, not because of the state of mind of the defendant when he committed his crime, but because the victim’s testimony was not a properly taken Marian statement.
The American murder cases to which the majority refers provide it no more support. See ante, at 363 (citing United States v. Woods, 28 F. Cas. 762, 763 (No. 16,760) (CC DC 1834); Lewis v. State, 17 Miss. 115, 120 (1847); Montgomery v. State, 11 Ohio 424, 425-426 (1842); Nelson v. State, 26 Tenn. 542, 543 (1847); Smith v. State, 28 Tenn. 9, 23 (1848)). Like Woodcock and Dingier, these are dying declaration cases. While it is true that none refers to the forfeiture exception, it is also true that none of these cases involved a previously given proper Marian deposition or its equivalent.
There are other explanations as well for the absence of authority to which the majority points. The defendant’s state of mind only arises as an issue in forfeiture cases where the witness has made prior statements against the defendant and where there is a possible motive for the killing other than to prevent the witness from testifying. (Where that *396motive is certain — for example, where the defendant knows the witness only because she has previously testified against him — the prior statements would be admitted under the majority’s purpose rule, and the question of intent would not come up.) We can see from modern cases that this occurs almost exclusively in the domestic violence context, where a victim of the violence makes statements to the police and where it is not certain whether the defendant subsequently killed her to prevent her from testifying, to retaliate against her for making statements, or in the course of another abusive incident. But 200 years ago, it might have been seen as futile for women to hale their abusers before a Marian magistrate where they would make such a statement. See, e. g., State v. Rhodes, 61 N. C. 453, 459 (1868) (per curiam) (“We will not inflict upon society the greater evil of raising the curtain upon domestic privacy, to punish the lesser evil of trifling violence”).
I also recognize the possibility that there are too few old records available for us to draw firm conclusions. Indeed, the “continuing confusion about the very nature of the law of evidence at the end of the eighteenth century underscores how primitive and undertheorized the subject then was.” J. Langbein, The Origins of Adversary Criminal Trial 248 (2003).
Regardless, the first explanation — that the forfeiture doctrine could not have helped admit an improperly taken Marian deposition — provides a sufficient ground to conclude that the majority has found nothing in the common-law murder cases, domestic or foreign, that contradicts the traditional legal principles supporting application of the rule of forfeiture here. See Williams, Criminal Law § 18, at 39 (relying on sources at common law for the proposition that the accused “necessarily intends that which must be the consequence of the act” (internal quotation marks omitted)); LaFave, Substantive Criminal Law § 5.2(a), at 341 (“[T]he traditional *397view is that a person who acts . . . intends a result of his act.. . when he knows that that result is practically certain to follow from his conduct, whatever his desire may be as to that result”).
The majority next points to a second line of common-law cases, cases in which a court admitted a murdered witness’ “dying declaration.” But those cases do not support the majority’s conclusion. A dying declaration can come into evidence when it is “made in extremity” under a sense of impending death, “when every hope of this world is gone: when every motive to falsehood is silenced, and the mind is induced by the most powerful considerations to speak the truth.” Woodcock, supra, at 502, 168 Eng. Rep., at 353; see King v. Drummond, 1 Leach 337, 338, 168 Eng. Rep. 271, 272 (1784) (“[T]he mind, impressed with the awful idea of approaching dissolution, acts under a sanction equally powerful with that which it is presumed to feel by a solemn appeal to God upon an oath”); see also Hawkins 619, n. 10; Mattox v. United States, 156 U. S. 237, 243-244 (1895). The majority notes that prosecutors did not attempt to obtain admission of dying declarations on forfeiture grounds before trying to meet these strict “dying declaratiofn]” requirements. See ante, at 364. This failure, it believes, supports its conclusion that admission pursuant to the forfeiture exception required a showing that the defendant killed the witness with the purpose of securing the absence of that witness at trial.
There is a simpler explanation, however, for the fact that parties did not argue forfeiture in “dying declaration” cases. And it is the explanation I have already mentioned. The forfeiture exception permitted admission only of a properly taken Marian deposition. And where death was at issue, the forfeiture exception was irrelevant. In other words, if the Marian deposition was proper, the rule of forfeiture was unnecessary; if the deposition was improper, the rule of forfeiture was powerless to help. That is why we find lawyers *398in “dying declaration” cases arguing that the dying declaration was either a proper Marian deposition (in which case it was admitted) or it was a “dying declaration” (in which case it was admitted), or both. See, e. g., Dingier, 2 Leach, at 562, 168 Eng. Rep., at 383-384 (discussing the admission of statements either “as a deposition taken pursuant to the [Marian] statutes” or, in the alternative, “as the dying declaration of a party conscious of approaching dissolution”); King v. Radbourne, 1 Leach 457, 460-461, 168 Eng. Rep. 330, 332 (1787) (same); People v. Restell, 3 Hill 289 (N. Y. 1842) (same); see also Chitty 79-81. Under these circumstances, there would have been little reason to add the word “forfeiture.”
For the same reason, we can find “dying declarations” admitted in murder cases where no proper Marian deposition existed, see, e. g., King v. Woodcock, 1 Leach 500, 168 Eng. Rep. 352; 1 East, Pleas of the Crown, at 356, or in cases involving, say, wills or paternity disputes, where Marian statements were not at all at issue, see 5 J. Wigmore, Evidence § 1431, p. 277, n. 2 (J. Chadbourn rev. ed. 1974) (citing such cases from the 18th and 19th centuries). Cf. Langbein, supra, at 245-246, nn. 291, 292 (at common law, there existed both oath-based and cross-examination-based rationales for the hearsay rule, with the latter only becoming dominant around the turn of the 19th century (citing Gallanis, The Rise of Modern Evidence Law, 84 Iowa L. Rev. 499, 516-550 (1999))).
The upshot is that the majority fails to achieve its basic objective. It cannot show that the common law insisted upon a showing that a defendant’s purpose or motive in killing a victim was to prevent the victim from testifying. At the least its authority is consistent with my own view, that the prosecution in such a case need show no more than intent (based on knowledge) to do so. And the most the majority might show is that the common law was not clear on the point.
*399IV
A
The majority makes three arguments in response. First, it says that I am wrong about unconfronted statements at common law. According to the majority, when courts found wrongful procurement, they admitted a defendant’s statements without regard to whether they were confronted. See ante, at 369-373. That being so, the majority's argument goes, one must wonder why no one argued for admissibility under the forfeiture rule in, say, Woodcock or Dingier. See ante, at 362-363. The reason, the majority concludes, is that the forfeiture rule would not have helped secure admission of the (unconfronted) prior statements in those cases, because the forfeiture rule applied only where the defendant purposely got rid of the witness. See ante, at 361. But the majority’s house of cards has no foundation; it is built on what is at most common-law silence on the subject. The cases it cites tell us next to nothing about admission of unconfronted statements.
Fenwick’s Case, see ante, at 369-370, n. 3, for example, was a parliamentary attainder proceeding; Parliament voted to admit unconfronted statements but it is not clear what arguments for admission Parliament relied upon. See generally 13 How. St. Tr. 537. Hence it is not clear that Parliament admitted unconfronted statements pursuant to a forfeiture theory. In fact, the forfeiture rule in a felony case was described in Fenwick’s Case as applying where the witness “hath given information against [the defendant] to a proper magistrate,” id., at 594 (remarks of Lovel), i. e., a magistrate who normally would have had the defendant before him as well.
Harrison’s Case, see ante, at 369-370, did admit an unconfronted statement, but it was a statement made before a coroner. See 12 How. St. Tr., at 852. Coroner’s statements seem to have had special status that may sometimes have *400permitted the admission of prior unconfronted testimonial statements despite lack of cross-examination. But, if so, that special status failed to survive the Atlantic voyage. See Crawford, 541 U. S., at 47, n. 2 (early American authorities “flatly rejected any special status for coroner statements”).
The American case upon which the majority primarily relies, Rex v. Barber, 1 Root 76 (Conn. Super. Ct. 1775), see ante, at 370, consists of three sentences that refer to “[o]ne White, who had testified before the justice and before the grand-jury against Barber.” 1 Root, at 76. White was “sent away” at Barber’s “instigation” and the “court admitted witnesses to relate what White had before testified.” Ibid. I cannot tell from the case whether White’s statement was made before a grand jury or was taken before a justice where cross-examination would have been possible. At least some commentators seem to think the latter. See W. Best, The Principles of the Law of Evidence 467, 473, n. (e.) (American ed. 1883) (listing Barber as a case “of preliminary investigation before a magistrate” where “evidence ha[d] been admitted, there having been a right of cross-examination”); 2 Bishop §§ 1194-1197, at 1020-1024 (explaining that where a witness had been “kept out of the way” by the defendant, his prior testimony is admissible if “the defendant had the opportunity to cross-examine the witness against him, not otherwise,” and giving as a “[fjamiliar illustration” of this principle cases before a committing magistrate including Barber (footnotes omitted)); J. Stephen, A Digest of the Law of Evidence 161, American Note, General (1902) (citing Barber for the proposition that evidence at a preliminary hearing was admissible if “the party against whom it is offered was present”).
The majority’s final authority, Williams v. State, 19 Ga., at 403, see ante, at 371, involved the admission of an “examination” taken by “the committing Magistrate.” Such examinations were ordinarily given in the presence of the defendant. *401See R. Greene & J. Lumpkin, Georgia Justice 99 (1835) (describing procedures relevant to a magistrate’s examination of a witness in Georgia); see also M. M’Kinney, The American Magistrate and Civil Officer 235 (1850) (testimony of the accuser and his witnesses taken by a magistrate “must be done in the presence of the party accused, in order that he may have the advantage of cross-examining the witnesses”).
At the same time, every Supreme Court case to apply the forfeiture rule has done so in the context of previously confronted testimony. See, e. g., Reynolds, 98 U. S., at 158 (admitting previously confronted statements pursuant to a forfeiture rule); Diaz, 223 U. S., at 449 (same); Mattox, 156 U. S., at 240 (same); Motes, 178 U. S., at 470-471 (same).
Of course, modern courts have changed the ancient common-law forfeiture rule — in my view, for the better. They now admit unconfronted prior testimonial statements pursuant to such a rule. See, e. g., United States v. Carlson, 547 F. 2d 1346, 1357-1360 (CA8 1976) (the earliest case to do so); United States v. Mastrangelo, 693 F. 2d 269 (CA2 1982); United States v. Rouco, 765 F. 2d 983 (CA11 1985); see also Davis, 547 U. S., at 834. But, as the dates of these cases indicate, the admission of unconfronted statements under a forfeiture exception is a fairly recent evidentiary development. The majority evidently finds this elephant of a change acceptable — as do I. Without it, there would be no meaningful modern-day forfeiture exception. Why then does the majority strain so hard at what, comparatively speaking, is a gnat (and a nonexistent gnat at that)?
In sum, I have tried to show the weakness of the foundation upon which the majority erects its claim that the common law applied the forfeiture rule only where it was a defendant’s purpose or motive (not his intent based on knowledge) to keep the witness away. The majority says that “the most natural reading of the language used at common law” supports its view. Ante, at 368. As I have shown, that is not so. See supra, at 383-384. The majority *402next points to “the absence of common-law cases admitting prior statements on a forfeiture theory” where the defendant prevented, but did not purposely prevent, the witness from testifying. Ante, at 368. As I have pointed out, this absence proves nothing because (1) the relevant circumstances (there has been a prior testimonial statement, the witness is now unavailable due to defendant’s actions, and the defendant knows that the witness will not testify but that is not his purpose) are likely to arise almost exclusively when the defendant murders the witness, and (2) a forfeiture theory was ordinarily redundant or useless in such cases. See supra, at 393-394. The majority, describing its next argument as “conclusive,” points to “innumerable cases” where courts did not admit “unconfronted inculpatory testimony by murder victims” against a defendant. Ante, at 368. The majority is referring to those dying declaration cases in which unconfronted statements were not admitted because the witness was not sufficiently aware of his impending death when he made them. See ante, at 363-364. But as I have explained, the forfeiture rule would have been unhelpful under these circumstances. See supra, at 397-398. Finally, the majority points to a “subsequent history” in the United States where questions about the defendant’s state of mind did not begin to arise until the 1980’s. Ante, at 368. I have explained why that history does not support its view. See supra, at 401. Having only begun to swallow the elephant in the late 1970’s and early 198Q’s, it makes sense that courts would not have previously considered the gnat.
While I have set forth what I believe is the better reading of the common-law cases, I recognize that different modern judges might read that handful of cases differently. All the more reason then not to reach firm conclusions about the precise metes and bounds of a contemporary forfeiture exception by trying to guess the state of mind of 18th-century lawyers when they decided not to make a particular argument, i. e., forfeiture, in a reported case. That is why, in *403Part II, supra, I have set forth other, more conclusive reasons in support of the way I would read the exception.
Second, the plurality objects to that aspect of the forfeiture rule that requires a judge to make a preliminary assessment of the defendant’s wrongful act in order to determine whether the relevant statements should be admitted. See ante, at 374-375. But any forfeiture rule requires a judge to determine as a preliminary matter that the defendant’s own wrongdoing caused the witness to be absent. Regardless, preliminary judicial determinations are not, as the majority puts it, “akin ... to ‘dispensing with jury trial.’” Ante, at 365 (quoting Crawford, 541 U. S., at 62). We have previously said that courts may make preliminary findings of this kind. For example, where a defendant is charged with conspiracy, the judge is permitted to make an initial finding that the conspiracy existed so as to determine whether a statement can be admitted under the co-conspirator exception to the hearsay rule. See Bourjaily v. United States, 483 U. S. 171, 175-176 (1987) (“The inquiry made by a court concerned with these matters is not whether the proponent of the evidence wins or loses his ease on the merits, but whether the evidentiary Rules have been satisfied”). And even the plurality is forced to admit that it is “sometimes” necessary for a “judge ... to inquire into guilt of the charged offense in order to make a preliminary evidentiary ruling.” Ante, at 375, n. 6.
Third, the plurality seems to believe that an ordinary intent requirement, rather than a purpose or motive requirement, would let in too much out-of-court testimonial evidence. See ante, at 374-376. Ordinarily a murderer would know that his victim would not be able to testify at a murder trial. Hence all of the victim’s prior testimonial statements would come in at trial for use against a defendant. To insist upon a showing of purpose rather than plain (knowledge-based) intent would limit the amount of unconfronted evidence that the jury might hear.
*404This argument fails to account for the fact that overcoming a constitutional objection does not guarantee admissibility of the testimonial evidence at issue. The States will still control admissibility through hearsay rules and exceptions. And why not? What important constitutional interest is served, say, where a prior testimonial statement of a victim of abuse is at issue, by a constitutional rule that lets that evidence in if the defendant killed a victim purposely to stop her from testifying, but keeps it out if the defendant killed her knowing she could no longer testify while acting out of anger or revenge?
B
Even the majority appears to recognize the problem with its “purpose” requirement, for it ends its opinion by creating a kind of presumption that will transform purpose into knowledge-based, intent — at least where domestic violence is at issue; and that is the area where the problem is most likely to arise.