dissenting.
. As the Court points out, it is unusual to find a statute of limitations keyed not to the time of the plaintiff’s injury, but to other related events. Still, I believe that Congress has written such a statute here, and we should respect its decision.
The language of the statute, 31 U. S. C. § 3731(b)(1), is reasonably clear. It says that “[a] civil action under section 3730 may not be brought . . . more than 6 years after the date on which the violation” of federal false claims law “is committed.” (Emphasis added.) Section 3730 lists three kinds of civil actions, including a retaliation action under § 3730(h). Thus, a retaliation action is a “civil action under section 3730,” and § 3731(b)(1)’s 6-year limitations period applies.
The Court tries to overcome the force of this syllogism with the help of two textual arguments. First, it points to the subsection that follows § 3731(b) — § 3731(c) — which says that “ ‘[i]n any action brought under section 3730, the United States shall be required to prove all essential elements of the cause of action, including damages, by a preponderance of the evidence.’ ” See ante, at 418. The Court then reasons that, read in context, the phrase “action brought under section 3730” could not refer to all the civil actions listed under § 3730, for the United States is not ordinarily a party to private retaliation suits brought under § 3730(h). Ibid. Rather, the phrase “action brought under section 3730” must refer only to the false claims actions listed in §§ 3730(a) and (b). Ibid. Thus, according to the Court, if in § 3731(c), the phrase “action brought under section 3730” refers only to a subset of the actions listed under §3730, one can read the similar phrase in § 3731(b)(1) to contain a similar limitation. Ibid.
The problem with this argument lies in its conclusion. The reason that § 3731(c) may apply only to §§ 3730(a) and *424(b) actions has nothing to do with the phrase “action brought under section 3730.” Rather, any limitation on §3731(c)’s application comes from different words, namely, “the United States.” These latter words make clear not that the phrase “under section 3730” has a different meaning than in (b), but that (c) comes into play only in cases in which the United States is a party (and only in such cases, cf. ibid.). Because it is these words — the subject of the subsection, “the United States” — that determines whether (c) has application in any given case, there is nothing in § 3731(c) that would make it “reasonable,” ibid., to read the phrase “action under section 3730” in § 3781(b)(1) to apply, as the Court concludes, to only “two out of three actions under section 3730.”
The subsections surrounding §§ 3731(b) and (c) further undermine the Court’s extratextual limitation on “[a] civil action under section 3730.” In § 3731(a), Congress apparently used the phrase “under section 3730” to mean all three § 3730 actions. § 3731(a) (a “subpoena requiring the attendance of a witness at a trial or hearing conducted under section 3730 of this title may be served at any place in the United States”). And in § 3731(d), Congress used the very words that the Court seeks to find in § 3731(b), but that do not there exist — namely, the words “under subsection (a) or (b) of section 3730” — when it meant to narrow a provision’s compass to two out of the three §3730 causes of action. § 3731(d) (“[A] final judgment rendered in favor of the United States in any criminal proceeding charging fraud or false statements ... shall estop the defendant from denying the essential elements of the offense in any action which involves the same transaction as in the criminal proceeding and which is brought under subsection (a) or (b) of section 3730”); see also ante, at 421-422. The statutory context therefore shows that Congress did not intend for the phrase “[a] civil action under section 3730” to mean anything other than what it says.
*425Second, the Court points to language in § 3731(b)(1) that specifies when the limitations period begins to run: “the date on which the violation” of the false claims provision, § 3729, “is committed.” See ante, at 415-416. It then points out that a retaliation action does not necessarily involve an actual false claims violation, because (it assumes) a retaliation plaintiff need only show “a suspected violation.” Ante, at 417 (emphasis in original). Thus, adopting relator’s and the Government’s reading, the Court reasons, would require reading some words into § 3731(b)(1) — so that it would say “ ‘the [suspected or actual] violation’ ” — which would distort the statute more than reading some other, different words into the statute — so that it would say “ ‘[a] civil action under section 3730[(a) or (b)].’” Ibid.
The difficulty with the Court’s choice of the latter linguistic addition is that the two sets of textual insertions — on the one hand “suspected or actual,” on the other hand “(a) or (b)” — are not equivalent. Statutes of limitations, when referring to starting points, generally refer not to actual events, but to alleged events. Thus, a plaintiff’s tort action is timely if he files it within, say, three years of the alleged negligently caused injury; a plaintiff’s breach-of-contract action is timely if filed within, say, one year of the alleged breach. And a plaintiff who loses such an action because the defendant shows, say, that there was no such injury or no such breach, has not, for that reason, brought the action outside the limitations period. Rather, the suit is still timely even though the violation remains nothing more than “alleged” after trial. Such a plaintiff has simply lost a timely filed action on the merits.
The provision before us is no different. Section 3731(b)(l)’s 6-year time clock begins to run on “the date on which the violation” of federal false claims law, §3729, “is committed.” Thus, any § 3730 plaintiff — even one bringing a false claims action under § 3730(a) or § 3730(b) — has six *426years from the moment of a suspected — that is, an unproven — violation of the False Claims Act’s antifraud provision. Thus, as naturally interpreted, the words “the date on which the violation ... is committed” refer to the date on which the suspected violation occurs.
I recognize that there is a relevant distinction in this case. In the typical case (say, the tort or contract case) the plaintiff must ultimately prove all the relevant allegations. Here, the retaliation victim need not prove that her employer did in fact violate federal false claims law, but only that she believed that there was such a violation. See ibid. But that distinction does not make the difference. Given the clear link between claimed violations of federal false claims law and retaliation actions (the latter depend on the former) and given that triggering events in statutes of limitations implicitly mean alleged triggering events, § 3731(b)(1) remains most naturally read as implicitly referring to alleged or suspected violations of federal false claims law. And at the very least, reading the statute in this way, especially in light of the other statutory indicators, see supra, at 423-424, does far less violence to § 3731(b)(1)’s text than does the Court’s addition of “(a) or (b).”
The Court’s far stronger argument is not textual. It concerns the limitations provision’s purpose. That purpose, after all, includes providing victims of retaliation a reasonable time within which they can file an action to vindicate • their rights. Cf. S. Rep. No. 99-345, p. 34 (1986) (addition of a retaliation cause of action intended “to halt companies . . . from using the threat of economic retaliation to silence ‘whistleblowers’ ” and to “assure those who may be considering exposing fraud that they are legally protected from retaliatory acts”). How can we reconcile that purpose with a reading of the statute that, as a matter of logic, could allow the limitations period to begin to run, perhaps even to terminate, before the forbidden retaliation occurs? See ante, at 421.
*427The answer, in my view, is that Congress could have had a particular qui tom-related purpose in mind. That is, it could have intended to provide a fairly lengthy limitations period, namely, six years from the time the false claims conduct occurs, applicable to all related actions, and then to put an end to all such litigation. This makes particular sense given the reasonable assumption that false claims and retaliation actions are likely to be litigated together. See, e. g., App. 11-35 (relator’s complaint pursuing both qui tam and retaliation claims in same suit); United States ex rel. Lujan v. Hughes Aircraft Co., 162 F. 3d 1027, 1030 (CA9 1998) (same).
Of course, as the Court emphasizes, such an unusual provision exacts a price, namely, possible injury to an individual who suffers retaliation that comes late in the day. But apparently there is no such individual. Neither the Court nor petitioners have been able to find any actual example. See, e. g., Tr. of Oral Arg. 5, 6; see also Brief for United States as Amicus Curiae 27-28 (United States is unaware of any such example). Nor have I.
By contrast, the Court’s reading of the statute exacts a different, but certain, price. It substitutes for a fairly lengthy — and uniform — 6-yéar limitations term, a crazy quilt of limitations periods stitched together from the laws of 51 jurisdictions which, in some instances, might require a plaintiff to bring a retaliation claim within 90 days, six months, or one year after the retaliation takes place. See, e. g., Ky. Rev. Stat. Ann. § 61.103(2) (Lexis 2004) (90-day limitations period for certain whistle-blower actions); Fla. Stat. § 112.3187(8)(a) (2003) (180-day limitations period); Hughes Aircraft Co., supra, at 1035 (California’s 1-year limitations period for wrongful termination in violation of public policy applies to § 3730(h) action). Rather than shed crocodile tears for the imagined plight of a nonexistent whistle-blower as petitioners ask us to do, I would read the statute to do what the statute says Congress wanted: to provide a rela*428tively long, single, uniform limitations period that, in practice, seems to protect the many real potential plaintiffs, such as relator, who will otherwise find themselves shut out of court. Such a reading also avoids the attendant practical difficulties and uncertainties inherent in requiring federal courts to borrow state statutes of limitations. See Jones v. R. R. Donnelley & Sons Co., 541 U. S. 369, 377-380, and n. 13 (2004) (discussing problems with this practice).
For these reasons, I respectfully dissent.