United States Supreme Court.
*168 O'Connor, J., delivered the opinion of the Court, in which Rehnquist, C. J., and Scalia, Kennedy, Souter, and Thomas, JJ., joined. Souter, J., filed a concurring opinion, post, p. 182. Stevens, J., filed an opinion concurring in part and concurring in the judgment, in which Souter, J., joined, post, p. 182. Breyer, J., filed a dissenting opinion, in which Ginsburg, J., joined, post, p. 185.
Preeta D. Bansal, Solicitor General of New York, argued the cause for petitioner. With her on the briefs were Eliot *169 Spitzer, Attorney General, Daniel Smirlock, Deputy Solicitor General, and David Axinn, Robert H. Easton, Robin Forshaw, and Martin A. Hotvet, Assistant Solicitors General.
Deborah Wolikow Loewenberg, by appointment of the Court, 531 U.S. 1066, argued the cause for respondent. With her on the brief were John H. Blume and Keir M. Weyble.[*]
Justice O'Connor, delivered the opinion of the Court.
Title 28 U.S. C. § 2244(d)(2) (1994 ed., Supp. V) provides: "The time during which a properly filed application for State post-conviction or other collateral review with respect to the pertinent judgment or claim is pending shall not be counted toward any period of limitation under this subsection." This case presents the question whether a federal habeas corpus petition is an "application for State postconviction or other collateral review" within the meaning of this provision.
In 1992, several judgments of conviction for robbery were entered against respondent Sherman Walker in the *170 New York state courts. The last of these convictions came in June 1992, when respondent pleaded guilty to robbery in the first degree in the New York Supreme Court, Queens County. Respondent was sentenced to 7 to 14 years in prison on this conviction.
Respondent unsuccessfully pursued a number of state remedies in connection with his convictions. It is unnecessary to describe all of these proceedings herein. Respondent's last conviction was affirmed on June 12, 1995. Respondent was later denied leave to appeal to the New York Court of Appeals. Respondent also sought a writ of error coram nobis, which the Appellate Division denied on March 18, 1996. Respondent's last conviction became final in April 1996, prior to the April 24, 1996, effective date of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), 110 Stat. 1214.
In a single document dated April 10, 1996, respondent filed a complaint under Rev. Stat. § 1979, 42 U.S. C. § 1983, and a petition for habeas corpus under 28 U.S. C. § 2254 in the United States District Court for the Eastern District of New York. On July 9, 1996, the District Court dismissed the complaint and petition without prejudice. With respect to the habeas petition, the District Court, citing § 2254(b), concluded that respondent had not adequately set forth his claim because it was not apparent that respondent had exhausted available state remedies. The District Court noted that, for example, respondent had failed to specify the claims litigated in the state appellate proceedings relating to his robbery convictions.
On May 20, 1997, more than one year after AEDPA's effective date, respondent filed another federal habeas petition in the same District Court. It is undisputed that respondent had not returned to state court since the dismissal of his first federal habeas filing. On May 6, 1998, the District Court dismissed the petition as time barred because *171 respondent had not filed the petition within a "reasonable time" from AEDPA's effective date.
The United States Court of Appeals for the Second Circuit reversed the District Court's judgment, reinstated the habeas petition, and remanded the case for further proceedings. Walker v. Artuz, 208 F.3d 357 (2000). The Court of Appeals noted at the outset that, because respondent's conviction had become final prior to AEDPA's effective date, he had until April 24, 1997, to file his federal habeas petition. The court also observed that the exclusion from the limitation period of the time during which respondent's first federal habeas petition was pending in the District Court would render the instant habeas petition timely.
The Court of Appeals held that respondent's first federal habeas petition had tolled the limitation period because it was an application for "other collateral review" within the meaning of § 2244(d)(2). The court characterized the disjunctive "or" between "post-conviction" and "other collateral" as creating a "distinct break" between two kinds of review. Id., at 359. The court also stated that application of the word "State" to both "post-conviction" and "other collateral" would create a "linguistic oddity" in the form of the construction "State other collateral review." Id., at 360. The court further reasoned that the phrase "other collateral review" would be meaningless if it did not refer to federal habeas petitions. The court therefore concluded that the word "State" modified only "post-conviction."
The Court of Appeals also found no conflict between its interpretation of the statute and the purpose of AEDPA. The court found instead that its construction would promote the goal of encouraging petitioners to file their federal habeas applications as soon as possible.
We granted certiorari, 531 U.S. 991 (2000), to resolve a conflict between the Second Circuit's decision and the decisions of three other Courts of Appeals. See Jiminez v. Rice, 222 F.3d 1210 (CA9 2000); Grooms v. Johnson, 208 F.3d 488 *172 (CA5 1999) (per curiam); Jones v. Morton, 195 F.3d 153 (CA3 1999). One other Court of Appeals has since adopted the Second Circuit's view. Petrick v. Martin, 236 F.3d 624 (CA10 2001). We now reverse.
Our task is to construe what Congress has enacted. We begin, as always, with the language of the statute. See, e. g., Williams v. Taylor, 529 U.S. 420, 431 (2000); Public Employees Retirement System of Ohio v. Betts, 492 U.S. 158, 175 (1989); Watt v. Energy Action Ed. Foundation, 454 U.S. 151, 162 (1981). Respondent reads § 2244(d)(2) to apply the word "State" only to the term "post-conviction" and not to the phrase "other collateral." Under this view, a properly filed federal habeas petition tolls the limitation period. Petitioner contends that the word "State" applies to the entire phrase "post-conviction or other collateral review." Under this view, a properly filed federal habeas petition does not toll the limitation period.
We believe that petitioner's interpretation of § 2244(d)(2) is correct for several reasons. To begin with, Congress placed the word "State" before "post-conviction or other collateral review" without specifically naming any kind of "Federal" review. The essence of respondent's position is that Congress used the phrase "other collateral review" to incorporate federal habeas petitions into the class of applications for review that toll the limitation period. But a comparison of the text of § 2244(d)(2) with the language of other AEDPA provisions supplies strong evidence that, had Congress intended to include federal habeas petitions within the scope of § 2244(d)(2), Congress would have mentioned "Federal" review expressly. In several other portions of AEDPA, Congress specifically used both the words "State" and "Federal" to denote state and federal proceedings. For example, 28 U.S. C. § 2254(i) (1994 ed., Supp. V) provides: "The ineffectiveness or incompetence of counsel *173 during Federal or State collateral post-conviction proceedings shall not be a ground for relief in a proceeding arising under section 2254." Likewise, the first sentence of 28 U.S. C. § 2261(e) (1994 ed., Supp. V) provides: "The ineffectiveness or incompetence of counsel during State or Federal post-conviction proceedings in a capital case shall not be a ground for relief in a proceeding arising under section 2254." The second sentence of § 2261(e) states: "This limitation shall not preclude the appointment of different counsel, on the court's own motion or at the request of the prisoner, at any phase of State or Federal post-conviction proceedings on the basis of the ineffectiveness or incompetence of counsel in such proceedings." Finally, 28 U.S. C. § 2264(a)(3) (1994 ed., Supp. V) excuses a state capital prisoner's failure to raise a claim properly in state court where the failure is "based on a factual predicate that could not have been discovered through the exercise of due diligence in time to present the claim for State or Federal post-conviction review."
Section 2244(d)(2), by contrast, employs the word "State," but not the word "Federal," as a modifier for "review." It is well settled that "`[w]here Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.' " Bates v. United States, 522 U.S. 23, 29-30 (1997) (quoting Russello v. United States, 464 U.S. 16, 23 (1983)). We find no likely explanation for Congress' omission of the word "Federal" in § 2244(d)(2) other than that Congress did not intend properly filed applications for federal review to toll the limitation period. It would be anomalous, to say the least, for Congress to usher in federal review under the generic rubric of "other collateral review" in a statutory provision that refers expressly to "State" review, while denominating expressly both "State" and "Federal" proceedings in other parts of the same statute. The anomaly is underscored by the fact that the words *174 "State" and "Federal" are likely to be of no small import when Congress drafts a statute that governs federal collateral review of state court judgments.
Further, were we to adopt respondent's construction of the statute, we would render the word "State" insignificant, if not wholly superfluous. "It is our duty `to give effect, if possible, to every clause and word of a statute.' " United States v. Menasche, 348 U.S. 528, 538-539 (1955) (quoting Montclair v. Ramsdell, 107 U.S. 147, 152 (1883)); see also Williams v. Taylor, 529 U.S. 362, 404 (2000) (describing this rule as a "cardinal principle of statutory construction"); Market Co. v. Hoffman, 101 U.S. 112, 115 (1879) ("As early as in Bacon's Abridgment, sect. 2, it was said that `a statute ought, upon the whole, to be so construed that, if it can be prevented, no clause, sentence, or word shall be superfluous, void, or insignificant' "). We are thus "reluctan[t] to treat statutory terms as surplusage" in any setting. Babbitt v. Sweet Home Chapter, Communities for Great Ore., 515 U.S. 687, 698 (1995); see also Ratzlaf v. United States, 510 U.S. 135, 140 (1994). We are especially unwilling to do so when the term occupies so pivotal a place in the statutory scheme as does the word "State" in the federal habeas statute. But under respondent's rendition of § 2244(d)(2), Congress' inclusion of the word "State" has no operative effect on the scope of the provision. If the phrase "State post-conviction or other collateral review" is construed to encompass both state and federal collateral review, then the word "State" places no constraint on the class of applications for review that toll the limitation period. The clause instead would have precisely the same content were it to read "post-conviction or other collateral review."
The most that could then be made of the word "State" would be to say that Congress singled out applications for "State post-conviction" review as one example from the universe of applications for collateral review. Under this approach, however, the word "State" still does nothing to *175 delimit the entire class of applications for review that toll the limitation period. A construction under which the word "State" does nothing more than further modify "postconviction" relegates "State" to quite an insignificant role in the statutory provision. We believe that our duty to "give each word some operative effect" where possible, Walters v. Metropolitan Ed. Enterprises, Inc., 519 U.S. 202, 209 (1997), requires more in this context.
The Court of Appeals characterized petitioner's interpretation as producing the "linguistic oddity" of "State other collateral review," which is "an ungainly construction that [the Court of Appeals did] not believe Congress intended." 208 F. 3d, at 360. But nothing precludes the application of the word "State" to the entire phrase "post-conviction or other collateral review," regardless of the resulting construction that one posits. The term "other collateral" is easily understood as a unit to which "State" applies just as "State" applies to "post-conviction." Moreover, petitioner's interpretation does not compel the verbal formula hypothesized by the Court of Appeals. Indeed, the ungainliness of "State other collateral review" is a very good reason why Congress might have avoided that precise verbal formulation in the first place. The application of the word "State" to the phrase "other collateral review" more naturally yields the understanding "other State collateral review."
The Court of Appeals also reasoned that petitioner's reading of the statute fails to give operative effect to the phrase "other collateral review." The court claimed that "the phrase `other collateral review' would be meaningless if it did not refer to federal habeas petitions." Ibid. This argument, however, fails because it depends on the incorrect premise that there can be no form of state "collateral" review "other" than state "post-conviction" review within the meaning of § 2244(d)(2). To the contrary, it is possible for "other collateral review" to include review of a state court judgment that is not a criminal conviction.
*176 Section 2244(d)(1)'s 1-year limitation period applies to "an application for a writ of habeas corpus by a person in custody pursuant to the judgment of a State court." Section 2244(d)(2) provides for tolling during the pendency of "a properly filed application for State post-conviction or other collateral review with respect to the pertinent judgment or claim." Nothing in the language of these provisions requires that the state court judgment pursuant to which a person is in custody be a criminal conviction. Nor does 28 U.S. C. § 2254 (1994 ed. and Supp. V) by its terms apply only to those in custody pursuant to a state criminal conviction. See, e. g., § 2254(a) ("a person in custody pursuant to the judgment of a State court"); § 2254(b)(1) ("a person in custody pursuant to the judgment of a State court"); § 2254(d) ("a person in custody pursuant to the judgment of a State court"); § 2254(e)(1) ("a person in custody pursuant to the judgment of a State court").
Incarceration pursuant to a state criminal conviction may be by far the most common and most familiar basis for satisfaction of the "in custody" requirement in § 2254 cases. But there are other types of state court judgments pursuant to which a person may be held in custody within the meaning of the federal habeas statute. For example, federal habeas corpus review may be available to challenge the legality of a state court order of civil commitment or a state court order of civil contempt. See, e. g., Francois v. Henderson, 850 F.2d 231 (CA5 1988) (entertaining a challenge brought in a federal habeas petition under § 2254 to a state court's commitment of a person to a mental institution upon a verdict of not guilty by reason of insanity); Leonard v. Hammond, 804 F.2d 838 (CA4 1986) (holding that constitutional challenges to civil contempt orders for failure to pay child support were cognizable only in a habeas corpus action). These types of state court judgments neither constitute nor require criminal convictions. Any state collateral review that is available with respect to these judgments, strictly speaking, *177 is not post-conviction review. Accordingly, even if "`"State post-conviction review" means all collateral review of a conviction provided by a state,' " 208 F. 3d, at 360 (quoting Barrett v. Yearwood, 63 F. Supp. 2d 1245, 1250 (ED Cal. 1999)), the phrase "other collateral review" need not include federal habeas petitions in order to have independent meaning.
Congress also may have employed the construction "postconviction or other collateral" in recognition of the diverse terminology that different States employ to represent the different forms of collateral review that are available after a conviction. In some jurisdictions, the term "postconviction" may denote a particular procedure for review of a conviction that is distinct from other forms of what conventionally is considered to be postconviction review. For example, Florida employs a procedure that is officially entitled a "Motion to Vacate, Set Aside, or Correct Sentence." Fla. Rule Crim. Proc. 3.850 (2001). The Florida courts have commonly referred to a Rule 3.850 motion as a "motion for post-conviction relief" and have distinguished this procedure from other vehicles for collateral review of a criminal conviction, such as a state petition for habeas corpus. See, e. g., Bryant v. State, 780 So. 2d 978, 979 (Fla. App. 2001) ("[A] petition for habeas corpus cannot be used to circumvent the two-year period for filing motions for post-conviction relief"); Finley v. State, 394 So. 2d 215, 216 (Fla. App. 1981) ("[T]he remedy of habeas corpus is not available as a substitute for post-conviction relief under Rule 3.850"). Congress may have refrained from exclusive reliance on the term "post-conviction" so as to leave no doubt that the tolling provision applies to all types of state collateral review available after a conviction and not just to those denominated "post-conviction" in the parlance of a particular jurisdiction.
Examination of another AEDPA provision also demonstrates that "other collateral" need not refer to any form of *178 federal review in order to have meaning. Title 28 U.S. C. § 2263 (1994 ed., Supp. V) establishes the limitation period for filing § 2254 petitions in state capital cases that arise from jurisdictions meeting the "opt-in" requirements of § 2261. Section 2263(b)(2) provides that the limitation period "shall be tolled from the date on which the first petition for postconviction review or other collateral relief is filed until the final State court disposition of such petition." The reference to "the final State court disposition of such petition" makes it clear that only petitions filed in state court, and not petitions for federal review, toll the limitation period in capital cases. Congress therefore used the phrases "postconviction review" and "other collateral relief" in a disjunctive clause where the term "other collateral," whatever its precise content, could not possibly include anything federal within its ambit. This illustration vitiates any suggestion that "other collateral" relief or review must include federal relief or review in order for the term to have any significance apart from "post-conviction" review.
Consideration of the competing constructions in light of AEDPA's purposes reinforces the conclusion that we draw from the text. Petitioner's interpretation of the statute is consistent with "AEDPA's purpose to further the principles of comity, finality, and federalism." Williams, 529 U. S., at 436. Specifically, under petitioner's construction, § 2244(d)(2) promotes the exhaustion of state remedies while respecting the interest in the finality of state court judgments. Under respondent's interpretation, however, the provision would do far less to encourage exhaustion prior to seeking federal habeas review and would hold greater potential to hinder finality.
The exhaustion requirement of § 2254(b) ensures that the state courts have the opportunity fully to consider federallaw challenges to a state custodial judgment before the lower federal courts may entertain a collateral attack upon that *179 judgment. See, e. g., O'Sullivan v. Boerckel, 526 U.S. 838, 845 (1999) ("[T]he exhaustion doctrine is designed to give the state courts a full and fair opportunity to resolve federal constitutional claims before those claims are presented to the federal courts"); Rose v. Lundy, 455 U.S. 509, 518-519 (1982) ("A rigorously enforced total exhaustion rule will encourage state prisoners to seek full relief first from the state courts, thus giving those courts the first opportunity to review all claims of constitutional error"). This requirement "is principally designed to protect the state courts' role in the enforcement of federal law and prevent disruption of state judicial proceedings." Id., at 518. The exhaustion rule promotes comity in that "`it would be unseemly in our dual system of government for a federal district court to upset a state court conviction without an opportunity to the state courts to correct a constitutional violation.' " Ibid. (quoting Darr v. Burford, 339 U.S. 200, 204 (1950)); see also O'Sullivan, supra, at 844 ("Comity thus dictates that when a prisoner alleges that his continued confinement for a state court conviction violates federal law, the state courts should have the first opportunity to review this claim and provide any necessary relief").
The 1-year limitation period of § 2244(d)(1) quite plainly serves the well-recognized interest in the finality of state court judgments. See generally Calderon v. Thompson, 523 U.S. 538, 555-556 (1998). This provision reduces the potential for delay on the road to finality by restricting the time that a prospective federal habeas petitioner has in which to seek federal habeas review.
The tolling provision of § 2244(d)(2) balances the interests served by the exhaustion requirement and the limitation period. Section 2244(d)(2) promotes the exhaustion of state remedies by protecting a state prisoner's ability later to apply for federal habeas relief while state remedies are being pursued. At the same time, the provision limits the *180 harm to the interest in finality by according tolling effect only to "properly filed application[s] for State post-conviction or other collateral review."
By tolling the limitation period for the pursuit of state remedies and not during the pendency of applications for federal review, § 2244(d)(2) provides a powerful incentive for litigants to exhaust all available state remedies before proceeding in the lower federal courts. But if the statute were construed so as to give applications for federal review the same tolling effect as applications for state collateral review, then § 2244(d)(2) would furnish little incentive for individuals to seek relief from the state courts before filing federal habeas petitions. The tolling provision instead would be indifferent between state and federal filings. While other statutory provisions, such as § 2254(b) itself, of course, would still provide individuals with good reason to exhaust, § 2244(d)(2) would be out of step with this design. At the same time, respondent's interpretation would further undermine the interest in finality by creating more potential for delay in the adjudication of federal law claims.
A diminution of statutory incentives to proceed first in state court would also increase the risk of the very piecemeal litigation that the exhaustion requirement is designed to reduce. Cf. Rose, 455 U. S., at 520. We have observed that "strict enforcement of the exhaustion requirement will encourage habeas petitioners to exhaust all of their claims in state court and to present the federal court with a single habeas petition." Ibid. But were we to adopt respondent's construction of § 2244(d)(2), we would dilute the efficacy of the exhaustion requirement in achieving this objective. Tolling the limitation period for a federal habeas petition that is dismissed without prejudice would thus create more opportunities for delay and piecemeal litigation without advancing the goals of comity and federalism that the exhaustion requirement serves. We do not believe that Congress designed the statute in this manner.
*181 The Court of Appeals reasoned that its interpretation of the statute would further Congress' goal "to spur defendants to file their federal habeas petitions more quickly." 208 F. 3d, at 361. But this view fails to account sufficiently for AEDPA's clear purpose to encourage litigants to pursue claims in state court prior to seeking federal collateral review. See, e. g., §§ 2254(b), 2254(e)(2), 2264(a). Section 2244(d)(1)'s limitation period and § 2244(d)(2)'s tolling provision, together with § 2254(b)'s exhaustion requirement, encourage litigants first to exhaust all state remedies and then to file their federal habeas petitions as soon as possible.
Respondent contends that petitioner's construction of the statute creates the potential for unfairness to litigants who file timely federal habeas petitions that are dismissed without prejudice after the limitation period has expired. But our sole task in this case is one of statutory construction, and upon examining the language and purpose of the statute, we are convinced that § 2244(d)(2) does not toll the limitation period during the pendency of a federal habeas petition.
We also note that, when the District Court dismissed respondent's first federal habeas petition without prejudice, respondent had more than nine months remaining in the limitation period in which to cure the defects that led to the dismissal. It is undisputed, however, that petitioner neither returned to state court nor filed a nondefective federal habeas petition before this time had elapsed. Respondent's May 1997 federal habeas petition also contained claims different from those presented in his April 1996 petition. In light of these facts, we have no occasion to address the alternative scenarios that respondent describes. We also have no occasion to address the question that Justice Stevens raises concerning the availability of equitable tolling.
We hold that an application for federal habeas corpus review is not an "application for State post-conviction or other collateral review" within the meaning of 28 U.S. C. § 2244(d)(2). Section 2244(d)(2) therefore did not toll the *182 limitation period during the pendency of respondent's first federal habeas petition. The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion.
It is so ordered.
Justice Souter, concurring.
Although I join the Court's opinion in full, I have joined Justice Stevens's separate opinion pointing out that nothing bars a district court from retaining jurisdiction pending complete exhaustion of state remedies, and that a claim for equitable tolling could present a serious issue on facts different from those before us.
Justice Stevens, with whom Justice Souter joins, concurring in part and concurring in the judgment.
For substantially the reasons stated in the Court's opinion, ante, at 172-178, I agree that the better reading of 28 U.S. C. § 2244(d)(2) (1994 ed., Supp. V) is that it encompasses only "State" applications for "post-conviction or other collateral review." Thus, as the Court holds, "an application for federal habeas corpus review is not an `application for State post-conviction or other collateral review' within the meaning of 28 U.S. C. § 2244(d)(2)." Ante, at 181. I write separately to add two observations regarding the equitable powers of the federal courts, which are unaffected by today's decision construing a single provision of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), 110 Stat. 1214.
First, although the Court's pre-AEDPA decision in Rose v. Lundy, 455 U.S. 509, 522 (1982), prescribed the dismissal of federal habeas corpus petitions containing unexhausted claims, in our post-AEDPA world there is no reason why a district court should not retain jurisdiction over a meritorious claim and stay further proceedings pending the complete *183 exhaustion of state remedies. Indeed, there is every reason to do so when AEDPA gives a district court the alternative of simply denying a petition containing unexhausted but nonmeritorious claims, see 28 U.S. C. § 2254(b)(2) (1994 ed., Supp. V), and when the failure to retain jurisdiction would foreclose federal review of a meritorious claim because of the lapse of AEDPA's 1-year limitations period.
Second, despite the Court's suggestion that tolling the limitations period for a first federal habeas petition would undermine the "purposes" of AEDPA, see ante, at 178-182, neither the Court's narrow holding, nor anything in the text or legislative history of AEDPA, precludes a federal court from deeming the limitations period tolled for such a petition as a matter of equity. The Court's opinion does not address a federal court's ability to toll the limitations period apart from § 2244(d)(2). See ante, at 181. Furthermore, a federal court might very well conclude that tolling is appropriate based on the reasonable belief that Congress could not have intended to bar federal habeas review for petitioners who invoke the court's jurisdiction within the 1-year interval prescribed by AEDPA.
After all, federal habeas corpus has evolved as the product of both judicial doctrine and statutory law. See generally E. Chemerinsky, Federal Jurisdiction § 15 (3d ed. 1999). In the context of AEDPA's 1-year limitations period, which by its terms runs from "the date on which the judgment became final," see § 2244(d)(1)(A), the Courts of Appeals have uniformly created a 1-year grace period, running from the date of AEDPA's enactment, for prisoners whose state convictions became final prior to AEDPA.[1] Similarly, federal *184 courts may well conclude that Congress simply overlooked the class of petitioners whose timely filed habeas petitions remain pending in district court past the limitations period, only to be dismissed after the court belatedly realizes that one or more claims have not been exhausted.[2] See post, at 186 (Breyer, J., dissenting) (district courts on average take 268 days to dismiss petitions on procedural grounds; 10% remain pending more than 2 years). As a result, equitable considerations may make it appropriate for federal courts to fill in a perceived omission on the part of Congress by tolling AEDPA's statute of limitations for unexhausted federal habeas petitions. Today's ruling does not preclude that possibility, given the limited issue presented in this case and the Court's correspondingly limited holding.[3]
I concur in the Court's holding on the understanding that it does not foreclose either of the above safeguards against the potential for injustice that a literal reading of § 2244(d)(2) might otherwise produce.
*185 Justice Breyer, with whom Justice Ginsburg joins, dissenting.
The federal habeas corpus statute limits the period of time during which a state prisoner may file a federal habeas petition to one year, ordinarily running from the time the prisoner's conviction becomes final in the state courts. See 28 U.S. C. § 2244(d) (1994 ed., Supp. V). Section 2244(d)(2) tolls that 1-year period while "a properly filed application for State post-conviction or other collateral review . . . is pending." The question before us is whether this tolling provision applies to federal, as well as state, collateral review proceedings. Do the words "other collateral review" encompass federal habeas corpus proceedings? I believe that they do.
To understand my conclusion, one must understand why the legal issue before us is significant. Why would a state prisoner ever want federal habeas corpus proceedings to toll the federal habeas corpus limitations period? After all, the very point of tolling is to provide a state prisoner adequate time to file a federal habeas petition. If the prisoner has already filed that petition, what need is there for further tolling?
The answer to this questionand the problem that gives rise to the issue before usis that a federal court may be required to dismiss a state prisoner's federal habeas petition, not on the merits, but because that prisoner has not exhausted his state collateral remedies for every claim presented in the federal petition. See 28 U.S. C. § 2254(b)(1) (1994 ed., Supp. V) (requiring petitioners to exhaust state remedies before filing federal habeas petition); cf. Rose v. Lundy, 455 U.S. 509, 510 (1982) (holding, under predecessor to current § 2254, that district courts cannot reach the merits of "mixed" petitions containing both exhausted and unexhausted claims). Such a dismissal means that a prisoner wishing to pursue the claim must return to state court, *186 pursue his state remedies, and then, if he loses, again file a federal habeas petition in federal court. All this takes time. The statute tolls the 1-year limitations period during the time the prisoner proceeds in the state courts. But unless the statute also tolls the limitations period during the time the defective petition was pending in federal court, the state prisoner may find, when he seeks to return to federal court, that he has run out of time.
This possibility is not purely theoretical. A Justice Department study indicates that 63% of all habeas petitions are dismissed, and 57% of those are dismissed for failure to exhaust state remedies. See U. S. Dept. of Justice, Office of Justice Programs, Bureau of Justice Statistics, Federal Habeas Corpus Review: Challenging State Court Criminal Convictions 17 (1995) (hereinafter Federal Habeas Corpus Review). And it can take courts a significant amount of time to dispose of even those petitions that are not addressed on the merits; on the average, district courts took 268 days to dismiss petitions on procedural grounds. Id., at 23-24; see also id., at 19 (of all habeas petitions, nearly half were pending in the district court for six months or longer; 10% were pending more than two years). Thus, if the words "other collateral review" do not include federal collateral review, a large group of federal habeas petitioners, seeking to return to federal court after subsequent state-court rejection of an unexhausted claim, may find their claims time barred. Moreover, because district courts vary substantially in the time they take to rule on habeas petitions, two identically situated prisoners can receive opposite results. If Prisoner A and Prisoner B file mixed petitions in different district courts six months before the federal limitations period expires, and the court takes three months to dismiss Prisoner A `s petition, but seven months to dismiss Prisoner B `s petition, Prisoner A will be able to return to federal court after exhausting state remedies, but Prisoner B due to no fault of his ownmay not.
*187 On the other hand, if the words "other collateral review" include federal collateral review, state prisoners whose federal claims have been dismissed for nonexhaustion will simply add to the 1-year limitations period the time they previously spent in both state and federal proceedings. Other things being equal, they will be able to return to federal court after pursuing the state remedies that remain available. And similarly situated prisoners will not suffer different outcomes simply because they file their petitions in different district courts.
The statute's language, read by itself, does not tell us whether the words "State post-conviction or other collateral review" include federal habeas proceedings. Rather, it is simply unclear whether Congress intended the word "State" to modify "post-conviction" review alone, or also to modify "other collateral review" (as the majority believes). Indeed, most naturally read, the statute refers to two distinct kinds of applications: (1) applications for "State postconviction" review and (2) applications for "other collateral review," a broad category that, on its face, would include applications for federal habeas review. The majority's reading requires either an unusual intonation"State postconviction-or-other-collateral review "or a slight rewrite of the language, by inserting the word "State" where it does not appear, between "other" and "collateral." Regardless, I believe that either reading is possible. The statute's words, by themselves, have no singular "plain meaning."
Neither do I believe that the various interpretive canons to which the majority appeals can solve the problem. Invoking the principle that "`Congress acts intentionally and purposely in the disparate inclusion or exclusion' " of particular words, Bates v. United States, 522 U.S. 23, 29-30 (1997) (quoting Russello v. United States, 464 U.S. 16, 23 (1983)), the majority attempts to ascertain Congress' intent by looking to the tolling provision's statutory neighbors. It points to other provisions where Congress explicitly used the *188 words "State" and "Federal" together, expressing its intent to cover both kinds of proceedings. See ante, at 172-173 (citing 28 U.S. C. § 2254(i) (1994 ed., Supp. V); § 2261(e); § 2264(a)(3)). And it reasons that Congress' failure to do so here displays a different intent.
But other statutory neighbors show that, when Congress wished unambiguously to limit tolling to state proceedings, "it knew how to do so." Custis v. United States, 511 U.S. 485, 492 (1994). In the special tolling provision governing certain capital cases, Congress said explicitly that the limitations period is tolled "from the date on which the first petition for post-conviction review or other collateral relief is filed until the final State court disposition of such petition, " thus making it clear that federal proceedings, for example, petitions for certiorari, do not count. 28 U.S. C. § 2263(b)(2) (1994 ed., Supp. V) (emphasis added). Does Congress' failure to include a similar qualification in § 2244's tolling provision show that it means that provision to cover both federal and state proceedings? In fact, the "argument from neighbors" shows only that Congress might have spoken more clearly than it did. It cannot prove the statutory point.
The majority also believes that only its interpretation gives effect to every word in the statutein particular the word "State." It asks: If Congress meant to cover federal habeas review, why does the word "State" appear in the statute? Federal habeas proceedings are a form of postconviction proceedings. So, had Congress meant to cover them, it would have just said "post-conviction and other collateral review." See ante, at 174.
But this argument proves too much, for one can ask with equal force: If Congress intended to exclude federal habeas proceedings, why does the word "post-conviction" appear in the statute? State postconviction proceedings are a form of collateral review. So, had Congress meant to exclude federal collateral proceedings, it could have just said "State collateral *189 review," thereby clearly indicating that the phrase applies only to state proceedings.
In fact, this kind of argument, viewed realistically, gets us nowhere. Congress probably picked out "State postconviction" proceedings from the universe of collateral proceedings and mentioned it separately because state postconviction proceedings are a salient example of collateral proceedings. But to understand this is not to understand whether the universe from which Congress picked "State post-conviction" proceedings as an example is the universe of all collateral proceedings, or the universe of state collateral proceedings. The statute simply does not say.
Indeed, the majority recognizes that neither the statute's language nor the application of canons of construction is sufficient to resolve the problem. It concedes that the phrase "other collateral review," if construed as "other [state] collateral review," would add little to the coverage that the words "State post-conviction . . . review" would provide in its absence. See ante, at 176 (noting that a state criminal conviction is "by far the most common" basis for seeking federal habeas review). The majority resolves this difficulty by noting that "other collateral review" could also include either review of state civil confinement proceedings or state postconviction review to which a State refers by some other name, such as state "habeas" proceedings. See ante, at 176-177.
But it is difficult to believe that Congress had state civil proceedings in mind, given that other provisions within § 2244 indicate that Congress saw criminal proceedings as its basic subject matter. For instance, the exceptions to the bar against successive petitions in § 2244(d) seem to presume that the petition at issue challenges a criminal conviction. See 28 U.S. C. § 2244(b)(2)(A) (1994 ed., Supp. V) (requiring a "new rule of constitutional law, made retroactive to cases on collateral review by the Supreme Court"); § 2244(b)(2)(B) (requiring new evidence establishing that, *190 "but for constitutional error, no reasonable factfinder would have found the applicant guilty of the underlying offense"). Nor does it seem likely that Congress would have expected federal courts applying the tolling provision to construe "post-conviction" review to exclude state "habeas" petitions challenging convictions. The statute in which the words "State post-conviction proceedings" appear is a federal statute, and federal courts would be likely to apply those words to whatever state proceedings in fact fall within this federal description, whatever different labels different States might choose to attach. It is simpler, more meaningful, and just as logical to assume that Congress meant the words "other collateral review" to cast a wider neta net wide enough to include federal collateral proceedings such as those that precede a dismissal for nonexhaustion.
Faced with this statutory ambiguity, I would look to statutory purposes in order to reach a proper interpretation. And, while I agree that Congress sought to "`further the principles of comity, finality, and federalism,' " ante, at 178 (quoting Williams v. Taylor, 529 U.S. 420, 436 (2000)), I would also ask whether Congress would have intended to create the kind of "unexhausted petition" problem that I described at the outset. The answer is no. Congress enacted a statute that all agree gave state prisoners a full year (plus the duration of state collateral proceedings) to file a federal habeas corpus petition. Congress would not have intended to shorten that time dramatically, at random, and perhaps erase it altogether, "den[ying] the petitioner the protections of the Great Writ entirely," Lonchar v. Thomas, 517 U.S. 314, 324 (1996), simply because the technical nature of the habeas rules led a prisoner initially to file a petition in the wrong court.
The majority's argument assumes a congressional desire to strengthen the prisoners' incentive to file in state court first. But that is not likely to be the result of today's holding. After all, virtually every state prisoner already *191 knows that he must first exhaust state-court remedies; and I imagine that virtually all of them now try to do so. The problem arises because the vast majority of federal habeas petitions are brought without legal representation. See Federal Habeas Corpus Review 14 (finding that 93% of habeas petitioners in study were pro se ). Prisoners acting pro se will often not know whether a change in wording between state and federal petitions will be seen in federal court as a new claim or a better way of stating an old one; and they often will not understand whether new facts brought forward in the federal petition reflect a new claim or better support for an old one. Insofar as that is so, the Court's approach is likely to lead not to fewer improper federal petitions, but to increased confusion, as prisoners hesitate to change the language of state petitions or add facts, and to greater unfairness. And it will undercut one significant purpose of the provision before usto grant state prisoners a fair and reasonable time to bring a first federal habeas corpus petition.
Nor is it likely that prisoners will deliberately seek to delay by repeatedly filing unexhausted petitions in federal court, as the Court suggests. See ante, at 180. First, prisoners not under a sentence of death (the vast majority of habeas petitioners) have no incentive to delay adjudication of their claims. Rather, "[t]he prisoner's principal interest . . . is in obtaining speedy federal relief." Rose v. Lundy, 455 U. S., at 520. Second, the prisoner who chooses to go into federal court with unexhausted claims runs the risk that the district court will simply deny those claims on the merits, as it is permitted to do, see 28 U.S. C. § 2254(b)(2) (1994 ed., Supp. V), before the prisoner has had the opportunity to develop a record in state court. Third, district courts have the power to prevent vexatious repeated filings by, for instance, ordering that a petition filed after a mixed petition is dismissed must contain only exhausted claims. See Slack v. McDaniel, 529 U.S. 473, 489 (2000). Thus, *192 the interest in reducing "piecemeal litigation," ante, at 180, is not likely to be significantly furthered by the majority's holding.
Finally, the majority's construction of the statute will not necessarily promote comity. Federal courts, understanding that dismissal for nonexhaustion may mean the loss of any opportunity for federal habeas review, may tend to read ambiguous earlier state-court proceedings as having adequately exhausted a federal petition's current claims. For similar reasons, wherever possible, they may reach the merits of a federal petition's claims without sending the petitioner back to state court for exhaustion. To that extent, the majority's interpretation will result in a lesser, not a greater, respect for the state interests to which the majority refers. In addition, by creating pressure to expedite consideration of habeas petitions and to reach the merits of arguably exhausted claims, it will impose a heavier burden on the district courts. (While Justice Stevens' sound suggestions that district courts hold mixed petitions in abeyance and employ equitable tolling, see ante, at 182-184 (opinion concurring in part and concurring in judgment), would properly ameliorate some of the unfairness of the majority's interpretation, they will also add to the burdens on the district courts in a way that simple tolling for federal habeas petitions would not.)
In two recent cases, we have assumed that Congress did not want to deprive state prisoners of first federal habeas corpus review, and we have interpreted statutory ambiguities accordingly. In Stewart v. Martinez-Villareal, 523 U.S. 637 (1998), we held that a federal habeas petition filed after the initial filing was dismissed as premature should not be deemed a "second or successive" petition barred by § 2244, lest "dismissal . . . for technical procedural reasons . . . bar the prisoner from ever obtaining federal habeas review." Id., at 645. And in Slack v. McDaniel, we held that a federal habeas petition filed after dismissal of an initial filing for nonexhaustion should not be deemed a "second *193 or successive petition," lest "the complete exhaustion rule" become a "`trap' " for "`the unwary pro se prisoner.' " 529 U. S., at 487 (quoting Rose, supra, at 520). Making the same assumption here, I would interpret the ambiguous provision before us to permit tolling for federal habeas petitions.
In both Martinez-Villareal and Slack, the Court discerned the purpose of an ambiguous statutory provision by assuming that (absent a contrary indication) congressional purpose would mirror that of most reasonable human beings knowledgeable about the area of the law in question. And the Court kept those purposes firmly and foremost in mind as it sought to understand the statute. See Slack, supra, at 486-487; Martinez-Villareal, supra, at 644 (refusing to adopt an interpretation whose "implications for habeas practice would be far reaching and seemingly perverse"). Today it takes a different approachan approach that looks primarily, though not exclusively, to linguistic canons to dispel the uncertainties caused by ambiguity. Where statutory language is ambiguous, I believe these priorities are misplaced. Language, dictionaries, and canons, unilluminated by purpose, can lead courts into blind alleys, producing rigid interpretations that can harm those whom the statute affects. If generalized, the approach, bit by bit, will divorce law from the needs, lives, and values of those whom it is meant to servea most unfortunate result for a people who live their lives by law's light. The Court was right in Martinez-Villareal and Slack to see purpose as key to the statute's meaning and to understand Congress as intending the same; it is wrong to reverse its interpretive priorities here.
With respect, I dissent.
[*] A brief of amici curiae urging reversal was filed for the Commonwealth of Massachusetts et al. by Thomas F. Reilly, Attorney General of Massachusetts, and Catherine E. Sullivan and William J. Meade, Assistant Attorneys General, and by the Attorneys General for their respective States as follows: Bill Pryor of Alabama, M. Jane Brady of Delaware, Earl Anzai of Hawaii, Thomas J. Miller of Iowa, J. Joseph Curran, Jr., of Maryland, Michael C. Moore of Mississippi, Jeremiah W. (Jay) Nixon of Missouri, Mike McGrath of Montana, Don Stenberg of Nebraska, Frankie Sue Del Papa of Nevada, Betty D. Montgomery of Ohio, D. Michael Fisher of Pennsylvania, Mark Barnett of South Dakota, Paul G. Summers of Tennessee, Mark L. Shurtleff of Utah, Mark L. Earley of Virginia, Christine O. Gregoire of Washington, and Darrel V. McGraw, Jr., of West Virginia.
Leon Friedman and Joshua L. Dratel filed a brief for the National Association of Criminal Defense Lawyers as amicus curiae urging affirmance.
Kent S. Scheidegger filed a brief for the Criminal Justice Legal Foundation as amicus curiae.
[1] See, e. g., Gaskins v. Duval, 183 F.3d 8, 9 (CA1 1999); Ross v. Artuz, 150 F.3d 97, 100-103 (CA2 1998); Burns v. Morton, 134 F.3d 109, 111-112 (CA3 1998); Brown v. Angelone, 150 F.3d 370, 374-376 (CA4 1998); United States v. Flores, 135 F.3d 1000, 1002, n. 7, 1006 (CA5 1998); Austin v. Mitchell, 200 F.3d 391, 393 (CA6 1999); Lindh v. Murphy, 96 F.3d 856, 866 (CA7 1996) (en banc), rev'd on other grounds, 521 U.S. 320 (1997); Ford v.Bowersox, 178 F.3d 522, 523 (CA8 1999); Calderon v. District Court, 128 F.3d 1283, 1286-1287 (CA9 1997), overruled on other grounds, 163 F.3d 530, 539-540 (CA9 1998); Hoggro v. Boone, 150 F.3d 1223, 1225-1226 (CA10 1998); Wilcox v. Florida Dept. of Corrections, 158 F.3d 1209, 1211 (CA11 1998).
[2] The question whether a claim has been exhausted can often be a difficult one, not just for prisoners unschooled in the immense complexities of federal habeas corpus law, see post, at 190-191 (Breyer, J., dissenting), but also for district courts, see, e. g., Morgan v.Bennett, 204 F.3d 360, 369-371 (CA2 2000) (disagreeing with District Court's conclusion that claim had not been exhausted); Bear v. Boone, 173 F.3d 782, 784-785 (CA10 1999) (same).
[3] Thus the court below, which resolved the case based on its reading of 28 U.S. C. § 2244(d)(2) (1994 ed., Supp. V), and which therefore did not reach the question whether it "should exercise its equitable powers to exclude the [time] during which the first [habeas] petition was pending," 208 F.3d 357, 362 (CA2 2000), is free to consider the issue on remand.
