SCALIA, J., delivered the opinion of the Court, in which REHNQUIST, C. J., and O‘CONNOR, SOUTER, THOMAS, GINSBURG, and BREYER, JJ., joined. SOUTER, J., filed a concurring opinion, post, p. 434. GINSBURG, J., filed a concurring opinion, in which SOUTER and BREYER, JJ., joined, post, p. 434. STEVENS, J., filed a dissenting opinion, in which KENNEDY, J., joined, post, p. 436.

James A. Christopherson argued the cause and filed briefs for petitioner. With him on the briefs was Joel R. Myler.

Paul A. Engelmayer argued the cause for the United States. With him on the brief were Solicitor General Days, Acting Assistant Attorney General Keeney, Deputy Solicitor General Dreeben, and David S. Kris.

JUSTICE SCALIA delivered the opinion of the Court.

This case presents the question whether a district court has authority to grant a postverdict motion for judgment of acquittal filed one day outside the time limit prescribed by Federal Rule of Criminal Procedure 29(c).

I

Petitioner Charles Carlisle, along with several co-defendants, was tried by jury in the United States District Court for the Western District of Michigan for conspiracy to possess with intent to distribute marijuana, in violation of 21 U. S. C. §§ 841, 846, 84 Stat. 1260, 1265. He did not move during the trial for a judgment of acquittal under Federal Rule of Criminal Procedure 29(a). On July 13, 1993, the jury returned a guilty verdict and was discharged. On July 23, 1993, Carlisle filed a “Motion for a Judgment of Acquittal Pursuant to Federal Rule of Criminal Procedure 29(c),” arguing that there was insufficient evidence to sustain his conviction. App. 6-9. Rule 29(c) provides that “a motion for judgment of acquittal may be made or renewed within 7 days after the jury is discharged or within such further time as the court may fix during the 7-day period.” Excluding the intermediate Saturday and Sunday (as Federal Rule of Criminal Procedure 45(a) requires), the 7-day period in this case ended on July 22, 1993. The United States’ response to Carlisle‘s motion argued that it should be denied as untimely and, alternatively, that there was sufficient evidence to sustain the conviction. The District Court denied Carlisle‘s motion on August 19, 1993. Its written opinion did not address the timeliness issue, but concluded that the evidence was sufficient for a rational trier of fact to find beyond a reasonable doubt that Carlisle knew about, and knowingly and voluntarily joined, the charged conspiracy.

When Carlisle appeared for sentencing on October 14, 1993, the District Court announced that it was reversing its ruling. When it made its decision in August, the court said, it had prepared two opinions, one granting and one denying the motion, and it had now decided to substitute the former for the latter. The court subsequently entered an order that (i) withdrew the opinion and order denying the motion to acquit and (ii) granted “Carlisle‘s motion for a judgment of acquittal pursuant to Rule 29(c), filed July 23, 1993.” App. 45. An opinion accompanying the order concluded that there was insufficient evidence to prove that Carlisle knowingly and voluntarily joined the conspiracy to possess and distribute marijuana. In a footnote, the opinion acknowledged that the motion for judgment of acquittal was filed one day late, but concluded:

“. . . I can conceive of no prejudice to the United States which will result from consideration of a motion that is one day lat[e] in this case. Because I believe that refusal to hear this motion would result in grave injustice, and because [Rule 29(c)] permits the Court to extend the deadline, I will consider this motion as if it were filed in a timely manner.” Id., at 37.

The United States Court of Appeals for the Sixth Circuit reversed the judgment of acquittal and remanded to the District Court for reinstatement of the jury‘s verdict and for sentencing. It held that under Rule 29 a district court has no jurisdiction to grant an untimely motion for judgment of acquittal, and that a district court has no jurisdiction to enter a judgment of acquittal sua sponte after the case has been submitted to the jury.

48 F. 3d 190, 192 (1995). We granted certiorari.

515 U. S. 1191 (1995).

II

Petitioner argues that district courts “should be given the power to go outside the strict time limits of Federal Rule of Criminal Procedure 29(c)” when (1) there is a claim that the defendant was legally innocent, (2) the motion is filed prior to sentencing, and (3) the motion was not timely filed because of attorney error. Brief for Petitioner 8. Petitioner seeks to root this argument in, among other places, the Federal Rules of Criminal Procedure.

Rule 29 is reproduced in its entirety below.¹ Subdivision (c) provides, in relevant part, that “[i]f the jury returns a verdict of guilty . . . , a motion for judgment of acquittal may

be made or renewed within 7 days after the jury is discharged or within such further time as the court may fix during the 7-day period.” Federal Rule of Criminal Procedure 45(b) provides that whereas certain untimely acts may be accorded validity upon a showing of excusable neglect, “the court may not extend the time for taking any action under Rul[e] 29 . . . except to the extent and under the conditions stated in [the Rule].” These Rules are plain and unambiguous. If, as in this case, a guilty verdict is returned, a motion for judgment of acquittal must be filed, either within seven days of the jury‘s discharge, or within an extended period fixed by the court during that 7-day period. There is simply no room in the text of Rules 29 and 45(b) for the granting of an untimely postverdict motion for judgment of acquittal, regardless of whether the motion is accompanied by a claim of legal innocence, is filed before sentencing, or was filed late because of attorney error.

Unable to offer any reading of Rule 29(c) that would permit an untimely motion for judgment of acquittal to be granted, Carlisle contends that Rule 29(a) gives a district court authority to enter a judgment of acquittal sua sponte at any time before sentencing. Rule 29(a), entitled “Motion Before Submission to Jury,” provides in relevant part:

“The court on motion of a defendant or of its own motion shall order the entry of judgment of acquittal of one or more offenses charged in the indictment or information after the evidence on either side is closed if the evidence is insufficient to sustain a conviction of such offense or offenses.”

It would be quite a surprise to find a district court‘s sua sponte power to grant judgment of acquittal after submission of the case to the jury hidden away in a provision entitled “Motion Before Submission to Jury.” We are not inclined to adopt an interpretation that creates such a surprise unless the intent that the text exceed its caption is clear.

Here, to the contrary, the structure of Rule 29 indicates that subdivision (a) is limited as its caption says.

Petitioner‘s proposed reading would create an odd system in which defense counsel could move for judgment of acquittal for only seven days after the jury‘s discharge, but the court‘s power to enter such a judgment would linger. In

United States v. Smith, 331 U. S. 469 (1947), we declined to read former Federal Rule of Criminal Procedure 33, which placed a 5-day limit on the making of a motion for new trial, as “permit[ting] the judge to order retrial without request and at any time,”

331 U. S., at 473. “[I]t would be a strange rule,” we said, “which deprived a judge of power to do what was asked when request was made by the person most concerned, and yet allowed him to act without petition,” and such an arrangement “would almost certainly subject trial judges to private appeals or application by counsel or friends of one convicted,”

id., at 474, 475. The same is true here.² In addition, petitioner‘s reading makes a farce of subdivision (b) of Rule 29, which provides that a court may reserve decision on the motion for judgment of acquittal and decide it after submission to the jury. There would be no need for this procedure if, even without reserving, the court had continuing power to grant judgment of acquittal on its own. In

sum, even without the captions (and a fortiori with them) it is clear that subdivisions (a) and (b) of Rule 29 pertain to motions made before submission, and subdivisions (c) and (d) to motions made after discharge.

The Government offers an alternative theory of a court‘s power to act sua sponte under Rule 29: Because Rule 29(a) refers to both a “motion of a defendant” and a court‘s “own motion,” whereas Rule 29(c) refers only to “a motion” simpliciter, the latter must refer to motions both of defendants and of courts, permitting both such “motions” to be made within seven days after the jury‘s discharge. We do not find this reading plausible. Rule 29(c) not only provides that “a motion for judgment of acquittal” may be made or renewed within seven days after the jury is discharged. It goes on to provide, in its second and third sentences: “If a verdict of guilty is returned the court may on such motion set aside the verdict and enter judgment of acquittal. If no verdict is returned the court may enter judgment of acquittal.” The phrase “on such motion” is notably absent from the third sentence—conveying the idea that, where a jury has not returned a verdict, a court can act without motion, but where a jury has returned a guilty verdict, it cannot. But if “on such motion” includes action taken by a court on its own initiative, the limiting phrase “on such motion” in the second sentence has no effect, and a court may act on its own whether or not a verdict has been returned. That is to say, the inclusion of the phrase “on such motion” in one sentence but not in the other would be inexplicable.³

Petitioner contends that even if Rule 29 does not permit a court to grant an untimely motion for judgment of acquittal, Federal Rule of Criminal Procedure 2 vests the court with supervisory power to enter judgment of acquittal. Rule 2 provides:

“These rules are intended to provide for the just determination of every criminal proceeding. They shall be construed to secure simplicity in procedure, fairness in administration and the elimination of unjustifiable expense and delay.”

This Rule is of no aid to petitioner. It sets forth a principle of interpretation to be used in construing ambiguous rules, not a principle of law superseding clear rules that do not achieve the stated objectives. It does not, that is to say, provide that rules shall be construed to mean something other than what they plainly say—which is what petitioner‘s proposed construction of Rule 29(c) would require.

We must acknowledge that there is precedent in this Court for using Rule 2 as a basis for deviating from time limits imposed by the Federal Rules of Criminal Procedure. In

Fallen v. United States, 378 U. S. 139 (1964), we cited Rule 2 in the course of excusing the failure of an incarcerated paraplegic pro se petitioner to comply with the time limit for filing a notice of appeal under former Federal Rule of Criminal Procedure 37(a). Concluding that the petitioner “had done all that could reasonably be expected” to file a timely appeal, including mailing a notice of appeal to the clerk‘s office two days before the notice was due, we “decline[d] to read the Rules so rigidly as to bar a determination of his appeal on the merits.”

378 U. S., at 144. Fallen has been made obsolete by an amendment to Rule 37(a).⁴ And

sua sponte, in order to invite such a motion later. It is our hope and belief that no such district judge exists.

of course Fallen was a narrow ruling when it was announced, as is evident from

Berman v. United States, 378 U. S. 530 (1964) (per curiam), a decision announced on the same day as Fallen, summarily affirming the dismissal of an appeal that had been filed one day late.

Finally, petitioner cannot rely on Federal Rule of Criminal Procedure 57 as the source of the District Court‘s authority in this case. The version of Rule 57 in effect when criminal proceedings against petitioner commenced (and which he relied upon at oral argument) states, in relevant part, that, “[i]n all cases not provided for by rule, the district judges . . . may regulate their practice in any manner not inconsistent with these rules.” The relevant portion of the current version of Rule 57 is captioned “Procedure When There Is No Controlling Law,” and states: “A judge may regulate practice in any manner consistent with federal law, these rules, and local rules of the district.” Fed. Rule Crim. Proc. 57(b). We need not decide which version of this Rule controls the present case, because neither authorizes the District Court‘s action here. A rule permitting a party to submit and prevail on an untimely motion for judgment of acquittal is “inconsistent” (or not “consistent“) with Rule 29‘s 7-day filing limit; and the question of when a motion for judgment of acquittal may be granted does not present a case “not provided for” by Rule 29; and Rule 29 is the “controlling law” governing this question.

III

As alternative authority for the District Court‘s action, petitioner invokes courts’ “inherent supervisory power.” Brief for Petitioner 9. We have recognized that federal

neglect.” See Fed. Rule Crim. Proc. 37(a) (1966). When Rule 37(a) was abrogated and replaced by Federal Rule of Appellate Procedure 4(b), the substance of this amendment was transferred to Rule 4(b). See Fed. Rule App. Proc. 4(b) (1968).

courts “may, within limits, formulate procedural rules not specifically required by the Constitution or the Congress.”

United States v. Hasting, 461 U. S. 499, 505 (1983). Whatever the scope of this “inherent power,” however, it does not include the power to develop rules that circumvent or conflict with the Federal Rules of Criminal Procedure. As we recognized in

Bank of Nova Scotia v. United States, 487 U. S. 250, 254-255 (1988), holding that federal courts may not invoke supervisory power to circumvent Rule 52(a): “[F]ederal courts have no more discretion to disregard the Rule‘s mandate than they do to disregard constitutional or statutory provisions.” Whether the action of the District Court here is described as the granting of an untimely motion, or the sua sponte entry of a judgment of acquittal, it contradicted the plain language of Rule 29(c), and effectively annulled the 7-day filing limit.

In

Chambers v. NASCO, Inc., 501 U. S. 32, 47 (1991), we said that we would not “lightly assume that Congress has intended to depart from established principles’ such as the scope of a court‘s inherent power,” id., at 47 (quoting

Weinberger v. Romero-Barcelo, 456 U. S. 305, 313 (1982)). Similarly, in

Link v. Wabash R. Co., 370 U. S. 626, 629-632 (1962), we said that since a district court‘s authority to dismiss sua sponte for lack of prosecution was a “sanction of wide usage,” we would not assume, in the absence of a clear expression, that Federal Rule of Civil Procedure 41(b), which allowed a party to move for dismissal for lack of prosecution, abrogated this “long . . . unquestioned” power. That cautionary principle does not apply in the present case, not only because of the clarity of the text, but also because we are unaware of any “long unquestioned” power of federal district courts to acquit for insufficient evidence sua sponte, after return of a guilty verdict. Indeed, we are aware of only two cases prior to the enactment of the Federal Rules of Criminal Procedure that could be read as asserting in dictum the existence of such a power.

United States v. McCracken, 26 F. Cas. 1069, 1069 (No. 15,664) (ED Va. 1878);

United States v. Hayden, 26 F. Cas. 236, 238 (No. 15,333) (NDNY 1877).⁵

The case law of this Court that petitioner relies upon does not establish any “inherent power” to act in contravention of applicable Rules. In

Gaca v. United States, 411 U. S. 618 (1973) (per curiam), which reinstated an appeal that had been dismissed for want of timely prosecution, there was no suggestion that reinstatement was contrary to any statute or rule of procedure. And in

United States v. Nobles, 422 U. S. 225 (1975), which approved exercise of a District Court‘s inherent authority to order the disclosure of certain witness statements, we felt it necessary to make sure that such exercise did not conflict with Federal Rule of Criminal Procedure 16. Petitioner‘s best case is

Thompson v. INS, 375 U. S. 384 (1964), which, contrary to former Federal Rule of Civil Procedure 73(a), gave effect to a notice of appeal filed more than 60 days from the entry of judgment. Thompson, however, is not pertinent here, since it expressly relied upon the “‘unique circumstances‘” that the cause of the failure to meet the Rule‘s deadline was an erroneous ruling or assurance by the District Court itself.

375 U. S., at 387 (quoting

Harris Truck Lines, Inc. v. Cherry Meat Packers, Inc., 371 U. S. 215, 217 (1962) (per curiam)).

IV

Petitioner‘s three remaining arguments need not detain us long. First, he argues that the District Court had power to enter a judgment of acquittal in this case under the All Writs Act, 28 U. S. C. § 1651, through the writ of coram nobis. Apart from the fact that the District Court was not asked to

verdict. The dissent apparently thinks it an adequate explanation for this lack of support that, prior to our decision in

United States v. Smith, 331 U. S. 469, 474 (1947) (suggesting that sua sponte grant of a new trial may raise double jeopardy concerns), district courts could order new trials where there was insufficient evidence to sustain the jury verdict. Post, at 442-443. But if these district courts truly had latent inherent power to enter a judgment of acquittal, surely at least some of them would have been willing to give a legally innocent defendant that to which he was entitled—viz., a judgment of acquittal rather than just a new trial.

issue, and did not purport to be issuing, a writ of coram nobis, that writ would not have lain here, since it was traditionally available only to bring before the court factual errors “material to the validity and regularity of the legal proceeding itself,” such as the defendant‘s being under age or having died before the verdict. See

United States v. Mayer, 235 U. S. 55, 67-68 (1914). Moreover, “[t]he All Writs Act is a residual source of authority to issue writs that are not otherwise covered by statute. Where a statute specifically addresses the particular issue at hand, it is that authority, and not the All Writs Act, that is controlling.”

Pennsylvania Bureau of Correction v. United States Marshals Service, 474 U. S. 34, 43 (1985). As we noted a few years after enactment of the Federal Rules of Criminal Procedure, “it is difficult to conceive of a situation in a federal criminal case today where [a writ of coram nobis] would be necessary or appropriate.”

United States v. Smith, 331 U. S., at 475, n. 4. In the present case, Rule 29 provides the applicable law.

Second, petitioner asserts that the failure to allow the District Court to enter a judgment of acquittal would violate the Due Process Clause of the Fifth Amendment. His argument on this point consists of nothing more than bald assertions that Rule 29(c) as applied to the facts of this case transgresses principles of fundamental fairness, “shocks the conscience,” and interferes with rights “implicit in the concept of ordered liberty.” Brief for Petitioner 28-29 (internal quotation marks omitted) (citing

Herrera v. Collins, 506 U. S. 390 (1993);

Rochin v. California, 342 U. S. 165, 172 (1952);

Palko v. Connecticut, 302 U. S. 319, 325-326 (1937)). Petitioner has failed to proffer any historical, textual, or controlling precedential support for his argument that the inability of a district court to grant an untimely postverdict motion for judgment of acquittal violates the Fifth Amendment, and we decline to fashion a new due process right out of thin air.

Third, petitioner argues that prohibiting a district court from granting a motion for judgment of acquittal filed one day late will lead to needless appeals and habeas corpus proceedings, where it will be more difficult for defendants to obtain relief than in motions directed to the trial court. Assuming, arguendo, that these contentions are accurate, we cannot permit them to alter our analysis, for we are not at liberty to ignore the mandate of Rule 29 in order to obtain “optimal” policy results. Cf.

United States v. Robinson, 361 U. S. 220, 229-230 (1960). We are similarly unmoved by petitioner‘s contention that the “rationale” behind Rule 29(c)‘s time limit does not apply where the motion for judgment of acquittal is filed a mere eight days after the trial. The only evident “rationale” behind Rule 29(c)‘s 7-day time limit is that a motion for judgment of acquittal filed eight days after trial is a motion filed one day later than justice and equity demand. As we said in a case involving the filing deadline of the Federal Land Policy and Management Act of 1976, 43 U. S. C. § 1744 (1988 ed.): “If 1-day late filings are acceptable, 10-day late filings might be equally acceptable, and so on in a cascade of exceptions that would engulf the rule erected by the filing deadline; yet regardless of where the cutoff line is set, some individuals will always fall just on the other side of it.”

United States v. Locke, 471 U. S. 84, 101 (1985).

V

Finally, we may respond to some of the many arguments put forward by the dissent. The dissent makes the sweeping assertion that “a district court clearly has the inherent authority to ensure that a legally innocent defendant is not wrongfully convicted,” post, at 442. Perhaps so. As the dissent itself recognizes, however, that power has come to an end once an appeal has been taken. Post, at 452-453. We are in accord, then, that there is some point at which the district court is rendered powerless to enter a judgment of acquittal, and the disagreement between us and the dissent comes down to nothing more cosmic than the question of timing—which we find answered by the text of Rule 29.

In an effort to explain why, if a Rule 29(c) motion is in any event unnecessary, it makes any sense to impose a 7-day deadline upon the making of it, the dissent maintains that the untimeliness of a motion gives a district court discretion to ignore it. Post, at 445. This presents the disedifying prospect of a court vested with “the inherent authority to ensure that a legally innocent defendant is not wrongfully convicted,” post, at 442, exercising its discretion to let an innocent defendant be wrongfully convicted. Quite obviously, this explanation of the deadline is incompatible with the premise that underlies the dissent‘s entire argument. As for the dissent‘s concern, post, at 448, that our decision runs afoul of Rule 2‘s mandate that the rules “be construed to secure simplicity in procedure, fairness in administration and the elimination of unjustifiable expense and delay“: We see neither simplicity, nor fairness, nor elimination of delay in a regime that makes it discretionary whether an untimely motion for judgment of acquittal will be entertained.

The dissent asserts that “permissive rules do not withdraw pre-existing inherent powers.” Post, at 452. That assertion is really not relevant to the present case since, as we have discussed, the power to enter postverdict judgments of acquittal sua sponte was not a “pre-existing inherent power.” See supra, at 426-428, and n. 5. But besides the lack of factual predicate for its application here, the principle the dissent proposes would produce some extraordinary consequences. For example, as the cases cited by the dissent illustrate, see post, at 439-440, courts previously have ordered new trials sua sponte. Federal Rule of Criminal Procedure 33, however, provides that “[t]he court on motion of a defendant may grant a new trial . . . .” Following the dissent‘s logic, Rule 33, being permissive, does not preclude a court from granting a new trial without motion, thereby leaving open to the court a course of action that may well violate the Double Jeopardy Clause. But see Advisory Committee‘s Notes on 1966 Amendment of Fed. Rule Crim. Proc. 33, 18 U. S. C. App., p. 801 (“The amendments to the first two sentences make it clear that a judge has no power to order a new trial on his own motion, that he can act only in response to a motion timely made by a defendant. Problems of double jeopardy arise when the court acts on its own motion“). Similarly, a pre-existing practice, if there was one, would allow a subpoena to be served by a party or a minor despite Federal Rule of Criminal Procedure 17(d) (“A subpoena may be served by the marshal, by a deputy marshal or by any other person who is not a party and who is not less than 18 years of age“); would allow a judge from another district to take over a jury trial from a disabled judge despite Federal Rule of Criminal Procedure 25(a) (“If . . . the judge before whom a jury trial has commenced is unable to proceed with the trial, any other judge regularly sitting in or assigned to the court . . . may proceed with and finish the trial“); and would allow a court to correct a technical error in a sentence more than seven days after the imposition of the sentence, despite Federal Rule of Criminal Procedure 35(c) (“The court, acting within 7 days after the imposition of sentence, may correct a sentence that was imposed as a result of arithmetical, technical, or other clear error“).

The decisions of Justice Harlan relied upon by the dissent to support the proposition that permissive rules do not eliminate inherent powers are not germane. We have discussed Link above, see supra, at 426. In

United States v. Ohio Power Co., 353 U. S. 98, 104 (1957), Justice Harlan noted that this Court has proceeded on the assumption that we have inherent authority to “affect judgments by action which would otherwise be out of time under [our own] Rules.” That statement would be relevant if the present case involved a district court‘s departure from one of its own rules—which of course it does not. In

Fernandez v. United

States, 81 S. Ct. 642 (1961), 5 L. Ed. 2d 683 (Harlan, J., in chambers), Justice Harlan recognized that the provision of former Federal Rule of Criminal Procedure 46(a) that a “person arrested for an offense not punishable by death shall be admitted to bail” (emphasis added) did not withdraw district courts’ authority to revoke bail in a noncapital case.

Fernandez, supra, at 644, and n. 7, 5 L. Ed. 2d, at 685, and n. 7. What admitting to bail implies with respect to revocation of bail is not comparable to what granting judgment on motion implies with respect to granting judgment without motion. What the dissent needs, in the Fernandez context, is a case holding that a statute which permits bail for “persons arrested for noncapital offenses” does not preclude bail for persons arrested for capital offenses. Of course, such a case will not be found.

Finally, the dissent contends that

United States v. Sisson, 399 U. S. 267 (1970), supports existence of the “inherent power” petitioner invokes. See post, at 448-449. We think not. Sisson did not “implicitly conclude” that it was proper to enter a postverdict judgment of acquittal without motion, because the propriety of the judgment of acquittal was irrelevant to the decision. The only issue was whether the judgment appealed from was a judgment of acquittal (proper or improper), because that would mean that the Government‘s appeal under the former 18 U. S. C. § 3731 (which did not apply to judgments of acquittal) must be dismissed. See

United States v. Wilson, 420 U. S. 332, 351 (1975) (appeal in Sisson “was barred solely by the statute“).

*

*

*

We conclude that the District Court had no authority to grant petitioner‘s motion for judgment of acquittal filed one day outside the time limit prescribed by Rule 29(c). We therefore affirm the judgment of the Sixth Circuit.

It is so ordered.

JUSTICE SOUTER, concurring.

In Part I of his dissenting opinion, JUSTICE STEVENS makes a persuasive argument that, absent a rule to the contrary, district judges have an “inherent authority” to enter a judgment of acquittal, although, for the reasons offered by the majority, ante, at 426, I am not persuaded that this inherent authority extends to the power to act sua sponte to grant a judgment of acquittal after the jury has returned a verdict. In any event, I accept the received view that inherent power generally is subject to legislative abrogation, see

Bank of Nova Scotia v. United States, 487 U. S. 250, 254-255 (1988); ante, at 426, and although Congress‘s power is not necessarily plenary, its limits are not implicated here. While there may be some point at which legislative interference with a court‘s inherent authority would run afoul of Article III, see

Chambers v. NASCO, Inc., 501 U. S. 32, 58 (1991) (SCALIA, J., dissenting) (“Some elements of that inherent authority are so essential to ‘[t]he judicial Power,’ U. S. Const., Art. III, § 1, that they are indefeasible“), it is not seriously contended that Rule 29(c) is an unconstitutional interference with the court‘s inherent authority. I therefore join the Court‘s opinion.

JUSTICE GINSBURG, with whom JUSTICE SOUTER and JUSTICE BREYER join, concurring.

I join the opinion of the Court and highlight features of the case key to my judgment.

It is anomalous to classify time prescriptions, even rigid ones,^* under the heading “subject matter jurisdiction.” That most basic requirement relates to the subject matter of the case or controversy or the status of the parties to it. See 13 C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure § 3522, p. 78 (2d ed. 1984); Restatement (Second) of Judgments § 11 (1982) (defining “subject matter juris-

diction” as the “authority [of the court] to adjudicate the type of controversy involved in the action“); cf.

United States v. Kember, 648 F. 2d 1354, 1357-1358 (CADC 1980) (per curiam) (commenting on “manifold settings in which we employ the term [jurisdiction]” and distinguishing fundamental “jurisdiction” questions from issues of a less basic character);

Center for Nuclear Responsibility, Inc. v. United States Nuclear Regulatory Comm‘n, 781 F. 2d 935, 945, n. 4 (CADC 1986) (Ginsburg, J., dissenting) (questioning “profligate use” of the word “jurisdiction,” in diverse contexts, “to mean many things—from the absence of a constitutional grant of judicial power to a statutory limit on time to appeal“).

Federal Rule of Criminal Procedure 29(c) concerns a matter less basic. It is simply a time prescription. Rule 29(c)‘s prescription is a tight one, to be sure. Federal Rule of Criminal Procedure 45(b) makes that clear by precluding extensions, even for “excusable neglect,” after expiration of the seven days specified in Rule 29(c). But like limitation periods generally, see, e. g.,

Irwin v. Department of Veterans Affairs, 498 U. S. 89, 95 (1990) (“[t]ime requirements in lawsuits . . . are customarily subject to ‘equitable tolling‘“), the 29(c)/45(b) constraint is not utterly exceptionless.

This Court has recognized one sharply honed exception to rules of the 29(c)/45(b) genre. That exception covers cases in which the trial judge has misled a party who could have—and probably would have—taken timely action had the trial judge conveyed correct, rather than incorrect, information. See

Thompson v. INS, 375 U. S. 384, 386-387 (1964) (per curiam) (had trial judge not misinformed party that his new trial motion was made “in ample time,” party “could have, and presumably would have, filed the appeal within 60 days of the entry of the original judgment, rather than waiting, as he did, until after the trial court had disposed of the [new trial motion]“);

Harris Truck Lines, Inc. v. Cherry Meat Packers, Inc., 371 U. S. 215, 216-217 (1962) (per curiam) (instructing that petitioner‘s appeal be heard on the merits

where petitioner had received from trial court an improperly grounded 14-day extension of the time to file his appeal); see also 4A Wright & Miller, Federal Practice and Procedure § 1168, at 501 (describing

Thompson and

Harris Truck Lines as “based on a theory similar to estoppel“).. As the Court observes, however, this exception “is not pertinent here.” See ante, at 428.

Carlisle‘s counsel was not misled by any trial court statement or action; rather, he neglected to follow plain instructions. Rule 29(c) clearly instructs that a motion for a judgment of acquittal be filed “within 7 days after the jury is discharged or within such further time as the court may fix during the 7-day period.” Just as clearly, Rule 45(b) excludes motions for enlargement once seven days have run. I agree that a rule like 29(c) is framed to resist ad hoc relaxation. A time line must be drawn at some point, and it is not unreasonable to draw the line as the rulemakers did, rather than extend it out to the day set for sentencing.

It bears emphasis, finally, that the Government recognizes legal avenues still open to Carlisle to challenge the sufficiency of the evidence to warrant his conviction: on appeal (subject to “plain error” standard); and through a postconviction motion, under 28 U.S.C. § 2255, asserting ineffective assistance of counsel. Brief for United States 38-39. In the rare situations JUSTICE STEVENS envisions-delay of a meritorious motion due to an Act of God, see post, at 454, or cases comparable to those in which we would read and grant an out-of-time rehearing petition, see post, at 450-451-these modes of relief should provide an adequate corrective.

JUSTICE STEVENS, with whom JUSTICE KENNEDY joins, dissenting.

As long as a federal court retains jurisdiction over a criminal case, it has the authority to ensure that no conviction is entered unless the prosecutor has proved the defendant‘s guilt. The exercise of the court‘s inherent power to set aside a jury verdict unsupported by evidence is not contingent on the filing of a timely motion by the defendant. The question in this case, therefore, is not whether Rule 29 of the Federal Rules of Criminal Procedure authorizes the court to grant an untimely motion for judgment of acquittal; I agree with the Court that it does not. Rather, the question is whether that Rule withdraws the court‘s pre-existing authority to refrain from entering judgment of conviction against a defendant whom it knows to be legally innocent.

Viewed in this light, the majority places more reliance on the negative implication in Rule 29 than its permissive language can bear. Assuming it exists at all, this negative implication is far too weak to justify the conclusion that Rule 29 manifests that Congress desired to withdraw a federal court‘s inherent authority to acquit an innocent defendant.

I

Trial judges are kept busy responding to motions, objections, and requests by the litigants. It is quite wrong, however, to assume that a judge is nothing more than a referee whose authority is limited to granting or denying motions advanced by the parties. As Learned Hand tersely noted, a “judge, at least in a federal court, is more than a moderator; he is affirmatively charged with securing a fair trial, and he must intervene sua sponte to that end, when necessary.”

Brown v. Walter, 62 F.2d 798, 799 (CA2 1933). That duty encompasses not only the avoidance of error before it occurs, but the correction of error that may have occurred earlier in a proceeding.

The basic principle has been stated many times. There is a “‘power inherent in every court of justice so long as it retains control of the subject matter and of the parties, to correct that which has been wrongfully done by virtue of its process.’

Arkadelphia Co. v. St. Louis Southwestern Ry. Co., 249 U.S. 134, 146. See

Northwestern Fuel Co. v. Brock, 139 U.S. 216, 219.”

United States v. Morgan, 307 U.S. 183, 197 (1939). Although that statement was made in a civil case, we have made it clear that a federal court has even broader discretion to notice error independently in the trial of a criminal case than in civil cases.

Crawford v. United States, 212 U.S. 183, 194 (1909).

Examples of the exercise of the federal courts’ inherent powers are abundant in both our civil and our criminal jurisprudence.¹ Indeed, when he was serving on the Court of Appeals for the Ninth Circuit, then-Judge Kennedy, after considering a series of cases that recognized various inherent judicial powers,² correctly pointed out:

“Exercise of judicial power by entry of orders not expressly sanctioned by rule or statute in order to correct the legal process or avert its misfunction has been approved in varied circumstances.”

Arizona v. Manypenny, 672 F.2d 761, 765, cert. denied,

459 U.S. 850 (1982).

When a federal court declines to enter a judgment of conviction against a defendant whom it should have directed the jury to acquit, it clearly corrects the legal process and averts its misfunctioning. Given the various sua sponte powers that district courts unquestionably may exercise in order to ensure that legally innocent defendants are not convicted, it is clear that they also possess the inherent authority sua sponte to enter postverdict acquittals when the Government has failed to prove that a defendant is guilty.

District courts have long exercised their inherent power to direct an acquittal sua sponte when the prosecution fails to prove its case at the close of evidence. See

Wiborg v. United States, 163 U.S. 632, 659 (1896);

Cady v. United States, 293 F. 829 (CADC 1923);

Nosowitz v. United States, 282 F. 575, 578 (CA2 1922).³ They have also long exercised their inherent power to set aside a jury verdict for insufficiency of the evidence sua sponte. See

United States v. Harding, 26 F. Cas. 131, 136 (No. 15,301) (ED Pa. 1846);

United States v. Fullerton, 25 F. Cas. 1225 (No. 15,176) (SDNY 1870); see also F. Wharton, Criminal Law of the United States 669 (1846) (“Where, however, evidence is not sufficient in law to authorize a verdict, a new trial will be granted, even though no objection be made at the trial“); id., at 643(s) (explaining that the judge reserves “it to himself, if there be an improper conviction, to arrest the judgment or set aside the verdict“);

Charles v. State, 4 Port. 107, 109-110 (Ala. 1836).⁴

The District Courts’ longstanding exercise of these inherent powers is entirely consistent with the conclusion that a district court acts within its power when it enters a judgment of acquittal upon setting aside an unsupported jury verdict. To be sure, the early cases reveal that District Courts typically ordered new trials, rather than acquittals, upon concluding that the jury‘s verdict was not supported by legally sufficient evidence. However, subsequent cases demonstrate that as courts became concerned that the new trial remedy trenched on the prohibition against double jeopardy, they began to enter judgments of acquittals. See

Ex parte United States, 101 F.2d 870, 878 (CA7 1939), aff‘d by an equally divided Court in

United States v. Stone, 308 U.S. 519 (1939).

The earliest cases involve appellate courts entering judgments of acquittal in order to remedy a district court‘s failure to direct the jury to acquit. See

Nosowitz v. United States, 282 F. 575 (CA2 1922);

Cherry v. United States, 78 F.2d 334 (CA7 1935);

Reiner v. United States, 92 F.2d 823 (CA9 1937); see also

France v. United States, 164 U.S. 676 (1897) (remanding to the District Court with directions to enter such judgment);

Romano v. United States, 9 F.2d 522 (CA2 1925) (same). Later cases reveal that District Courts soon followed suit, either by ruling on reserved, preverdict acquittal motions or by granting postverdict motions to acquit. See

Ex parte United States, 101 F.2d 870 (CA7 1939);

United States v. Standard Oil Co., 23 F. Supp. 937 (WD Wis. 1938), aff‘d in

United States v. Socony-Vacuum Oil Co., 310 U.S. 150, 165, n. 1 (1940);

State v. Meen, 171 Wis. 36 (1920) (same); see also Advisory Committee‘s Notes to Rule 29 (endorsing these practices). Moreover, prior to the adoption of Rule 29 in 1944, the Fifth Circuit explained that, even after a jury returns a verdict, a court “may and should” sua sponte review the sufficiency of the evidence.

Ansley v. United States, 135 F.2d 207, 208 (1943).

In light of this history, it makes no sense to conclude that a federal district court lacks the inherent power to enter sua sponte a postverdict judgment of acquittal. A trial court‘s postverdict entry of a judgment of acquittal is in substance no different from an appellate court‘s order directing entry of that same judgment. Moreover, the double jeopardy concerns that may bar a district court from ordering a new trial to remedy its failure to have directed an acquittal cannot sensibly be understood to prohibit the district court from providing a defendant some measure of relief from a legally insufficient guilty verdict. See

United States v. Smith, 331 U.S. 469, 474 (1947). Finally, given that a motion was not thought to be needed in order for the District Court to exercise its inherent power either to direct an acquittal, or to set aside an unsupported verdict and order a new trial, there is no reason to conclude that a district court is utterly powerless to remedy a wrongful conviction in the exceedingly rare circumstance in which an unforeseen accident results in the defendant‘s failure to file a motion for acquittal.

In all events, a district court clearly has the inherent authority to ensure that a legally innocent defendant is not wrongfully convicted. It would be most strange to conclude that this authority, which enables a district court to keep a case from the jury altogether when the Government fails to prove its case, does not permit that same court to revise a guilty verdict that the jury returns despite the Government‘s insufficient proof. That conclusion is particularly difficult to fathom when one considers that the latter action may be appealed by the Government, while the former may not.

United States v. Wilson, 420 U.S. 332, 345 (1975). Not surprisingly, therefore, numerous courts have recognized that, prior to the passage of Rule 29, district courts possessed the inherent power to acquit defendants sua sponte. See

United States v. Hughes, 759 F. Supp. 530, 532-536 (WD Ark.), aff‘d sub nom.

United States v. Haren, 952 F.2d 190 (CA8 1991);

United States v. DiBernardo, 880 F.2d 1216, 1225, n. 4 (CA11 1989);

United States v. Coleman, 811 F.2d 804 (CA3 1987);

United States v. Giampa, 758 F.2d 928, 936, n. 1 (CA3 1985);

Arizona v. Manypenny, 672 F.2d, at 765;

Ansley v. United States, 135 F.2d, at 208; see also

United States v. Weinstein, 452 F.2d 704, 713, 714 (CA2 1971);

United States v. Broadus, 664 F. Supp. 592, 595-598 (DC 1987).

The majority states that no pre-Rule case establishes the precise power at issue here. Ante, at 427-428, n. 5. That is true but unremarkable. The majority does not dispute that, prior to the passage of Rule 29, trial courts possessed the inherent power to remedy unsupported guilty verdicts by ordering new trials sua sponte. After Rule 29 was adopted, this Court pointed out the double jeopardy concerns raised by the sua sponte exercise of the new trial remedy. See

United States v. Smith, 331 U.S., at 474. Since that time, numerous cases have concluded that courts may remedy unsupported jury verdicts by entering judgments of acquittal. The majority offers no principled reason for concluding that this more recent remedy is beyond the power of district courts, even though the prior remedy was not.

In sum, the error-correcting power that is “inherent in every court of justice so long as it retains control of the subject matter and of the parties,”

Morgan, 307 U.S., at 197, encompasses the kind of error at issue in this case. Therefore, absent some express indication that Congress intended to withdraw the power that implicitly attends its initial grant of jurisdiction, a district court acts well within its discretion when it sets aside a jury verdict and acquits a defendant because the prosecution failed to prove its case.

II

Because the Acts of Congress investing federal judges with jurisdiction to try criminal cases are the source of a district court‘s power to set aside unsupported jury verdicts, I have no occasion to disagree with the Court‘s view that petitioner errs in relying on Rule 29 as the source of the District Court‘s authority in this case. I do, however, strongly disagree with the Court‘s own reliance on that Rule for the quite different conclusion that it clearly prohibits the power exercised by the District Court here.

In Part III of its opinion, the majority asserts that the District Court‘s action “contradicted the plain language of Rule 29(c), and effectively annulled the 7-day filing limit,” ante, at 426, and that “the clarity of the text” suffices to prohibit the District Court‘s action, ibid. The majority assumes that these conclusory assertions follow implicitly from its determination in Part II of its opinion that Rule 29 does not authorize the District Court to set aside a jury verdict sua sponte.

In my view, the Rule serves three salutary purposes that are in no tension with a district court‘s inherent power to enter a judgment of acquittal sua sponte. None of these purposes would be frustrated if the Rule were understood to coexist with, though not to authorize, a district court‘s power to avoid imposing sentence on an innocent defendant in the truly exceptional case in which evidence of guilt is wholly lacking.

First, subdivision (a) confirms the view that a judge has a duty to direct an acquittal if the prosecution has failed to prove its case at the close of evidence. The Rule‘s affirmation of that duty is in no way inconsistent with a court‘s exercise of its postverdict power to enter sua sponte a judgment of acquittal. As then-Judge Kennedy explained for the Ninth Circuit in

Arizona v. Manypenny, 672 F.2d, at 764: “We do not read the mention in Rule 29(a) of a court granting such a judgment ‘on its own motion’ before submission to a jury as an elimination of a court‘s inherent power to grant such a judgment after submission to the jury.”

Second, subdivision (b) accommodates the defendant‘s right to move for a directed acquittal with the Government‘s right to seek appellate review. Indeed, the subdivision was amended in 1994 for the very purpose of striking a more proper balance between those two interests. See Advisory Committee‘s Notes to Fed. Rule Crim. Proc. 29(b), 18 U.S.C. App., pp. 784-785. As a result, a district court‘s sua sponte decision to acquit after the jury returns a guilty verdict can hardly be said to undermine the purpose of subdivision (b). The defendant‘s interests are obviously fully protected by an acquittal, while the Government‘s right to appeal is pro-tected because the jury has already returned its verdict of guilt. See

United States v. Wilson, 420 U.S., at 345.⁵

Third, subdivision (c) requires defense counsel to file the postverdict motion for judgment of acquittal promptly, while the trial judge presumably retains a firm recollection of the evidence and therefore is able to rule expeditiously and efficiently. The untimeliness of a later motion provides the judge with a sufficient reason for denying it without even reading it or reviewing the transcript. Thus, a judge‘s entirely discretionary decision to enter sua sponte an acquittal after the 7-day period in no way annuls the 7-day deadline. Defendants are still bound by that time limitation, and the Rule thus serves the useful function of limiting a defendant‘s right to require a judge to reconsider the sufficiency of the evidence. As then-Judge Kennedy explained: ”Rule 29(c) creates a deadline by which defendants must present motions for judgment of acquittal to the court; it does not address the court‘s inherent power to grant such a judgment.”

Arizona v. Manypenny, 672 F.2d, at 764.

The majority nevertheless maintains that the Rule must be read to require judges, in some instances, to enter judgments of conviction against defendants they know to be innocent. The majority does not argue that Rule 29 expressly prohibits a district court from acting on its own to set aside an unsupported jury verdict. Rather, it relies solely on the negative inference that it draws from the absence of three words in one sentence of Rule 29(c). Ante, at 423.⁶ Specifically, the majority seizes upon the “notabl[e] absen[ce]” of the phrase “on such motion” in the third sentence of the Rule, ibid., and concludes that this omission “convey[s] the idea that, where a jury has not returned a verdict, a court can act without motion, but where a jury has returned a guilty verdict, it cannot,” ibid.

In light of the pre-Rule precedent establishing a district court‘s inherent power to review sua sponte a jury verdict for sufficiency of the evidence, see

Ansley v. United States, 135 F.2d, at 208, the majority reads far too much into the omission. The caption to Rule 29(c) makes clear that the subdivision only contemplates judicial action taken in response to a motion. The first sentence explains that a motion may be made after a jury‘s discharge whether or not a guilty verdict has been returned. The next sentence sets forth the action that the district court may take when such a motion is filed after the jury returns a guilty verdict. In a similar vein, the third sentence sets forth the action that the district court may take when no verdict has been re-turned. The omission of the words “on such motion” from the third sentence surely just reflects a draftsman‘s sensible decision to avoid a patent redundancy rather than a cryptic intent to change the law by prohibiting a judge from exercising his or her inherent power to enter a judgment of acquittal.⁷