sistent with the typical state provision on this subject; see, e.g., Ill.S.Ct. Rule 402(f).

The language of the amendment identifies with more precision than the present language the necessary relationship between the statements and the plea or discussion. See the dispute between the majority and concurring opinions in United States v. Herman, 544 F.2d 791 (5th Cir. 1977), concerning the meanings and effect of the phrases "connection to" and "relevant to" in the present rule. Moreover, by relating the statements to "plea discussions" rather than "an offer to plead," the amendment ensures "that even an attempt to open plea bargaining [is] covered under the same rule of inadmissibility." United States v. Brooks, 536 F.2d 1137 (6th Cir. 1976).

The last sentence of Rule 11(e)(6) is amended to provide a second exception to the general rule of nonadmissibility of the described statements. Under the amendment, such a statement is also admissible "in any proceeding wherein another statement made in the course of the same plea or plea discussions has been introduced and the statement ought in fairness be considered contemporaneously with it." This change is necessary so that, when evidence of statements made in the course of or as a consequence of a certain plea or plea discussions are introduced under circumstances not prohibited by this rule (e.g., not "against" the person who made the plea), other statements relating to the same plea or plea discussions may also be admitted when relevant to the matter at issue. For example, if a defendant upon a motion to dismiss a prosecution on some ground were able to admit certain statements made in aborted plea discussions in his favor, then other relevant statements made in the same plea discussions should be admissible against the defendant in the interest of determining the truth of the matter at issue. The language of the amendment follows closely that in Fed.R.Evid. 106, as the considerations involved are very similar.

The phrase "in any civil or criminal proceeding" has been moved from its present position, following the word "against," for purposes of clarity. An ambiguity presently exists because the word "against" may be read as referring either to the kind of proceeding in which the evidence is offered or the purpose for which it is offered. The change makes it clear that the latter construction is correct. No change is intended with respect to provisions making evidence rules inapplicable in certain situations. See, e.g., Fed.R.Evid. 104(a) and 1101(d).

Unlike ABA Standards Relating to Pleas of Guilty §3.4 (Approved Draft, 1968), and ALI Model Code of PreArraignment Procedure §350.7 (Proposed Official Draft, 1975), rule 11(e)(6) does not also provide that the described evidence is inadmissible "in favor of" the defendant. This is not intended to suggest, however, that such evidence will inevitably be admissible in the defendant's favor. Specifically, no disapproval is intended of such decisions as United States v. Verdoorn, 528 F.2d 103 (8th Cir. 1976), holding that the trial judge properly refused to permit the defendants to put into evidence at their trial the fact the prosecution had attempted to plea bargain with them, as "meaningful dialogue between the parties would, as a practical matter, be impossible if either party had to assume the risk that plea offers would be admissible in evidence.”

NOTES OF ADVISORY COMMITTEE ON RULES-1982
AMENDMENT

Note to Subdivision (c)(1). Subdivision (c)(1) has been amended by specifying "the effect of any special parole term" as one of the matters about which a defendant who has tendered a plea of guilty or nolo contendere is to be advised by the court. This amendment does not make any change in the law, as the courts are in agreement that such advice is presently required by Rule 11. See, e.g., Moore v. United States, 592 F.2d 753 (4th Cir. 1979); United States v. Eaton, 579 F.2d 1181 (10th Cir. 1978); Richardson v. United States, 577 F.2d 447 (8th Cir. 1978); United States v. Del Prete, 567 F.2d 928 (9th Cir. 1978);

United States v. Watson, 548 F.2d 1058 (D.C.Cir. 1977); United States v. Crusco, 536 F.2d 21 (2d Cir. 1976); United States v. Yazbeck, 524 F.2d 641 (1st Cir. 1975); United States v. Wolak, 510 F.2d 164 (6th Cir. 1975). In United States v. Timmreck, 441 U.S. 780 (1979), 99 S.Ct. 2085, 60 L.Ed.2d 634 (1979), the Supreme Court assumed that the judge's failure in that case to describe the mandatory special parole term constituted "a failure to comply with the formal requirements of the Rule."

The purpose of the amendment is to draw more specific attention to the fact that advice concerning special parole terms is a necessary part of Rule 11 procedure. As noted in Moore v. United States, supra:

Special parole is a significant penalty. *** Unlike ordinary parole, which does not involve supervision beyond the original prison term set by the court and the violation of which cannot lead to confinement beyond that sentence, special parole increases the possible period of confinement. It entails the possibility that a defendant may have to serve his original sentence plus a substantial additional period, without credit for time spent on parole. Explanation of special parole in open court is therefore essential to comply with the Rule's mandate that the defendant be informed of "the maximum possible penalty provided by law."

As the aforecited cases indicate, in the absence of specification of the requirement in the rule it has sometimes happened that such advice has been inadvertently omitted from Rule 11 warnings.

The amendment does not attempt to enumerate all of the characteristics of the special parole term which the judge ought to bring to the defendant's attention. Some flexibility in this respect must be preserved although it is well to note that the unique characteristics of this kind of parole are such that they may not be readily perceived by laymen. Moore v. United States supra, recommends that in an appropriate case the judge

inform the defendant and determine that he understands the following:

(1) that a special parole term will be added to any prison sentence he receives;

(2) the minimum length of the special parole term that must be imposed and the absence of a statutory maximum;

(3) that special parole is entirely different fromand in addition to-ordinary parole; and

(4) that if the special parole is violated, the defendant can be returned to prison for the remainder of his sentence and the full length of his special parole term.

The amendment should not be read as meaning that a failure to comply with this particular requirement will inevitably entitle the defendant to relief. See United States v. Timmreck, supra. Likewise, the amendment makes no change in the existing law to the effect that many aspects of traditional parole need not be communicated to the defendant by the trial judge under the umbrella of Rule 11. For example, a defendant need not be advised of all conceivable consequences such as when he may be considered for parole or that, if he violates his parole, he will again be imprisoned.

Bunker v. Wise, 550 F.2d 1155, 1158 (9th Cir. 1977).

Note to Subdivision (c)(4). The amendment to subdivision (c)(4) is intended to overcome the present conflict between the introductory language of subdivision (c), which contemplates the advice being given "[b]efore accepting a plea of guilty or nolo contendere," and thus presumably after the plea has been tendered, and the "if he pleads" language of subdivision (c)(4) which suggests the plea has not been tendered.

As noted by Judge Doyle in United States v. Sinagub, 468 F.Supp. 353 (W.D.Wis.1979):

Taken literally, this wording of subsection (4) of 11(c) suggests that before eliciting any plea at an arraignment, the court is required to insure that a defendant understands that if he or she pleads guilty or nolo contendere, the defendant will be

waiving the right to trial. Under subsection (3) of 11(c), however, there is no requirement that at this pre-plea stage, the court must insure that the defendant understands that he or she enjoys the right to a trial and, at trial, the right to the assistance of counsel, the right to confront and cross-examine witnesses against him or her, and the right not to be compelled to incriminate himself or herself. It would be incongruous to require that at the preplea stage the court insure that the defendant understands that if he enters a plea of guilty or nolo contendere he will be waiving a right, the existence and nature of which need not be explained until after such a plea has been entered. I conclude that the insertion of the words "that if he pleads guilty or nolo contendere," as they appear in subsection (4) of 11(c), was an accident of draftsmanship which occurred in the course of Congressional rewriting of 11(c) as it has been approved by the Supreme Court. Those words are to be construed consistently with the words "Before accepting a plea of guilty or nolo contendere," as they appear in the opening language of 11(c), and consistently with the omission of the words "that if he pleads" from subsections (1), (2), and (3) of 11(c). That is, as they appear in subsection (4) of 11(c), the words, "that if he pleads guilty or nolo contendere" should be construed to mean "that if his plea of guilty or nolo contendere is accepted by the court." Although this is a very logical interpretation of the present language, the amendment will avoid the necessity to engage in such analysis in order to determine the true meaning of subdivision (c)(4).

Note to Subdivision (c)(5). Subdivision (c)(5), in its present form, may easily be read as contemplating that in every case in which a plea of guilty or nolo contendere is tendered, warnings must be given about the possible use of defendant's statements, obtained under oath, on the record and in the presence of counsel, in a later prosecution for perjury or false statement. The language has prompted some courts to reach the remarkable result that a defendant who pleads guilty or nolo contendere without receiving those warnings must be allowed to overturn his plea on appeal even though he was never questioned under oath, on the record, in the presence of counsel about the offense to which he pleaded. United States v. Artis, No. 78-5012 (4th Cir. March 12, 1979); United States v. Boone, 543 F.2d 1090 (4th Cir. 1976). Compare United States v. Michaelson, 552 F.2d 472 (2d Cir. 1977) (failure to give subdivision (c)(5) warnings not a basis for reversal, "at least when, as here, defendant was not put under oath before questioning about his guilty plea"). The present language of subdivision (c)(5) may also have contributed to the conclusion, not otherwise supported by the rule, that "Rule 11 requires that the defendant be under oath for the entirety of the proceedings" conducted pursuant to that rule and that failure to place the defendant under oath would itself make necessary overturning the plea on appeal. United States v. Aldridge, 553 F.2d 922 (5th Cir. 1977).

When questioning of the kind described in subdivision (c)(5) is not contemplated by the judge who is receiving the plea, no purpose is served by giving the (c)(5) warnings, which in such circumstances can only confuse the defendant and detract from the force of the other warnings required by Rule 11. As correctly noted in United States v. Sinagub, supra,

subsection (5) of section (c) of Rule 11 is qualitatively distinct from the other sections of the Rule. It does not go to whether the plea is knowingly or voluntarily made, nor to whether the plea should be accepted and judgment entered. Rather, it does go to the possible consequences of an event which may or may not occur during the course of the arraignment hearing itself, namely, the administration of an oath to the defendant. Whether this event is to occur is wholly within the control of the presiding judge. If the event is not to occur, it is pointless to inform the defendant of its conse

quences. If a presiding judge intends that an oath not be administered to a defendant during an arraignment hearing, but alters that intention at some point, only then would the need arise to inform the defendant of the possible consequences of the administration of the oath.

The amendment to subdivision (c)(5) is intended to make it clear that this is the case.

The amendment limits the circumstances in which the warnings must be given, but does not change the fact, as noted in Sinagub that these warnings are "qualitatively distinct" from the other advice required by Rule 11(c). This being the case, a failure to give the subdivision (c)(5) warnings even when the defendant was questioned under oath, on the record and in the presence of counsel would in no way affect the validity of the defendant's plea. Rather, this failure bears upon the admissibility of defendant's answers pursuant to subdivision (e)(6) in a later prosecution for perjury or false statement.

NOTES OF ADVISORY COMMITTEE ON RULES-1983

AMENDMENT

Note to Subdivision (a). There are many defenses, objections and requests which a defendant must ordinarily raise by pretrial motion. See, e.g., 18 U.S.C. §3162(a)(2); Fed.R.Crim.P.12(b). Should that motion be denied, interlocutory appeal of the ruling by the defendant is seldom permitted. See United States v. MacDonald, 435 U.S. 850 (1978) (defendant may not appeal denial of his motion to dismiss based upon Sixth Amendment speedy trial grounds); DiBella v. United States, 369 U.S. 121 (1962) (defendant may not appeal denial of pretrial motion to suppress evidence); compare Abney v. United States, 431 U.S. 651 (1977) (interlocutory appeal of denial of motion to dismiss on double jeopardy grounds permissible). Moreover, should the defendant thereafter plead guilty or nolo contendere, this will usually foreclose later appeal with respect to denial of the pretrial motion "When a criminal defendant has solemnly admitted in open court that he is in fact guilty of the offense with which he is charged, he may not thereafter raise independent claims relating to the deprivation of constitutional rights that occurred prior to the entry of the guilty plea." Tollett v. Henderson, 411 U.S. 258, (1973). Though a nolo plea differs from a guilty plea in other respects, it is clear that it also constitutes a waiver of all nonjurisdictional defects in a manner equivalent to a guilty plea. Lott v. United States, 367 U.S. 421 (1961).

As a consequence, a defendant who has lost one or more pretrial motions will often go through an entire trial simply to preserve the pretrial issues for later appellate review. This results in a waste of prosecutorial and judicial resources, and causes delay in the trial of other cases, contrary to the objectives underlying the Speedy Trial Act of 1974, 18 U.S.C. §3161 et seq. These unfortunate consequences may be avoided by the conditional plea device expressly authorized by new subdivision (a)(2).

The development of procedures to avoid the necessity for trials which are undertaken for the sole purpose of preserving pretrial objections has been consistently favored by the commentators. See ABA Standards Relating to the Administration of Criminal Justice, standard 21-1.3(c) (2d ed. 1978); Model Code of Pre-Arraignment Procedure §SS 290.1(4)(b) (1975); Uniform Rules of Criminal Procedure, rule 444(d) (Approved Draft, 1974); 1 C. Wright, Federal Practice and Procedure - Criminal §175 (1969); 3 W. LaFave, Search and Seizure §11.1 (1978). The Supreme Court has characterized the New York practice, whereby appeals from suppression motions may be appealed notwithstanding a guilty plea, as a "commendable effort to relieve the problem of congested trial calendars in a manner that does not diminish the opportunity for the assertion of rights guaranteed by the Constitution." Lefkowitz v. Newsome, 420 U.S. 283, 293 (1975). That Court has never discussed conditional pleas as such, but has permitted without comment a federal appeal on issues preserved by a conditional plea. Jaben v. United States, 381 U.S. 214 (1965).

In the absence of specific authorization by statute or rule for a conditional plea, the circuits have divided on the permissibility of the practice. Two circuits have actually approved the entry of conditional pleas, United States v. Burke, 517 F.2d 377 (2d Cir. 1975); United States v. Moskow, 588 F.2d 882 (3d Cir. 1978); and two others have praised the conditional plea concept, United States v. Clark, 459 F.2d 977 (8th Cir. 1972); United States v. Dorsey, 449 F.2d 1104 (D.C.Cir. 1971). Three circuits have expressed the view that a conditional plea is logically inconsistent and thus improper, United States v. Brown, 499 F.2d 829 (7th Cir. 1974); United States v. Sepe, 472 F.2d 784, aff'd en banc, 486 F.2d 1044 (5th Cir. 1973); United States v. Cox, 464 F.2d 937 (6th Cir. 1972); three others have determined only that conditional pleas are not now authorized in the federal system, United States v. Benson, 579 F.2d 508 (9th Cir. 1978); United States v. Nooner, 565 F.2d 633 (10th Cir. 1977); United States v. Matthews, 472 F.2d 1173 (4th Cir. 1973); while one circuit has reserved judgment on the issue, United States v. Warwar, 478 F.2d 1183 (1st Cir. 1973). (At the state level, a few jurisdictions by statute allow appeal from denial of a motion to suppress notwithstanding a subsequent guilty plea, Cal. Penal Code §1538.5(m); N.Y.Crim. Proc. Law §710.20(1); Wis. Stat. Ann. §971.31(10), but in the absence of such a provision the state courts are also in disagreement as to whether a conditional plea is permissible; see cases collected in Comment, 26 U.C.L.A. L.Rev. 360, 373 (1978).)

The conditional plea procedure provided for in subdivision (a)(2) will, as previously noted, serve to conserve prosecutorial and judicial resources and advance speedy trial objectives. It will also produce much needed uniformity in the federal system on this matter; see United States v. Clark, supra, noting the split of authority and urging resolution by statute or rule. Also, the availability of a conditional plea under specified circumstances will aid in clarifying the fact that traditional, unqualified pleas do constitute a waiver of nonjurisdictional defects. See United States v. Nooner, supra (defendant sought appellate review of denial of pretrial suppression motion, despite his prior unqualified guilty plea, claiming the Second Circuit conditional plea practice led him to believe a guilty plea did not bar appeal of pretrial issues).

The obvious advantages of the conditional plea procedure authorized by subdivision (a)(2) are not outweighed by any significant or compelling disadvantages. As noted in Comment, supra, at 375: "Four major arguments have been raised by courts disapproving of conditioned pleas. The objections are that the procedure encourages a flood of appellate litigation, militates against achieving finality in the criminal process, reduces effectiveness of appellate review due to the lack of a full trial record, and forces decision on constitutional questions that could otherwise be avoided by invoking the harmless error doctrine." But, as concluded therein, those "arguments do not withstand close analysis." Ibid.

As for the first of those arguments, experience in states which have permitted appeals of suppression motions notwithstanding a subsequent plea of guilty is most relevant, as conditional pleas are likely to be most common when the objective is to appeal that kind of pretrial ruling. That experience has shown that the number of appeals has not increased substantially. See Comment, 9 Hous.L.Rev. 305, 315-19 (1971). The minimal added burden at the appellate level is certainly a small price to pay for avoiding otherwise unnecessary trials. As for the objection that conditional pleas conflict with the government's interest in achieving finality, it is likewise without force. While it is true that the conditional plea does not have the complete finality of the traditional plea of guilty or nolo contendere because "the essence of the agreement is that the legal guilt of the defendant exists only if the prosecution's case" survives on appeal, the plea

continues to serve a partial state interest in finality, however, by establishing admission of the defendant's factual guilt. The defendant stands guilty

and the proceedings come to an end if the reserved issue is ultimately decided in the government's favor.

Comment, 26 U.C.L.A. L.Rev. 360, 378 (1978).

The claim that the lack of a full trial record precludes effective appellate review may on occasion be relevant. Cf. United States v. MacDonald, supra (holding interlocutory appeal not available for denial of defendant's pretrial motion to dismiss, on speedy trial grounds, and noting that "most speedy trial claims * * * are best considered only after the relevant facts have been developed at trial"). However, most of the objections which would likely be raised by pretrial motion and preserved for appellate review by a conditional plea are subject to appellate resolution without a trial record. Certainly this is true as to the very common motion to suppress evidence, as is indicated by the fact that appellate courts presently decide such issues upon interlocutory appeal by the government.

With respect to the objection that conditional pleas. circumvent application of the harmless error doctrine, it must be acknowledged that "[a]bsent a full trial record, containing all the government's evidence against the defendant, invocation of the harmless error rule is arguably impossible." Comment, supra, at 380. But, the harmless error standard with respect to constitutional objections is sufficiently high, see Chapman v. California, 386 U.S. 18 (1967), that relatively few appellate decisions result in affirmance upon that basis. Thus it will only rarely be true that the conditional plea device will cause an appellate court to consider constitutional questions which could otherwise have been avoided by invocation of the doctrine of harmless

error.

To the extent that these or related objections would otherwise have some substance, they are overcome by the provision in Rule 11(a)(2) that the defendant may enter a conditional plea only "with the approval of the court and the consent of the government." (In this respect, the rule adopts the practice now found in the Second Circuit.) The requirement of approval by the court is most appropriate, as it ensures, for example, that the defendant is not allowed to take an appeal on a matter which can only be fully developed by proceeding to trial; cf. United States v. MacDonald, supra. As for consent by the government, it will ensure that conditional pleas will be allowed only when the decision of the court of appeals will dispose of the case either by allowing the plea to stand or by such action as compelling dismissal of the indictment or suppressing essential evidence. Absent such circumstances, the conditional plea might only serve to postpone the trial and require the government to try the case after substantial delay, during which time witnesses may be lost, memories dimmed, and the offense grown so stale as to lose jury appeal. The government is in a unique position to determine whether the matter at issue would be case-dispositive, and, as a party to the litigation, should have an absolute right to refuse to consent to potentially prejudicial delay. Although it was suggested in United States v. Moskow, supra, that the government should have no right to prevent the entry of a conditional plea because a defendant has no comparable right to block government appeal of a pretrial ruling pursuant to 18 U.S.C. §3731, that analogy is unconvincing. That statute requires the government to certify that the appeal is not taken for purposes of delay. Moreover, where the pretrial ruling is case-dispositive, § 3731 is the only mechanism by which the government can obtain appellate review, but a defendant may always obtain review by pleading not guilty.

Unlike the state statutes cited earlier, Rule 11(a)(2) is not limited to instances in which the pretrial ruling the defendant wishes to appeal was in response to defendant's motion to suppress evidence. Though it may be true that the conditional plea device will be most commonly employed as to such rulings, the objectives of the rule are well served by extending it to other pretrial rulings as well. See, e.g., ABA Standards, supra (declaring the New York provision "should be enlarged

to include other pretrial defenses"); Uniform Rules of Criminal Procedure, rule 444(d) (Approved Draft, 1974) ("any pretrial motion which, if granted, would be dispositive of the case").

The requirement that the conditional plea be made by the defendant "reserving in writing the right to appeal from the adverse determination of any specified pretrial motion," though extending beyond the Second Circuit practice, will ensure careful attention to any conditional plea. It will document that a particular plea was in fact conditional, and will identify precisely what pretrial issues have been preserved for appellate review. By requiring this added step, it will be possible to avoid entry of a conditional plea without the considered acquiescence of the government (see United States v. Burke, supra, holding that failure of the government to object to entry of a conditional plea constituted consent) and post-plea claims by the defendant that his plea should be deemed conditional merely because it occurred after denial of his pretrial motions (see United States v. Nooner, supra).

It must be emphasized that the only avenue of review of the specified pretrial ruling permitted under a rule 11(a)(2) conditional plea is an appeal, which must be brought in compliance with Fed.R.App.P. 4(b). Relief via 28 U.S.C. §2255 is not available for this purpose.

The Supreme Court has held that certain kinds of constitutional objections may be raised after a plea of guilty. Menna v. New York, 423 U.S. 61 (1975) (double jeopardy violation); Blackledge v. Perry, 417 U.S. 21 (1974) (due process violation by charge enhancement following defendant's exercise of right to trial de novo). Subdivision 11(a)(2) has no application to such situations, and should not be interpreted as either broadening or narrowing the Menna-Blackledge doctrine or as establishing procedures for its application.

Note to Subdivision (h). Subdivision (h) makes clear that the harmless error rule of Rule 52(a) is applicable to Rule 11. The provision does not, however, attempt to define the meaning of “harmless error," which is left to the case law. Prior to the amendments which took effect on Dec. 1, 1975, Rule 11 was very brief; it consisted of but four sentences. The 1975 amendments increased significantly the procedures which must be undertaken when a defendant tenders a plea of guilty or nolo contendere, but this change was warranted by the "two principal objectives" then identified in the Advisory Committee Note: (1) ensuring that the defendant has made an informed plea; and (2) ensuring that plea agreements are brought out into the open in court. An inevitable consequence of the 1975 amendments was some increase in the risk that a trial judge, in a particular case, might inadvertently deviate to some degree from the procedure which a very literal reading of Rule 11 would appear to require.

This being so, it became more apparent than ever that Rule 11 should not be given such a crabbed interpretation that ceremony was exalted over substance. As stated in United States v. Scarf, 551 F.2d 1124 (8th Cir. 1977), concerning amended Rule 11: "It is a salutary rule, and district courts are required to act in substantial compliance with it although *** ritualistic compliance is not required." As similarly pointed out in United States v. Saft, 558 F.2d 1073 (2d Cir. 1977),

the Rule does note say that compliance can be achieved only by reading the specified items in haec verba. Congress meant to strip district judges of freedom to decide what they must explain to a defendant who wishes to plead guilty, not to tell them precisely how to perform this important task in the great variety of cases that would come before them. While a judge who contents himself with literal application of the Rule will hardly be reversed, it cannot be supposed that Congress preferred this to a more meaningful explanation, provided that all the specified elements were covered.

Two important points logically flow from these sound observations. One concerns the matter of construing Rule 11: it is not to be read as requiring a litany or other ritual which can be carried out only by word-for

word adherence to a set "script." The other, specifically addressed in new subdivision (h), is that even when it may be concluded Rule 11 has not been complied with in all respects, it does not inevitably follow that the defendant's plea of guilty or nolo contendere is invalid and subject to being overturned by any remedial device then available to the defendant.

Notwithstanding the declaration in Rule 52(a) that "[a]ny error, defect, irregularity or variance which does not affect substantial rights shall be disregarded," there has existed for some years considerable disagreement concerning the applicability of the harmless error doctrine to Rule 11 violations. In large part, this is attributable to uncertainty as to the continued vitality and the reach of McCarthy v. United States, 394 U.S. 459 (1969). In McCarthy, involving a direct appeal from a plea of guilty because of noncompliance with Rule 11, the Court concluded

that prejudice inheres in a failure to comply with Rule 11, for noncompliance deprives the defendant of the Rule's procedural safeguards, which are designed to facilitate a more accurate determination of the voluntariness of his plea. Our holding [is] that a defendant whose plea has been accepted in violation of Rule 11 should be afforded the opportunity to plead anew * * *.

McCarthy has been most frequently relied upon in cases where, as in that case, the defendant sought relief because of a Rule 11 violation by the avenue of direct appeal. It has been held that in such circumstances a defendant's conviction must be reversed whenever the "district court accepts his guilty plea without fully adhering to the procedure provided for in Rule 11," United States v. Boone, 543 F.2d 1090 (4th Cir. 1976), and that in this context any reliance by the government on the Rule 52(a) harmless error concept "must be rejected." United States v. Journet, 544 F.2d 633 (2d Cir. 1976). On the other hand, decisions are to be found taking a harmless error approach on direct appeal where it appeared the nature and extent of the deviation from Rule 11 was such that it could not have had any impact on the defendant's decision to plead or the fairness in now holding him to his plea. United States v. Peters, No. 77-1700 (4th Cir., Dec. 22, 1978) (where judge failed to comply fully with Rule 11(c)(1), in that defendant not correctly advised of maximum years of special parole term but was told it is at least 3 years, and defendant thereafter sentenced to 15 years plus 3-year special parole term, government's motion for summary affirmance granted, as "the error was harmless"); United States v. Coronado, 554 F.2d 166 (5th Cir. 1977) (court first holds that charge of conspiracy requires some explanation of what conspiracy means to comply with Rule 11(c)(1), but then finds no reversible error "because the rule 11 proceeding on its face discloses, despite the trial court's failure sufficiently to make the required explicitation of the charges, that Coronado understood them").

But this conflict has not been limited to cases involving nothing more than a direct appeal following defendant's plea. For example, another type of case is that in which the defendant has based a post-sentence motion to withdraw his plea on a Rule 11 violation. Rule 32(d) says that such a motion may be granted "to correct manifest injustice," and some courts have relied upon this latter provision in holding that post-sentence plea withdrawal need not be permitted merely because Rule 11 was not fully complied with and that instead the district court should hold an evidentiary hearing to determine "whether manifest injustice will result if the conviction based on the guilty plea is permitted to stand." United States v. Scarf, 551 F.2d 1124 (8th Cir. 1977). Others, however, have held that McCarthy applies and prevails over the language of Rule 32(d), so that "a failure to scrupulously comply with Rule 11 will invalidate a plea without a showing of manifest injustice." United States v. Cantor, 469 F.2d 435 (3d Cir. 1972).

Disagreement has also existed in the context of collateral attack upon pleas pursuant to 28 U.S.C. §2255. On the one hand, it has been concluded that "[n]ot every violation of Rule 11 requires that the plea be set

aside" in a §2255 proceeding, and that "a guilty plea will be set aside on collateral attack only where to not do so would result in a miscarriage of justice, or where there exists exceptional circumstances justifying such relief." Evers v. United States, 579 F.2d 71 (10th Cir. 1978). The contrary view was that McCarthy governed in § 2255 proceedings because "the Supreme Court hinted at no exceptions to its policy of strict enforcement of Rule 11." Timmreck v. United States, 577 F.2d 377 (6th Cir. 1978). But a unanimous Supreme Court resolved this conflict in United States v. Timmreck, 441 U.S. 780 (1979), where the Court concluded that the reasoning of Hill v. United States, 368 U.S. 424 (1962) (ruling a collateral attack could not be predicated on a violation of Rule 32(a))

is equally applicable to a formal violation of Rule 11.* * *

Indeed, if anything, this case may be a stronger one 'for foreclosing collateral relief than the Hill case. For the concern with finality served by the limitation on collateral attack has special force with respect to convictions based on guilty pleas.

"Every inroad on the concept of finality undermines confidence in the integrity of our procedures; and, by increasing the volume of judicial work, inevitably delays and impairs the orderly administration of justice. The impact is greatest when new grounds for setting aside guilty pleas are approved because the vast majority of criminal convictions result from such pleas. Moreover, the concern that unfair procedures may have resulted in the conviction of an innocent defendant is only rarely raised by a petition to set aside a guilty plea."

This interest in finality is strongest in the collateral attack context the Court was dealing with in Timmreck, which explains why the Court there adopted the Hill requirement that in a §2255 proceeding the rule violation must amount to "a fundamental defect which inherently results in a complete miscarriage of justice" or "an omission inconsistent with the rudimentary demands of fair procedure." The interest in finality of guilty pleas described in Timmreck is of somewhat lesser weight when a direct appeal is involved (so that the Hill standard is obviously inappropriate in that setting), but yet is sufficiently compelling to make unsound the proposition that reversal is required even where it is apparent that the Rule 11 violation was of the harmless error variety.

Though the McCarthy per se rule may have been justified at the time and in the circumstances which obtained when the plea in that case was taken, this is no longer the case. For one thing, it is important to recall that McCarthy dealt only with the much simpler pre1975 version of Rule 11, which required only a brief procedure during which the chances of a minor, insignificant and inadvertent deviation were relatively slight. This means that the chances of a truly harmless error (which was not involved in McCarthy in any event, as the judge made no inquiry into the defendant's understanding of the nature of the charge, and the government had presented only the extreme argument that a court "could properly assume that petitioner was entering that plea with a complete understanding of the charge against him" merely from the fact he had stated he desired to plead guilty) are much greater under present Rule 11 than under the version before the Court in McCarthy. It also means that the more elaborate and lengthy procedures of present Rule 11, again as compared with the version applied in McCarthy, make it more apparent than ever that a guilty plea is not "a mere gesture, a temporary and meaningless formality reversible at the defendant's whim," but rather "'a grave and solemn act,' which is 'accepted only with care and discernment.'" United States v. Barker, 514 F.2d 208 (D.C.Cir.1975), quoting from Brady v. United States, 397 U.S. 742 (1970). A plea of that character should not be overturned, even on direct appeal, when there has been a minor and technical violation of Rule 11 which amounts to harmless error.

Secondly, while McCarthy involved a situation in which the defendant's plea of guilty was before the

court of appeals on direct appeal, the Supreme Court appears to have been primarily concerned with §2255type cases, for the Court referred exclusively to cases of that kind in the course of concluding that a per se rule was justified as to Rule 11 violations because of "the difficulty of achieving [rule 11's] purposes through a post-conviction voluntariness hearing." But that reasoning has now been substantially undercut by United States v. Timmreck, supra, for the Court there concluded § 2255 relief "is not available when all that is shown is a failure to comply with the formal requirements of the Rule," at least absent "other aggravating circumstances," which presumably could often only be developed in the course of a later evidentiary hearing.

Although all of the aforementioned considerations support the policy expressed in new subdivision (h), the Advisory Committee does wish to emphasize two important cautionary notes. The first is that subdivision (h) should not be read as supporting extreme or speculative harmless error claims or as, in effect, nullifying important Rule 11 safeguards. There would not be harmless error under subdivision (h) where, for example, as in McCarthy, there had been absolutely no inquiry by the judge into defendant's understanding of the nature of the charge and the harmless error claim of the government rests upon nothing more than the assertion that it may be "assumed" defendant possessed such understanding merely because he expressed a desire to plead guilty. Likewise, it would not be harmless error if the trial judge totally abdicated to the prosecutor the responsibility for giving to the defendant the various Rule 11 warnings, as this "results in the creation of an atmosphere of subtle coercion that clearly contravenes the policy behind Rule 11." United States v. Crook, 526 F.2d 708 (5th Cir. 1976).

Indeed, it is fair to say that the kinds of Rule 11 violations which might be found to constitute harmless error upon direct appeal are fairly limited, as in such instances the matter "must be resolved solely on the basis of the Rule 11 transcript" and the other portions (e.g., sentencing hearing) of the limited record made in such cases. United States v. Coronado, supra. Illustrative are: where the judge's compliance with subdivision (c)(1) was not absolutely complete, in that some essential element of the crime was not mentioned, but the defendant's responses clearly indicate his awareness of that element, see United States v. Coronado, supra; where the judge's compliance with subdivision (c)(2) was erroneous in part in that the judge understated the maximum penalty somewhat, but the penalty actually imposed did not exceed that indicated in the warnings, see United States v. Peters, supra; and where the judge completely failed to comply with subdivision (c)(5), which of course has no bearing on the validity of the plea itself, cf. United States v. Sinagub, supra.

The second cautionary note is that subdivision (h) should not be read as an invitation to trial judges to take a more casual approach to Rule 11 proceedings. It is still true, as the Supreme Court pointed out in McCarthy, that thoughtful and careful compliance with Rule 11 best serves the cause of fair and efficient administration of criminal justice, as it

will help reduce the great waste of judicial resources required to process the frivolous attacks on guilty plea convictions that are encouraged, and are more difficult to dispose of, when the original record is inadequate. It is, therefore, not too much to require that, before sentencing defendants to years of imprisonment, district judges take the few minutes necessary to inform them of their rights and to determine whether they understand the action they are taking.

Subdivision (h) makes no change in the responsibilities of the judge at Rule 11 proceedings, but instead merely rejects the extreme sanction of automatic reversal.

It must also be emphasized that a harmless error provision has been added to Rule 11 because some courts have read McCarthy as meaning that the general harmless error provision in Rule 52(a) cannot be utilized with respect to Rule 11 proceedings. Thus, the addition