We have before us John Minarik's second petition for relief under 28 U.S.C. § 2254. It was tendered to the District Court after the enactment of the Anti-terrorism and Effective Death Penalty Act of 1996 ("AEDPA") and transferred to this Court pursuant to the provisions of that act. Minarik's first § 2254 petition was filed prior to AEDPA's passage. We must decide whether the gatekeeping provisions made applicable to "second or successive petitions" by 28 U.S.C. § 2244 as amended by AEDPA should be applied in Minarik's case. We conclude that such application would have no impermissible retroactive effect and, accordingly, that AEDPA's modified version of § 2244 requires us to deny him permission to proceed with his successive petition.

On February 7, 1971, Minarik killed his former fiancee with an ax. In October 1971, Minarik pleaded guilty to the murder. The Allegheny County Court of Common Pleas convicted Minarik of first degree murder and sentenced him to life imprisonment. Minarik did not pursue a direct appeal. In 1977, Minarik filed a motion to withdraw his guilty plea. The Court of Common Pleas granted the motion. That decision, however, was ultimately overturned by the Pennsylvania Supreme Court, and Minarik was not permitted to withdraw his plea. Commonwealth v. Minarik, 493 Pa. 573, 427 A.2d 623 (Pa.1981).

In 1981, Minarik filed his first federal habeas corpus petition alleging two grounds for relief. First, Minarik claimed that he had not knowingly, intelligently and voluntarily entered his guilty plea because (1) the trial court failed to explain the requisite mental state required for first degree murder, and (2) he had no memory of the events surrounding the murder. Second, Minarik claimed that the Pennsylvania Supreme Court's reversal of the Court of Common Pleas' decision allowing Minarik to withdraw his guilty plea violated his due process and equal protection rights. An extensive evidentiary hearing was held to examine the circumstances surrounding Minarik's guilty plea. Two significant sources of testimony highlighted the hearing. First, expert witnesses testified about the possible effects of mixing alcohol and Triavil, an anti-depressant prescription drug that Minarik had been taking at the time of the murder. Second, Minarik's trial counsel, Ralph J. Cappy, testified that he had thoroughly discussed all of the elements of, and defenses to, the first degree murder charge with Minarik before he entered his plea. According to Cappy, Minarik insisted upon pleading guilty against his advice. The District Court denied Minarik's petition and this Court affirmed.

On April 24, 1996, President Clinton signed into law the Antiterrorism and Effective Death Penalty Act of 1996 ("AEDPA"). 110 Stat. 1214. AEDPA substantially revised the law governing federal habeas corpus codified in chapters 153 and 154 of Title 28. See 28 U.S.C. §§ 2241-66. AEDPA contains "gatekeeping" provisions that establish new procedural and substantive standards governing "second or successive" habeas petitions. Id. § 2244. Procedurally, the AEDPA amendments require petitioners to file a motion in the appropriate Court of Appeals requesting an order authorizing the District Court to consider their "second or successive" application. Id. § 2244(b)(3)(A). A three judge panel of the Court of Appeals may grant such a motion only where the petitioner establishes a prima facie case that the application satisfies AEDPA's new substantive standards regarding "second or successive" petitions. Id. § 2244(b)(3)(C). Notably, the new substantive standards governing the allowance of second or successive applications are more rigorous than the pre-AEDPA standard developed by the courts interpreting the prior version of § 2244. See James S. Liebman & Randy Hertz, Federal Habeas Corpus Practice and Procedure § 28.3a, at 271 (Supp.1997)("[AEDPA] sharply narrow[ed] the (already extremely narrow) circumstances in which new-claim successive petitions are permitted").

On October 6, 1997, following another unsuccessful bid for post conviction relief in state court, Minarik filed a second federal habeas corpus petition, the subject of this appeal, stating three grounds for relief. First, Minarik claims that his trial counsel's failure to discover the availability of an involuntary intoxication defense deprived him of effective assistance of counsel. Second, Minarik repeats his contention that he did not knowingly, intelligently and voluntarily enter his guilty plea. Third, Minarik claims that the state court violated his Fourteenth and Sixth Amendment rights when it refused to hold an evidentiary hearing regarding his claim that trial counsel disregarded his request to file a direct appeal.

Because Minarik had filed a previous habeas petition in 1981, the District Court transferred Minarik's second petition to this Court to permit us to perform our new gatekeeping function under 28 U.S.C. § 2244(b)(3)(A). Discerning a possible retroactivity problem, we requested that the parties brief the issue of whether 28 U.S.C. § 2244's "second or successive" petition provisions, as amended by AEDPA, apply in a case where the first petition was filed prior to AEDPA's enactment.

Two Supreme Court decisions guide our retroactivity analysis in this case. First, we must consider Landgraf v. USI Film Products, 511 U.S. 244, 114 S. Ct. 1483, 128 L. Ed. 2d 229 (1994), the landmark case which establishes the analytical framework governing retroactivity issues. Second, we must consult the Court's more recent decision in Lindh v. Murphy, 521 U.S. 320, 117 S. Ct. 2059, 138 L. Ed. 2d 481 (1997), where it provided additional guidance regarding Landgraf retroactivity analysis in a case involving AEDPA.

In Landgraf, the Court considered whether provisions of the Civil Rights Act of 1991 that provided expanded rights to recover compensatory and punitive damages in Title VII suits, and the right to a jury trial in cases involving claims for such damages, could be applied to cases pending when the Act took effect. Landgraf, 511 U.S. at 247, 114 S. Ct. 1483. The Court found in its case law a strong historical presumption against the retroactive application of statutes:

[T]he presumption against retroactive legislation is deeply rooted in our jurisprudence, and embodies a legal doctrine centuries older than our Republic. Elementary considerations of fairness dictate that individuals should have an opportunity to know what the law is and to conform their conduct accordingly; settled expectations should not be lightly disrupted. For that reason, the principle that the legal effect of conduct should ordinarily be assessed under the law that existed when the conduct took place has timeless and universal appeal. In a free dynamic society, creativity in both commercial and artistic endeavors is fostered by a rule of law that gives people confidence about the legal consequences of their actions.

Id. at 265-66, 114 S. Ct. 1483 (citations omitted). The Court noted that several constitutional provisions manifest similar anti-retroactivity principles.1 Recognizing the "limited scope" of the constitutional restrictions, however, the Court indicated that, absent a violation of such a constitutional provision, the traditional anti-retroactivity presumption permits retroactive application only where "Congress first makes its intention clear [so as to leave no question] that Congress itself has determined that the benefits of retroactivity outweigh the potential for disruption or unfairness." Id. at 268, 114 S. Ct. 1483. The Court then announced a two part test for statutory retroactivity problems:

When a case implicates a federal statute enacted after the events in suit, the court's first task is to determine whether Congress has expressly prescribed the statute's proper reach. If Congress has done so, of course, there is no need to resort to judicial default rules. When, however, the statute contains no such express command, the court must determine whether the new statute would have retroactive effect, i.e., whether it would impair rights a party possessed when he acted, increase a party's liability for past conduct or impose new duties with respect to transactions already completed. If the statute would operate retroactively, our traditional presumption teaches that it does not govern absent clear congressional intent favoring such a result.

Id. at 280, 114 S. Ct. 1483.

Finally, the Landgraf Court identified three categories in which "application of new statutes passed after the events in suit is unquestionably proper" even "absent specific legislative authorization." Id. at 273, 114 S. Ct. 1483. First, "when the intervening statute authorizes or affects the propriety of prospective relief, the application of the new provision is not retroactive." Id. at 273, 114 S. Ct. 1483. Second, courts may apply statutes "conferring or ousting jurisdiction, whether or not jurisdiction lay when the underlying conduct occurred or when the suit was filed." Landgraf, 511 U.S. at 274, 114 S. Ct. 1483. Third, "[c]hanges in procedural rules may often be applied in suits arising before their enactment without raising concerns about retroactivity." Landgraf, 511 U.S. at 275, 114 S. Ct. 1483.

Applying these principles, the Court concluded that the 1991 Act's damages provisions could not be applied to cases pending at the time of enactment because Congress failed to provide an "explicit command" regarding retroactivity and applying the new damages provisions would have a "genuine retroactive effect" by attaching new legal consequences to events completed before the Act's enactment. Id. at 280-84, 114 S. Ct. 1483.2

In Lindh, the Court considered whether AEDPA's new standards for granting habeas petitions in non-capital cases under 28 U.S.C. § 2254(d) could be applied to cases pending at the time AEDPA was enacted. 521 U.S. at 320, 117 S.Ct. at 2059. The District Court had denied Lindh's habeas corpus application in 1995. Shortly after oral argument in Lindh's appeal before the Seventh Circuit Court of Appeals, AEDPA became effective substantially modifying federal habeas corpus law. After en banc reconsideration, the Court of Appeals applied Landgraf and concluded that the AEDPA modified version of § 2254(d) could be applied to Lindh's case because it did not "attach new legal consequences" to events completed before enactment, and therefore did not result in a genuine retroactive effect. The Supreme Court reversed, holding that Congress did not intend to apply AEDPA's habeas corpus amendments to non-capital cases pending when AEDPA became effective.

The Lindh majority began by stating that Landgraf stood for the proposition that "where a statute [does] not clearly mandate an application with retroactive effect, a court [must] determine whether applying it as its terms ostensibly indicated would have [sic] genuinely retroactive effect; if so, the judicial presumption against retroactivity would bar its application." Id. at ----, 117 S.Ct. at 2062. The Court rejected, however, the respondent's contention that "whenever a new statute on its face could apply to the litigation of events that occurred before it was enacted, there are only two alternative sources of rules to determine its ultimate temporal reach: either an 'express command' from Congress or application of our Landgraf default rule." Id. Instead, the Court stated that:

[i]n determining whether a statute's terms would produce a retroactive effect ... and in determining a statute's temporal reach generally, our normal rules of construction apply. Although Landgraf's default rule would deny application when a retroactive effect would otherwise result, other construction rules may apply to remove even the possibility of retroactivity (as by rendering the statutory provision wholly inapplicable to a particular case).

Id. at ----, 117 S.Ct. at 2063.

The Court further stressed that Landgraf 's retroactivity analysis must be case-specific:

In sum, if the application of a [statutory] term would be retroactive as to [the particular party affected], the term will not be applied, even if in the absence of retroactive effect, we might find the term applicable.

Id.

Applying "normal rules of construction" to determine congressional intent regarding AEDPA's temporal reach, the Court observed that all of AEDPA's habeas corpus amendments are found in Title I of the Act, and that Title I's amendments can be divided into two categories: (i) amendments to chapter 153 of Title 28 governing all federal habeas corpus proceedings found in §§ 101-106 of the Act, and (ii) amendments establishing a new chapter 154 of Title 28 governing habeas proceedings against qualifying states in capital cases found in § 107 of the Act. See id. (citing 110 Stat. 1217-26). Notably, § 107(c) provides that "[c]hapter 154 ... shall apply to cases pending on or after the date of enactment of this Act." 110 Stat. 1226. According to the Court, "the negative implication of § 107(c), is that the new provisions of chapter 153 generally apply only to cases filed after the Act became effective." Lindh, 521 U.S. at ----, 117 S. Ct. at 2068. Thus, the AEDPA modified chapter 153 standards for granting applications for habeas relief could not be applied to Lindh's case because it was "pending" on appeal when AEDPA became effective.

We read Landgraf and Lindh as establishing the following principles that we must employ in resolving the issues before us:

1. There is a strong presumption against applying a statute in a manner that would attach "new legal consequences" to events completed before the statute's enactment, i.e., a manner that would "impair rights a party possessed when he acted, increase a party's liability for past conduct, or impose new duties." Landgraf, 511 U.S. at 280, 114 S. Ct. 1483.

2. If Congress has focused on the issue, "has determined that the benefits of retroactivity outweigh the potential for disruption or unfairness," and has provided unambiguous evidence of its conclusion by directing that retroactive effect be given, then, and only then, will the presumption be overridden.

3. Consistent with these principles, normal rules of statutory construction "may apply to remove ... the possibility of retroactivity." Nothing short of an unambiguous directive, however, will justify giving a statute a retroactive effect. Thus, when normal rules of statutory construction indicate that a statute is intended to be applied in a manner involving no retroactive effect, a Court need inquire no further. On the other hand, if such construction suggests that a retroactive effect may have been intended, the traditional presumption nevertheless bars retroactive application unless an unambiguous congressional directive is found.3

With these principles in mind we address Minarik's claim that applying AEDPA's "second or successive application" procedures and standards to his case would result in an impermissible retroactive application of the statute.

As a threshold matter, we note that the Lindh decision's construction of AEDPA's temporal reach does not control the outcome of this case. Lindh held that AEDPA's text, read in light of normal principles of statutory interpretation, evidences a congressional intent that AEDPA's chapter 153 amendments should generally be applied to petitions, like Minarik's, filed after April 24, 1996, the effective date of the Act, but not to petitions, like Lindh's, filed before. This does not resolve the issue before us, however. The finding of congressional intent in Lindh was based on the drawing of a negative inference from Congress's express mandate that AEDPA's new rules regarding certain death penalty cases apply to pending cases. Because Congress had expressly provided for application to pending capital cases, but not to pending non-capital cases, it was a fair inference that Congress did not intend retrospective application to the latter. Landgraf and Lindh make clear, however, that while such an inference is sufficient to eliminate the possibility of a retroactivity problem, it is not the kind of unambiguous statement that will justify overriding the judicial presumption against retroactivity in a case where a retroactivity problem exists. See Mathews, 161 F.3d at 166-68 (rejecting a similar "negative inference" from the statute's text as evidence of "clear intent" to justify a statute's retroactive effect).

Landgraf describes the statement of congressional intent necessary to override the presumption against retroactive application in terms of "express commands," "unambiguous directives," and "clear statements." 511 U.S. at 263, 264, 272-73, 286, 114 S. Ct. 1483. It specifically teaches that "a statement that a statute will become effective on a certain date does not even arguably suggest that it has application to conduct that occurred at an earlier date." Id. at 257, 114 S. Ct. 1483. It necessarily follows, we believe, that the negative inference drawn in Lindh--that the Act's chapter 153 amendments are applicable to cases filed after its enactment--does not constitute an unambiguous directive that those amendments be applied to all post-enactment filed petitions, including those in which the first petition was filed before the Act's passage.

Moreover, Lindh counsels that the only cases in which the Court has "found truly 'retroactive' effect adequately authorized by statute have involved statutory language so clear that it could sustain only one interpretation." 521 U.S. at ---- n. 4, 117 S. Ct. at 2064 n. 4. It was precisely because the negative inference drawn under normal principles of statutory construction did not satisfy this requirement that the Lindh Court discussed at some length the distinction between express congressional commands and manifestations of congressional intent gleaned by applying those principles. In short, it is apparent that the Court in Lindh meticulously examined AEDPA and failed to find any express command from Congress regarding retroactive application of the AEDPA's chapter 153 amendments. See Liebman & Hertz, Federal Habeas Corpus § 2.7b, at 29 ("Lindh also makes clear that [AEDPA's habeas corpus amendments] nowhere state Congress' intention to cause retroactive effects ... ").

Based on our reading of Landgraf and Lindh, we join two other courts of appeals in holding that AEDPA contains no unambiguous guidance regarding retroactive application of AEDPA's new "second or successive" petition standards and procedures to cases in which the first habeas petition was filed before AEDPA's enactment. See In re Green, 144 F.3d 384 (6th Cir.1998); In re Hanserd, 123 F.3d 922, 924 (6th Cir.1997); United States v. Ortiz, 136 F.3d 161, 165 (D.C.Cir.1998).

Having concluded that Congress did not clearly express its intent regarding the retroactivity issue presented in this case, we now turn to a case-specific analysis of whether applying AEDPA's § 2244(b) would have a genuine retroactive effect by "attach[ing] new legal consequences to events completed before [AEDPA's] enactment." Landgraf, 511 U.S. at 270, 114 S. Ct. 1483. If applying AEDPA's habeas corpus amendments would produce a genuine retroactive effect in Minarik's case, then Landgraf 's default rule prohibits their application. If on the other hand, no such retroactive effect would result, then Lindh requires us to apply the AEDPA amendments because Minarik's second petition wasfiled after April 24, 1996. To resolve this issue we treat 2244(b)'s new procedural provisions and substantive standards separately.4

AEDPA established a new procedure governing "second or successive" petitions for federal habeas corpus relief under § 2244(b). It provides:

Before a second or successive application permitted by this section is filed in the district court, the applicant shall move in the appropriate court of appeals for an order authorizing the district court to consider the application.

28 U.S.C. § 2244(b)(3)(A). Minarik maintains that subjecting him to this new procedure is an impermissible retroactive application of the statute. Section 2244(b)(3)(A), however, is a change in procedural law which falls within the firmly established "procedural change" category described in Landgraf that may be retrospectively applied. See Landgraf, 511 U.S. at 275, 114 S. Ct. 1483 ("Because rules of procedure regulate secondary rather than primary conduct, the fact that a new procedural rule was instituted after the conduct giving rise to the suit does not make application of the rule at trial retroactive.").

Because AEDPA, as read by the Lindh Court in the light of normal principles of statutory interpretation, calls for the application of § 2244(b)(3)(A) to cases filed after April 24, 1996, and because it is a rule of procedure that does not "attach new legal consequences to events completed before its enactment," petitioners in Minarik's position must seek permission of a Court of Appeals prior to proceeding on a second petition, even if their first petition was filed before the Act was adopted. See In re Hanserd, 123 F.3d 922, 934 (6th Cir.1997) ("Inmates who wish to file a second or successive petition should first file a motion in [the Court of Appeals] requesting permission under 28 U.S.C. §§ 2244, 2255, regardless of when the first motion to vacate sentence was filed.").

We now consider whether applying AEDPA's new substantive gatekeeping standards would have an impermissible retroactive effect if applied in Minarik's case. At all times here relevant, the unsuccessful prosecution of a § 2254 proceeding has had an adverse impact on the petitioner's right to prosecute a second or successive § 2254 proceeding. When Minarik filed his first federal habeas petition, the existing law provided that he could thereafter prosecute another such petition only if he could (1) show cause for, and prejudice from, the omission of his new claim or claims from his earlier petition (i.e., that his proceeding would not constitute an "abuse of the writ"), or (2) demonstrate "actual innocence." See 28 U.S.C. § 2244; McCleskey v. Zant, 499 U.S. 467, 111 S. Ct. 1454, 113 L. Ed. 2d 517 (1991); Kuhlmann v. Wilson, 477 U.S. 436, 106 S. Ct. 2616, 91 L. Ed. 2d 364 (1986).

AEDPA's passage significantly altered the showing that Minarik was required to make in order to proceed on new claims in a second petition. Section 2244(b), as amended by AEDPA, provides in relevant part:

(1) A claim presented in a second or successive habeas corpus application under section 2254 that was presented in a prior application shall be dismissed.

(2) A claim presented in a second or successive habeas corpus application under section 2254 that was not presented in a prior application shall be dismissed unless--

(A) the applicant shows that the claim relies on a new rule of constitutional law, made retroactive to cases on collateral review by the Supreme Court, that was previously unavailable; or

(B)(i) the factual predicate for the claim could not have been discovered previously through the exercise of due diligence; and

(ii) the facts underlying the claim, if proven and viewed in light of the evidence as a whole, would be sufficient to establish by clear and convincing evidence that, but for constitutional error, no reasonable factfinder would have found the applicant guilty of the underlying offense.

28 U.S.C. § 2244(b)(2).

These substantive gatekeeping provisions were intended to reduce the universe of cases in which a habeas petition may go forward on a second or successive petition. In those cases where a prisoner in state custody had a right to prosecute a second or successive petition prior to AEDPA's passage, but would be deprived of that right by these new gatekeeping provisions, we conclude that applying the AEDPA standard would have a "genuine retroactive effect" because it would attach a new and adverse consequence to pre-AEDPA conduct--the prosecution of the original proceeding.

By its terms, § 2244(b) requires that a claim not meeting its articulated standards "shall be dismissed" thus extinguishing any right the petitioner may have to relief. Its effect is not unlike that of AEDPA's statute of limitations which we recently declined to apply retroactively in Burns v. Morton, 134 F.3d 109, 111 (3d Cir.1998). On September 21, 1995, when Burns exhausted all of his state rights of direct appeal and collateral review, there was no statute of limitations applicable to federal habeas petitions under 28 U.S.C. § 2254. AEDPA, however, established a one year statute of limitations which, by its terms, begins to run as soon as the petitioner's rights of direct review have been exhausted or expired. See 28 U.S.C. § 2244(d). If applicable, AEDPA's limitations period would require Burns to file his petition on or before September 22, 1996. Burns, however, filed his petition on April 22, 1997. Because Burns had a right to proceed on his habeas claim prior to AEDPA's enactment, and because AEDPA's statute of limitations, if applied to his case, would extinguish his claim, we held that such an application would impermissibly attach new legal consequences to events completed before the statute's enactment. Burns, 134 F.3d at 111. We see no meaningful distinction between that case and Minarik's, assuming that Minarik had a right to proceed on his claim prior to AEDPA.

AEDPA's gatekeeping provisions when applied to cases in which there was a right to proceed under preexisting law also operate much like the RICO amendment we considered in Mathews, 161 F.3d at 163. Prior to the RICO amendment, securities fraud could serve as a predicate offense under RICO, thereby entitling a plaintiff to treble damages. The RICO amendment altered the text of the statute conferring federal jurisdiction over RICO claims to exclude jurisdiction over RICO claims predicated on "any conduct that would have been actionable as fraud in the purchase or sale of securities." Id. We pointed out that, while phrased in jurisdictional terms, the amendment's practical effect was to alter substantive rights because "prior to the passage of the Act, the [plaintiff] had a RICO cause of action based upon defendants' alleged actions, but afterward he would not." Id. Similarly, if a habeas petitioner had a right to initiate federal proceedings to secure release from confinement prior to AEDPA, and had no such rights thereafter, then AEDPA has altered substantive rights and thereby attached new legal consequences to pre-enactment conduct.5

We find additional support for our conclusion in cases from two other circuits. See Ortiz, 136 F.3d 161; In re Green, 144 F.3d 384; In re Sonshine, 132 F.3d 1133 (6th Cir.1997); In re Hanserd, 123 F.3d 922. These Courts have encountered a similar retroactivity issue in cases where petitioners challenged application of AEDPA's new "gatekeeping" standards governing "second or successive" § 2255 motions. In these cases, the petitioners filed their first § 2255 motions before, and their second motions after, AEDPA's effective date. These Courts have concluded that the AEDPA standard would have an impermissible retroactive effect if applied where the petitioner would have been allowed to file his second § 2255 motion under the pre-AEDPA standard, but would be precluded from doing so under the AEDPA standard. See Ortiz, 136 F.3d at 166 ("[T]he new standards and procedures under AEDPA for filing § 2255 motions could only be improperly retroactive as applied to [the petitioner] if he would have met the former cause-and-prejudice standard under McCleskey and previously would have been allowed to file a second motion, but could not file a second motion under AEDPA."); In re Hanserd, 123 F.3d at 930 ("When [the movant] filed his initial § 2255 motion, the law would have allowed him to raise a Bailey claim in a second motion ... Under AEDPA, however, he may not. Applying the new statute would thus attach a severe new legal consequence to his filing a first motion ... Because Congress has not expressed an intent that the new Act have such a retroactive effect, we could not apply AEDPA in this way."); Sonshine, 132 F.3d at 1134 (noting that retroactive effect exists only where the "difference matters" between pre- and post-AEDPA standards). Conversely, where the two standards lead to the same outcome, these courts have concluded that there is no genuine retroactive effect and the AEDPA standard may be applied. See Sonshine, 132 F.3d at 1135 ("[Petitioner's claim] would be barred under both AEDPA and the old abuse-of-writ standard. [Petitioner] would not have prevailed under pre-AEDPA law, as his petition would have been denied as an abuse of writ. AEDPA's restrictions thus do not attach new legal consequences for [petitioner], and AEDPA has no impermissible retroactive effect on this case."); Ortiz, 136 F.3d at 167 (applying AEDPA's new standard to deny the petitioner's second § 2255 motion carried no impermissible retroactive effect because petitioner "failed to meet the requirements of the former 'abuse of writ' standard of McCleskey, and the new AEDPA standards."); see also In re Hanserd, 123 F.3d at 932 ("Where the old and the new law lead to an identical result, there is no need to conduct a retroactivity analysis because the new law has not attached any new consequences to preenactment conduct or upset settled expectations."); United States v. Enigwe, 1998 WL 150974 (E.D.Pa. March 30, 1998) (applying AEDPA standards where first § 2255 motion was filed before AEDPA was not impermissibly retroactive because there was "no difference in outcome between pre- and post-AEDPA law"). We agree.

We conclude therefore that if Minarik can show that he would have been entitled to pursue his second petition under pre-AEDPA law, then the Landgraf default rule prohibits applying AEDPA's new substantive gatekeeping provisions to bar his claims. In the absence of such a showing, however, applying those standards to Minarik results in no genuine retroactive effect, and the AEDPA standard must be applied under the Supreme Court's holding in Lindh that AEDPA's habeas corpus amendments apply generally to cases filed after its effective date.

1. The Pre-AEDPA Law And Minarik's Second Petition

We first consider whether Minarik's second application is barred under the pre-AEDPA standard. For the reasons that follow, we conclude that it is.

Minarik filed his first federal habeas petition on December 23, 1981, ten years after pleading guilty to his fiancee's murder. His first petition raised two claims: (1) the due process clause was violated because hi § guilty plea was not knowingly, intelligently and voluntarily entered because (a) his counsel did not explain to him the mental state which the State would be required to prove and, (b) he had no recollection of the events surrounding the crime, and (2) the equal protection and due process clauses were violated when the Supreme Court of Pennsylvania failed to apply Commonwealth v. Minor, 467 Pa. 230, 356 A.2d 346 (1976), in his case.

In November of 1982, the District Court held a lengthy evidentiary hearing concerning the circumstances surrounding Minarik's guilty plea. Minarik called two expert witnesses to testify about the effects of alcohol and Triavil, a drug prescribed for Minarik several months before the murder. While neither knew when Minarik last took Triavil before the crime or how much alcohol he had consumed, each expressed the opinion, inter alia, that he was amnesiac following his arrest. Minarik testified that he had no memory of the events of the crime, that his attorney had failed to advise him regarding the intent that would have to be proved and that counsel, rather than Minarik, made the decision to plead guilty. The respondent called Minarik's trial counsel, Ralph J. Cappy, in rebuttal. He confirmed that Minarik's memory on the night of the murder lapsed when he was in the bushes outside the house. Counsel also testified that he reviewed every element of the offense charged and possible defenses with Minarik, and explained that Minarik had insisted on pleading guilty against counsel's advice. Counsel summed up the situation as follows:

Q. Did you make a recommendation yourself as to whether or not there should be a guilty plea or you should go to trial?

A. I know that was one point of disagreement that he and I always had.

I was always--my recollection is that I wanted to try this case. There were a number of reasons why I wanted to try it. I didn't believe in my heart that a jury would send a 21-year-old man, with his education, lack of prior record, to the electric chair.

We had somewhat of an equitable defense, in the sense that this was a homicide resulting from a lover's situation, with a very distraught defendant; we had alcohol involved; we had a potential for drugs and alcohol involved; we had a potential, as I recall the Behavior Clinic describing him as having had a personality disorder. None of these things by themselves rose to the level of the legitimate defense, but it was my opinion, taken together with his age, his family, it was a very good family, hard working good people, that our equitable defense, in my mind, would have--I mean, there's always a chance involved, but in my mind I didn't believe I would lose the case to the death penalty, therefore the worse I could do was life, therefore, why not take a chance and see if you could get a second or voluntary. Voluntary was probably the most realistic from analyzing it from a legal standpoint.

Quite frankly, John refused all through this. If my memory serves me correctly, he never wanted to go to trial. He steadfastly maintained that he was pleading guilty, period; he was going to take his punishment.

App. III at 526-527a.

Counsel further testified that he had talked with Minarik's family doctor who had prescribed Triavil for the depression Minarik experienced after the breakup with his fiancee. He engaged the services of Dr. Stanger, a psychiatrist, to secure an opinion regarding a possible insanity defense, and he reviewed reports supplied by the state from the Behavioral Clinic on Minarik's mental state. In connection with the possibility of an intoxication defense, counsel and the Chief Investigator of the Public Defender's Office interviewed everyone they could find who was present at the party Minarik attended earlier on the evening of the crime. They discovered two witnesses who were prepared to testify that Minarik "had announced his intended purpose prior to actually going to commit this act." Id. at 535a.

When asked whether he had investigated and considered "an intoxication defense" and an "involuntary intoxication defense," counsel gave the following testimony:

A. My investigation, as I recall it, did reveal the fact that at some time during the evening, earlier in the evening, John had become intoxicated and had placed himself on a couch and dozed off, or rested for a period of time. Subsequent to that time he either awoke or got up and seemed to be, to the witnesses who had talked with him, coherent, not intoxicated, and to one of those witnesses he had offered a ring, an engagement ring, that apparently had been the victim's ring, as I recall, and had told the witness, "I'll not need this ring after tonight," or "I'll no longer need this ring," and told this witness that he was going to, in essence, kill Rosemary.

Q. Did you investigate the possibility of defending against these charges on an involuntary intoxication theory?

A. John had told me that he was taking a medication--I had a discussion or discussions with the family doctor--

* * * * * *

The family physician had prescribed for him a particular drug. I know that's an issue in this case, but I'm still not familiar with exactly all the ramifications. I thought maybe we would have a chance to produce a defense that would reduce the degree by combining the alcohol with the drug and I endeavored to do that. I know I inquired of not only Dr. Stanger, but I even went down to Dr. Campbell, but I'm not quite sure it was Dr. Campbell, about what would be the effect of the drug with alcohol and I believe at the time it was grain alcohol, I'm not sure about that, but anyway it was some kind of a college party, could there be a reason, in combining these two drugs, I mean the drug and the alcohol, could that have set John off, or could we connect that with a mental deficiency which would result in a defense? Maybe not a complete blackout defense, but some defense which would reduce the degree from first to second, or even go into voluntary manslaughter. I remember specifically Dr. Stanger saying, "No, I believe ..."--the problem I'm having here is, I consulted with a number of people on this, not only other criminal defense attorneys, but physicians, some of them psychiatrists, these were friends of mine who were in a residency program, an internship program at the University of Pittsburgh, and I could get from no one any indication that an accommodation of this drug and alcohol, given the facts as described by the witnesses at the party, before he left on his way to Rosemary, no physician or physician in training could give me any indication that the alcohol and drug played a part in what he had done.

Id. at 535a-536a; 537a-538a.

The District Court credited trial counsel's testimony and concluded that Minarik's plea was knowingly, intelligently, and voluntarily entered. While it found that Minarik lacked memory of the events surrounding the crime at the time of his plea, the Court concluded that this did not prevent the plea from being knowing, intelligent and voluntary. Finally, it rejected the contention that the failure of the Pennsylvania court to follow the Minor case constituted a federal constitutional violation. On appeal, this Court affirmed in all respects.

Following an unsuccessful bid for post-conviction relief in state court, Minarik filed his second federal habeas petition on October 3, 1997. He asserted three claims: (1) trial counsel's failure to discover that he had available a complete defense of involuntary intoxication induced by Triavil and alcohol constituted ineffective assistance of counsel, (2) due process was violated because his guilty plea was not knowing, intelligent and voluntary, and (3) the state court's failure to give him an evidentiary hearing on his claim that trial counsel failed to file a direct appeal as he instructed violated due process. To support his second claim, Minarik renews his first petition's allegations of amnesia and lack of knowledge of the elements of the offense and adds several new allegations. In the course of making these three claims, Minarik asserts that he is actually innocent because he was involuntarily intoxicated at the time of the crime.

Minarik's second claim is, in substance, the same claim advanced and rejected in his first habeas proceeding.6 Accordingly, he can go forward on that claim only if he shows "actual innocence." Kuhlmann, 477 U.S. 436, 106 S. Ct. 2616, 91 L. Ed. 2d 364. Before addressing that issue, we turn to whether Minarik has shown "cause and prejudice" for failing to include his first and third claims in his first petition.

"The cause standard requires the petitioner to show that 'some objective factor external to the defense impeded counsel's efforts to raise the claim in state court." McCleskey v. Zant, 499 U.S. 467, 111 S. Ct. 1454, 113 L. Ed. 2d 517 (1991) (quoting Murray v. Carrier, 477 U.S. 478, 488, 106 S. Ct. 2639, 91 L. Ed. 2d 397 (1986)). While a showing that the factual basis for a claim was unavailable at the time of the first petition may constitute "cause," the fact that the petitioner was subjectively unaware of that factual basis is insufficient if the relevant facts were discoverable with due diligence. McCleskey, 499 U.S. at 497, 111 S. Ct. 1454.

We begin with Minarik's first claim: ineffective assistance of counsel for failing to discover the involuntary intoxication defense. While Minarik insists that he did not know that he had a defense of involuntary intoxication until the middle of the evidentiary hearing on his first petition, his habeas counsel had obviously discovered the potential effects of Triavil and alcohol some weeks earlier and no reason is suggested why a similar investigation at any point during the ten years preceding the first petition's filing would not have produced the same information. It follows that the factual basis for Minarik's first claim was discoverable in December of 1981, when the first petition wasfiled. Indeed, we do not understand Minarik to contend otherwise.

The "cause" that Minarik does rely upon before us is the fact that he had not exhausted his state remedies with respect to his first claim when he filed his first petition. This "cause" is legally insufficient, however, because it is not an "objective factor external to the defense." As we have explained, Minarik is deemed to have knowledge of all facts discoverable with reasonable diligence. With that knowledge, he had the alternative on December 23, 1981, of delaying the filing of his first federal habeas petition until he had exhausted what turned out to be the first claim of his second petition. We hold that his failure to do that and then litigate all his claims together constituted an abuse of the writ.

In Rose v. Lundy, 455 U.S. 509, 102 S. Ct. 1198, 71 L. Ed. 2d 379 (1982), the Supreme Court held that a habeas petition containing exhausted and unexhausted claims cannot be entertained by a federal court unless the petitioner withdraws, and thus relinquishes, the unexhausted claims. This "total exhaustion" rule, the Court concluded, would "encourage state prisoners to seek full relief first from the state courts, thus giving those courts the first opportunity to review all claims of error.... Equally as important, [under that rule] federal claims that have been fully exhausted in state courts will more often be accompanied by a complete factual record to aid the federal courts in their review." 455 U.S. at 519, 102 S. Ct. 1198. Finally, the Court emphasized 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. To the extent that the exhaustion requirement reduces piecemeal litigation, both the courts and the prisoners should benefit, for as a result the district court will be more likely to review all of the prisoner's claims in a single proceeding, thus providing for a more focused and thorough review." 455 U.S. at 520, 102 S. Ct. 1198.

The court went on to spell out what this meant for petitioning prisoners:

[O]ur interpretation of §§ 2254(b), (c) provides a simple and clear instruction to potential litigants: before you bring any claims to federal court, be sure that the first have taken each one to state court....Those prisoners who misunderstand this requirement and submit mixed petitions nevertheless are entitled to resubmit a petition with only exhausted claims or to exhaust the remainder of their claims.

* * * * * *

The prisoner's principal interest, of course, is in obtaining speedy federal relief on his claims.... A total exhaustion rule will not impair that interest since he can always amend the petition to delete the unexhausted claims, rather than returning to state court to exhaust all of his claims. By invoking this procedure, however, the prisoner would risk forfeiting consideration of his unexhausted claims in federal court....

455 U.S. at 520, 102 S. Ct. 1198.

While the Supreme Court said that petitioners "would risk" forfeiture of unexhausted claims rather than that they "would forfeit" such claims, this choice of wording was designed to leave room for cases in which there might be a legitimate justification for proceeding on the exhausted issues alone and thus no abuse of the writ. The Court's opinion makes it clear, however, that the prisoner's desire to get an adjudication of some of his claims earlier rather than later is not alone a legitimate justification for going forward on less than all of one's claims. Thus, if "a prisoner deliberately withholds one of two grounds for federal collateral relief at the time of filing his first application, in the hope of being granted two hearings or for some other such reason, he may be deemed to have waived his right to a hearing on a second application presenting the withheld grounds." Id. at 521, 102 S. Ct. 1198 (quoting from Sanders v. United States, 373 U.S. 1, 18, 83 S. Ct. 1068, 10 L. Ed. 2d 148 (1963)).

Rose did not announce a new rule of law. It neither questioned nor reversed existing precedent. The Court's holding reflected its "interpretation of a federal statute on the basis of its language and legislative history, and consistent with its underlying policies." Rose, 455 U.S. at 519-20, 102 S. Ct. 1198. Rose was argued on October 14, 1981, two months before Minarik filed his first federal habeas petition on December 23, 1981, and was decided on March 3, 1982, a little over two months after that filing date. The statute, its legislative history, and its underlying policies were the same on those dates as they were when Minarik's first federal petition was filed. It necessarily follows that the teachings of Rose were the law of the land at the time Minarik elected to file a federal habeas petition limited to the two claims contained in that petition.

Even if we were unpersuaded that Rose represented the law of the land when Minarik filed his petition, however, our ultimate conclusion would not differ. Rose undisputably became the law of the land on March 3, 1982, and the Federal Magistrate Judge's opinion dated April 23, 1982, expressly called the Rose holding to Minarik and his counsel's attention. That opinion preceded the District Court's evidentiary hearing concerning the circumstances surrounding Minarik's guilty plea by more than six months. The subject matter of that requested hearing included what he knew and had been told about the elements of the offense charged and the possible defenses thereto. At some point well before that hearing, Minarik necessarily must have known that he was going to call the two expert witnesses he ultimately called to testify about the effects of Triavil and alcohol. He likewise was aware that he would take the stand and testify that his trial counsel did not explain to him the state of mind required as an element of the offense. Finally, Minarik obviously knew his trial counsel was likely to be called to testify concerning his preparation and pre-plea discussions with Minarik. For that reason, Minarik took trial counsel's deposition in August of 1982.

In view of the actual knowledge Minarik possessed well before the evidentiary hearing on his first petition, we conclude that Minarik had a duty to advise the Court of his new claims before the hearing. These claims would obviously have required a second, extensive evidentiary hearing on essentially the same subject matter. Stated conversely, we hold that it was an abuse of the writ for Minarik to go forward without at least advising the Court of this withheld claim. If he had done so, the District Court would have been compelled by Rose to dismiss the petition without prejudice and require Minarik to file a new petition after exhausting his state remedies on the new claim. Indeed, this almost came to pass during the November 1982 hearing when Minarik's habeas counsel began to question Minarik's trial counsel about his investigation of a possible defense based on intoxication. The Court raised the issue sua sponte:

THE COURT: Let me interrupt you for one second. I'm becoming somewhat concerned. It appears to me that the issue here is the voluntary nature of the guilty plea. I suspect we are going further afield and raising other issues, and if we are raising other issues, then the entire case is subject to immediate dismissal.

So, I want to caution you that I think it's very inappropriate to be raising other issues. I think that we have been getting into them and I think we have been getting into them for quite a while now.

MR. POTTER: The issues are as stated in the habeas corpus petition and I submit that the cross examination is within the scope of the direct examination.

THE COURT: Well, I'm not saying one way or the other, but I'm just saying that I think we are at this juncture, clearly injecting issues, which if they are raised in the habeas corpus petition, that the entire petition is subject to dismissal because they have not been raised in the State Court.

* * * * * *

THE COURT: ... I have not in my own mind resolve d the issue, but clearly the only issue appropriate before this court is as to the nature of the guilty plea and any other matters, which might be waived in a collateral proceeding, I strongly suggest have to be raised in the State court first and they have not been raised.

MR. POTTER: I understand the Court's position.

Q. All right, Mr. Cappy, I will limit myself. We a re interested in amnesia. The defendant had it, didn't he? ...

App. III at 567a-568a; 568a-569a.

Minarik apparently chose not to call his new claim to the District Court's attention because he wanted a ruling on his first petition's claims sooner rather than later. While understandable, this motivation does not justify the kind of piecemeal litigation that Minarik seeks to pursue.

A similar analysis leads us to conclude that asserting Minarik's third claim at this late date also constitutes an abuse of the writ. On December 23, 1981, Minarik knew the content of the instructions he gave trial counsel as well as the fact that counsel filed no direct appeal. With this knowledge, he was not licensed to withhold this ineffective assistance of counsel claim for later litigation.

"Because [Minarik] has been unable to show 'cause and prejudice' sufficient to excuse his failure to present his evidence in support of his first federal petition, ... [he] may obtain review of his constitutional claims only if he falls within the 'narrow class of cases ... implicating a fundamental miscarriage of justice.' " Schlup v. Delo, 513 U.S. 298, 314-15, 115 S. Ct. 851, 130 L. Ed. 2d 808 (1995) (quoting McCleskey v. Zant, 499 U.S. 467, 111 S. Ct. 1454, 113 L. Ed. 2d 517 (1991)). This "requires the habeas petitioner to show that 'a constitutional violation has probably resulted in the conviction of one who is actually innocent'.... To establish the requisite probability, the petitioner must show that it is more likely than not that no reasonable juror would have convicted him in light of the new evidence. The petitioner thus is required to make a stronger showing than that needed to establish prejudice." Id. at 327, 115 S. Ct. 851 (quoting from Murray v. Carrier, 477 U.S. 478, 106 S. Ct. 2639, 91 L. Ed. 2d 397 (1986)).

Minarik argues that he is actually innocent because he was involuntarily intoxicated when he committed the crime. Our task is therefore to look to all of the information currently available concerning a potential "defense of involuntary intoxication," including that developed in the proceedings since Minarik's sentencing, to determine whether, more likely than not, no reasonable juror exposed to that information would have convicted him.

It is not clear to us that Pennsylvania law would characterize intoxication produced by the voluntary consumption of a prescription drug and alcohol as "involuntary" even if that consumption was without knowledge of a synergistic effect. See Commonwealth v. Todaro, 301 Pa.Super. 1, 446 A.2d 1305 (Pa.Super.1982) (holding such intoxication to be voluntary). Nothing in our analysis turns on this point, however. We assume for present purposes that evidence of involuntary intoxication of a degree that would have deprived Minarik of the ability to form the requisite intent, or control his actions, would constitute a complete defense or reduce what would otherwise be first degree murder to some lesser included offense.7 We further assume that "actual innocence" of the crime charged would include the situation where the new evidence would show the petitioner not guilty of first degree murder, though guilty of some lesser offense. Even under these assumptions, however, Minarik cannot show "actual innocence" because he has failed to establish the necessary factual basis.

In addition to the conclusory statement that he "had available a complete defense of involuntary intoxication," Minarik's petition asserts only that he is "innocent of crimes charged because he suffered an involuntary intoxication due to misprescribed Triavil which Public Defender Cappy missed in 1971." App. IV at 855a-856a. Minarik's objections to the Magistrate Judge's report in the District Court adds only that "[i]n 1971, Petitioner suffered an involuntary intoxication due to 'atropenic intoxication' because of 'inappropriately prescribed' Triavil by Dr. Provan." Obj. Mag. Rep. at 9. Minarik's main brief before us adds only that he discovered his complete defense "only while sitting in court during the federal evidentiary hearing in 1982 and listening to the testimony of Drs. Sadoff and Himmelhoch." Pet. Br. at 6 n. 1. The most specific information Minarik has provided is that contained in the following segment of his reply brief:

Triavil contains Elavil, a tricyclic antidepressant, and Trilafon, a member of the phenothiazine family, an antipsychotic agent. After his visit to Dr. Provan, while taking Triavil, John Minarik's behavior became increasingly bizarre.

Mr. Minarik spent Saturday afternoon, February 6, 1971, in the company of college friends who were planning a party. His friends, while at the party that night, noticed his visible personality change, his red face, and his dry mouth. Redness of face and dryness of mouth are two observable undesired side effects when a person is suffering from atropinic intoxication: the "class is atropine syndrome" associated with the misprescription of Triavil. Dr. Himmelhoch testified on November 18, 1982: "In medical school this syndrome is summarized as mad as a hatter, red as a beet, and dry as a bone." Neither Dr. Provan nor the prescription label identified the drug, and no warnings were given of side effects or danger associated with alcohol.

Pet. Rep. Br. at. 3-4.

The record before us contains no evidence concerning when before the crime Minarik last consumed alcohol or Triavil, nor how much he consumed of each. Nor is there any evidence that Minarik was observed to have been behaving in an aberrant manner after waking from his nap, or after the crime. Absent as well is any expert testimony supporting the theory that Minarik was out of control at the time of the crime. We thus have only general claims of involuntary intoxication against the background of sworn testimony from an experienced trial attorney that he and his investigator conducted an extensive investigation and found no evidentiary support for an intoxication defense, voluntary or involuntary.

It follows that Minarik has not come close to putting the integrity of the judgment against him into question. We cannot excuse his abuse of the writ on the ground that he has demonstrated actual innocence. Accordingly, we conclude that Minarik would have been precluded from filing his second habeas petition under pre-AEDPA law. It necessarily follows that applying the AEDPA gatekeeping standard cannot have a genuine retroactive effect upon Minarik and therefore Lindh requires their application. We now turn to determine whether Minarik may proceed on his second petition's claims under the new AEDPA standard.8

Minarik's second claim, that his guilty plea was not knowingly, intelligently and voluntarily entered into, is repetitive and must be dismissed under § 2244(b)(1) because it was "presented in a prior application." Minarik's first and third claims, however, are "new claims" that we must evaluate under § 2244(b)(2).

Section 2244(b)(2) provides two narrow grounds upon which a new claim may be pursued in a second petition, neither of which are present here. First, a petitioner may make a prima facie showing that his new claim falls within the ambit of a "new rule of constitutional law, made retroactive to cases on collateral review by the Supreme Court." Id. § 2244(b)(2)(A). Neither of Minarik's "new claims" rely upon such a new rule of constitutional law.

Second, § 2244(b)(2)(B) permits petitioners to proceed on new claims in certain cases where there is newly discovered evidence. It has two requirements. Initially, petitioners must demonstrate that the "factual predicate for the claim could not have been discovered previously through the exercise of reasonable diligence." Id. § 2244(b)(2)(B)(i). Additionally, petitioners must establish that "the facts underlying the claim, if proven and viewed in light of the evidence as a whole, would be sufficient to establish by clear and convincing evidence that, but for constitutional error, no reasonable fact finder would have found the applicant guilty of the underlying offense." Id. § 2244(b)(2)(B)(ii).

Minarik's first "new claim"--that his trial counsel was ineffective for failing to discover that Minarik had a complete involuntary intoxication defense--does not satisfy either of § 2244(b)(2)(B)'s conditions. As we concluded in our analysis of the pre-AEDPA law, this claim was "discoverable" when Minarik filed his first habeas application in 1981. In addition, Minarik cannot meet § 2244(b)(2)(b)(ii)'s innocence requirement for the substantially the same reasons that he cannot establish "actual innocence" under the pre-AEDPA standard.

Minarik's second "new claim"--that the state court committed constitutional error by failing to give him an evidentiary hearing on his claim that trial counsel ignored his instruction to file a direct appeal--also fails to meet § 2244(b)(2)(B)'s requirements. Minarik had actual knowledge of both his instructions and his attorney's failure to file such an appeal well before he filed his first habeas petition in 1981. Moreover, his counsel's failure to appeal is not a "fact" relevant to whether Minarik was "guilty of the underlying offense."

Section 2244(b)(2)(B) thus requires that we deny Minarik permission to proceed on his second petition.

We hold that anyone seeking to file a second or successive petition under 28 U.S.C. § 2254 after April 24, 1996, must move in the appropriate Court of Appeals for an order authorizing the District Court to consider the application. When such a motion is filed by a petitioner whose previous petition was filed before that date, the Court of Appeals must apply the substantive gatekeeping standards of 28 U.S.C. § 2244(b) as amended by AEDPA unless such application would bar a second or successive petition that could have been considered by the District Court under the law existing at the time the previous petition was filed.

Since Minarik's second petition is barred by both § 2244(b) as amended by AEDPA and the preexisting law, we will deny him permission to proceed in the District Court.