concurring.
Under our current caselaw, post-conviction collateral attack is available under Article 11.07, V.A.C.C.P., to raise errors of federal constitutional dimension. Ex parte *387Banks, 769 S.W.2d 539 (Tex.Cr.App.1989). Some federal constitutional claims may be procedurally defaulted, however, and this Court will not reach the merits of a forfeited claim on habeas corpus. See Ex parte Crispen, 777 S.W.2d 103 (Tex.Cr.App.1989). But, a forfeitable claim may nevertheless be cognizable on the merits if it meets the test for excusing forfeiture articulated in Ex parte Chambers, 688 S.W.2d 483 (Tex.Cr.App.1984) (Campbell, J., concurring), viz: if the “right” upon which relief is premised was “not recognized” as of the time of the procedural default. I have joined Judge Campbell’s separate opinion in Black v. State, 816 S.W.2d 350 (Tex.Cr.App., delivered this day), holding that claims predicated upon Penry v. Lynaugh, 492 U.S. 302, 109 S.Ct. 2934, 106 L.Ed.2d 256 (1989), meet the Chambers test, and are therefore cognizable even when raised for the first time on direct appeal or on collateral attack. Thus, I agree that under our caselaw Penry error is cognizable under Article 11.07, supra. I write separately to add, however, that I would reach the merits of applicant’s claim here irrespective of whether any forfeiture is excusable under Ex parte Chambers. I also wish to lodge several relatively minor objections to the majority opinion.
I.
For some time now I have appealed to the Court to construct a more principled standard for making threshold determinations of cognizability of federal constitutional claims under Article 11.07, supra. There is no federal constitutional requirement that we recognize any, much less every, federal constitutional claim in state post-conviction habeas corpus. Ex parte Crispen, supra (Clinton, J., concurring). Just as in the context of the law of contemporaneous objection, analysis of what is cognizable on collateral attack should be informed first and foremost by the State’s legitimate interest in the integrity and finality of convictions. The appellate process is past. Any standard for deciding what claims to entertain in a collateral proceeding should be rigorous. In Crispen, I proposed what seemed an acceptable standard, suggesting we limit cognizability of federal constitutional claims:
“to ‘exceptional’ constitutional defects so ‘fundamental’ as not to be susceptible to a determination of harm, such as those identified in Rose v. Clark, 478 U.S. 570, 577-78, 106 S.Ct. 3101, 3106, 92 L.Ed.2d 460, 470 (1986). [footnote omitted.] It is only ‘[e]rrors of this kind [that] justify collateral relief no matter how long a judgment may have been final and even though they may not have been preserved properly in the original trial.’ Rose v. Lundy, 455 U.S. [509,] at 544, 102 S.Ct. [1198,] at 1217, 71 L.Ed.2d [379,] at 404 [ (1982) ]. (Stevens, J., dissenting) (footnotes omitted).”
Subsequently, in Ex parte Dutchover, 779 S.W.2d 76 (Tex.Cr.App.1989) (Clinton, J., concurring), I further proposed a doctrine of excuses, applicable to those federal constitutional claims that do not meet the criteria suggested in Crispen:
“It may be that applicability of a constitutional harm analysis would prove too inflexible a criterion for cognizability to accommodate the collective judgment of the Court regarding which constitutional claims are sufficiently grievous to trump otherwise persuasive finality interests. A given applicant may be able to show to the satisfaction of the Court both a constitutional violation and such attendant harm as to impugn fundamental fairness of the whole proceeding against him. If the defect was not one he could have raised by due diligence in the ordinary course of appellate review, the Court might well choose to hold that the State’s interest in finality of judgments must give way.”
Id., at 79. It occurs to me now that finality considerations further counsel that these non-exceptional federal constitutional defects should only be entertained on collateral attack if the basis for the claim was recognized by the courts at the time of the applicant’s trial. That is to say, a non-exceptional constitutional claim should not be retroactively applied in state post-conviction writ of habeas corpus. Apart from exceptional claims, only already-extant con*388stitutional claims such as ineffective assistance of counsel at trial or on appeal, or failure of the State to disclose exculpatory evidence, where a record to substantiate the claim could not have been made by due diligence before perfection of the record on appeal, should be cognizable on collateral attack. The upshot is that for non-exceptional federal constitutional claims, I would not entertain a Chambers “right not recognized” analysis.
By contrast, however we choose to define an “exceptional” federal constitutional defect, once we have identified one, we Jhould consider it “exceptional” for all purposes. A “right not recognized” analysis would be superfluous. For it is pointless to say a claim is cognizable in post-conviction habeas corpus because it fits within an “exceptional” category, but then decline to treat it on the merits because of some procedural default at the trial or appellate level. A federal constitutional violation may be of sufficient character or magnitude that we are obliged to say that our interest in rectifying it overcomes otherwise compelling finality interests. We may therefore say that a claim premised upon such a violation, though it dc«s not affect jurisdiction of the convicting court, is cognizable on collateral attack, without more. Ex parte Crispen, supra (Clinton, J., concurring). An interest in finality is also what undergirds our rules governing contemporaneous objection and retroactivity. Once we say a federal constitutional claim is of such character as to defeat the State’s interest in finality for purposes of determining cognizability on collateral attack, no justification can exist for denying full re-troactivity to that claim, or for requiring an objection in the trial court or point of error on appeal. Finality is the common touchstone. That an “exceptional” federal constitutional claim is cognizable at all under Article 11.07, should mean by definition that we do not care whether it was, or could have been, raised at any point below.
The Supreme Court has not definitively said whether Penry error is subject to a constitutional harm analysis. An inquiry for harm was neither required of nor conducted by the Fifth Circuit upon remand. See Penry v. Lynaugh, 882 F.2d 141 (CA5 1989). There are some indications that the Supreme Court would consider a harm analysis appropriate.* See Hitchcock v. Dugger, 481 U.S. 393, at 399, 107 S.Ct. 1821, at 1824, 95 L.Ed.2d 347, at 353 (1987); Clemons v. Mississippi, 494 U.S. 738, at —, 110 S.Ct. 1441, at 1450-51, 108 L.Ed.2d 725, at 741-42 (1990); Parker v. Dugger, — U.S.-,-, 111 S.Ct. 731, 738, 739, 112 L.Ed.2d 812 (1991). But irrespective of whether it ultimately proves susceptible to a constitutional harm analysis, I would consider Penry an “exceptional” federal constitutional claim. The State has no overriding interest in upholding the finality of a sentence of death under circumstances in which the capital sentencing procedure does not meet threshold Eighth *389Amendment standards for “reliability.” Woodson v. North Carolina, 428 U.S. 280, 96 S.Ct. 2978, 49 L.Ed.2d 944 (1976). That full consideration be accorded mitigating evidence in a capital sentencing proceeding ultimately:
“rests squarely on the predicate that the penalty of death is qualitatively different from a sentence of imprisonment, however long. Death, in its finality, differs more from life imprisonment than a 100-year prison term differs from one of only a year or two. Because of that qualitative difference, there is a corresponding difference in the need for reliability in the determination that death is the appropriate punishment in a specific case.”
Id., 428 U.S. at 306, 96 S.Ct. at 2991, 49 L.Ed.2d at 961. The State’s “interest in leaving concluded litigation in a state of repose, that is, reducing the controversy to a final judgment not subject to further judicial revision,” Mackey v. United States, 401 U.S. 667, at 683, 91 S.Ct. 1160, at 1175, 28 L.Ed.2d 404, at 415 (1971) (Separate opinion by Harlan, J.), must fall away in the face of a Penny claim. Because death is irrevocable, reliability in the procedure for adjudging it is indispensable. If Penny stands for anything, it is that Article 37.-071, supra, by itself, is a constitutionally unreliable mechanism for determining that death is an appropriate penalty whenever evidence is proffered having mitigating potential that transcends its logical relevance to the special issues. Where the decision to execute is made under circumstances so suspect under the Eighth Amendment, the State’s interest in repose of its convictions cannot justify repose of the convicted.
II.
In three other respects I take issue with the majority opinion. First, the opinion in Penny itself does not purport to decide what would constitute a sufficient Penny objection or request. Nevertheless, in a footnote the majority gratuitously “finds” that applicant’s attorney made a sufficient objection at trial to have preserved Penny error, because his requested instruction is similar to that requested in Penny itself. At 385, n. 4. Yet the instruction applicant requested, had it been given, would have been insufficient to cure Penny error, for it would have instructed the jury to consider any mitigating evidence only as it pertains to the special instructions. But Penny requires that the jury be empowered to assess a sentence less than death on the basis of mitigation that goes beyond the scope of the special issues. This Court long ago observed that as to evidence that mitigates strictly within the parameters of the special issues, the logical relevance would be readily apparent, and no special jury instruction is necessary. Quinones v. State, 592 S.W.2d 933, 947 (Tex.Cr.App.1980), cert. denied, 449 U.S. 893,101 S.Ct. 256, 66 L.Ed.2d 121, reh. denied, 449 U.S. 1027, 101 S.Ct. 600, 66 L.Ed.2d 490. See also Burns v. State, 761 S.W.2d 353, at 358, n. 5 (Tex.Cr.App.1988). That observation remains pertinent. The specific instruction requested by applicant at his trial was properly denied. See James v. State, 805 S.W.2d 415, at 417, n. 3 (Tex.Cr.App.1990) (Opinion on remand from the United States Supreme Court).
In any event, the question of what constitutes a sufficient objection to invoke Pem~y need not be addressed. Under the Court’s holding in Black v. State, supra, no request or objection is necessary for preservation of applicant’s Penny error.
Second, a plurality of judges declares that it “will not entertain arguments addressing evidence which would have been proffered by defense counsel had the trial judge allowed an additional instruction on that evidence.” At 386, n. 6. Because the Court grants relief without resort to that potential evidence, this observation is also gratuitous. As in Hammond v. State, 799 S.W.2d 741, at 749-750, n. 7 (Tex.Cr.App.1990), the question is left for another day.
Finally, it is true that applicant’s mitigating evidence in this cause fits the “double-edged” mold, having a tendency to ameliorate blameworthiness at the same time that it may counsel an affirmative answer to the second special issue. I do not understand the holding in Penny to be limited to this kind of evidence, however. Any evidence that has a tendency to ameliorate blame*390worthiness wholly apart from the special issues will invoke Eighth Amendment concerns, whether the evidence also happens to be relevant to a special issue or not. See Franklin v. Lynaugh, 487 U.S. 164, at 185, 108 S.Ct. 2320, at 2333, 101 L.Ed.2d 155, at 173 (1988) (O’Connor, J., concurring) (“... If, however, petitioner had introduced mitigating evidence about his background or character or the circumstances of the crime that was not relevant to the special verdict questions, or that had relevance to the defendant’s moral culpability beyond the scope of the special verdict questions, the jury instructions would have provided the jury with no vehicle for expressing its ‘reasoned moral response’ to that evidence.”) (emphasis supplied). To the extent it may suggest otherwise, I disagree with the opinion of the majority.
With these observations, I join the judgment of the Court, but not its opinion.
That this Court, in its capacity as direct appellate court in capital cases, can meaningfully determine "beyond a reasonable doubt” that Penry error did or did not contribute to the assessment of the death penalty seems an awfully unlikely prospect to me. How can we predict with any level of confidence how a jury, presented with mitigating evidence going beyond special issues, would answer the value-laden, non-fact-specific question of whether a capital offender deserves to live in spite of his crime? Such a judgment simply cannot be predicted "beyond a reasonable doubt.” It seems to me we could not avoid simply substituting our own normative judgment for that which the jury was precluded from making — because of the Penry error itself.
Nor could we declare ourselves the ultimate arbiter of that normative question, see Clemons v. Mississippi, 494 U.S. 738, 110 S.Ct. 1441, 108 L.Ed.2d 725 (1990), without addressing other problems that would engender. For instance, it would certainly be unfair to force a capital accused to put on evidence Penry itself recognized to be "double-edged," thus himself contributing to the jury’s affirmative answers to special issues, in the hope that this Court, conducting an appellate Penry review, would adjudge him fit to live despite the jury’s answers to the special issues. Moreover, assume we were to be persuaded by mitigating evidence that a capital defendant who has been assessed the death penalty in the trial court nevertheless does not deserve to die. Even if we were to nullify the death sentence as a matter of Eighth Amendment dictates, neither Article 37.071, supra, nor Article 44.251, V.A.C.C.P, authorizes this Court to impose a life sentence, or any other alternative in its place.