concurring specially:
Respectfully, I disagree with the majority’s conclusion that Scott failed in his original federal habeas corpus proceeding to fairly present (as part of his challenge to the heinous, atrocious or cruel (“HAC”) aggravating factor) the aspect involving the deficiency in the guidance for the trial jury. For the reasons set out below, I am satisfied that the issue was fairly presented, and thus must be addressed. Addressing the issue it is clear that our prior resolution thereof, Scott v. Dugger, 891 F.2d 800 (11th Cir.1989), cert. denied, 498 U.S. 881, 111 S.Ct. 224, 112 L.Ed.2d 179 (1990), relied upon an erroneous legal principle. Notwithstanding that, I conclude that under the circumstances of this case, Scott has failed to satisfy the high standards required for relief under Fed. R.App.P. 60(b)(6) or recall of our mandate. Thus, I concur in the result reached by the majority.
I.
First, I explain the deficiency in the guidance for Scott’s trial jury with respect to the HAC aggravating circumstance. The guidance which the jury received in the instant case is identical to the guidance received by the jury in Shell v. Mississippi, 498 U.S. 1, 111 S.Ct. 313, 112 L.Ed.2d 1 (1990), which guidance the Supreme Court held was constitutionally deficient. In this case, the guidance received by the jury with respect to the HAC aggravating factor was as follows:
That the crime for which the Defendant is to be sentenced was especially heinous, atrocious or cruel. Heinous means extremely wicked or shockingly evil. Atrocious means outrageously wicked and vile. Cruel means designed to inflict a high degree of pain, utter indifference to or enjoyment of the suffering of others, pitilessness.
In Shell, the jury received virtually identical guidance with respect to the “especially heinous, atrocious or cruel” factor, namely:
The word heinous means extremely wicked or shockingly evil; atrocious means outrageously wicked and vile; and cruel means designed to inflict a high degree of pain with indifference to, or even enjoyment of, the suffering of others.
Shell v. Mississippi, 498 U.S. at 2, 111 S.Ct. at 313, (Marshall, J., concurring) (quoting Shell v. State, 554 So.2d 887, 905-06 (Miss.1989). The Supreme Court in Shell held that this language was constitutionally deficient: “Although the trial court in this case used a limiting instruction to define the ‘especially heinous, atrocious or cruel’ factor, that instruction is not constitutionally sufficient.” Id. Similarly, in Maynard v. Cartwright, 486 U.S. 356, 108 S.Ct. 1853, 100 L.Ed.2d 372 *1555(1988), the guidance received by the jury was virtually identical:
The term “heinous” means extremely wicked or shockingly evil; “atrocious” means outrageously wicked and vile; “cruel” means pitiless, or designed to inflict a high degree of pain, utter indifference to, or enjoyment of, the suffering of others.
Shell, 498 U.S. at 2, 111 S.Ct. at 314 (quoting the Maynard instruction as set out in Cartwright v. Maynard, 822 F.2d 1477, 1488 (10th Cir.1987) (en banc)). Thus, it is clear that the guidance received by the jury in the instant case was constitutionally deficient.
II.
The next issue that must be addressed is whether Shell and Maynard constitute new law such that Teague v. Lane, 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989), precludes Scott’s reliance thereon. This issue has been decided in Scott’s favor. The Supreme Court in Stringer v. Black, — U.S. —,—, 112 S.Ct. 1130, 1135, 117 L.Ed.2d 367 (1992), and this court in Glock v. Singletary, 36 F.3d 1014, n. 27 (11th Cir.1994), held that Maynard was foreordained by Godfrey v. Georgia, 446 U.S. 420, 100 S.Ct. 1759, 64 L.Ed.2d 398 (1980). Thus, Maynard is not new law, and is not subject to a Teague bar. As noted above, the relevant language in Maynard was virtually identical to that in the instant case.1 Thus, Scott’s claim — that the guidance to his jury on the HAC factor— was constitutionally deficient is not Teague barred.
III.
I turn next to the issue with respect to which I disagree with the majority — i.e., whether Scott fairly presented a claim of deficiency in the guidance of the jury in his original federal habeas corpus proceeding. It is undisputed that Scott fairly presented at least a closely related issue. It is also clear that the State did not interpose any claim of procedural bar in the previous federal habeas corpus proceeding; and thus, if Scott did fairly present the issue then, the State could not now interpose a procedural bar.
The State argues, and the district court in the Rule 60(b) proceedings below concluded, that Scott’s claim in the first federal proceedings was limited to a challenge to the arbitrariness of the application of the HAC factor by the Florida Supreme Court. The majority also so holds. Respectfully, I disagree.
As mentioned, it is undisputed that Scott did in fact raise in the first federal proceedings a closely related claim, i.e., a claim based upon the HAC aggravating circumstance. My close reading of Scott’s initial brief in the prior appeal to this court persuades me that Scott did in fact fairly present the issue now before us — i.e., the constitutional sufficiency of the guidance which his sentencing jury received. In that brief, at p. 71, Scott argued:
In the instant case, as in all Florida capital cases, there was a similar [i.e., similar to that in Maynard ] absence of guidance for the jury at the trial level. The jury was merely told of the aggravating circumstances. Presumably, the limitation of arbitrary application of the HAC circumstance would not seriously come into play until review by the Supreme Court of Florida. But because that review itself has been arbitrary and capricious with respect to the HAC circumstance, the Eighth Amendment has been violated. A review of Florida Supreme Court’s irrational treatment of the circumstance makes the violation obvious.
The foregoing language clearly asserts that the jury itself received insufficient guidance, similar to that in Maynard.2, Dissected into its component parts, Scott’s argument was: (1) that the jury received insufficient guidance, and (2) that the narrowing construction entailed in review by the Supreme Court of Florida did not cure the problem because that review was itself inconsistent and arbitrary. Although the rest of the brief, and also the reply brief, focused on the inconsis*1556tency of the review by the Florida Supreme Court, the quoted language sets the context of the argument. In other words, the context begins with the inadequacy of the guidance for the jury, and then focuses upon the reviewing function of the Supreme Court of Florida and whether or not that review function can cure the problem.
In its answer brief, the State addressed the merits of Scott’s claim and interposed no procedural bar. The State did not take issue with Scott’s assertion that there was an “absence of guidance for the jury at the trial level” similar to that in Maynard. Rather, the State disputed only Scott’s assertion that the review function by the Florida Supreme Court was arbitrary and inconsistent. Understandably responding to the only argument made by the State’s answer brief, Scott’s reply brief also focused only upon the consistency of the review function by the Florida Supreme Court.
Not only does the language quoted above — “similar absence of guidance for the jury at the trial level” — clearly raise the issue, the context of Scott’s argument clearly puts at issue the deficiency in the jury instructions. As noted above, the quoted language frames the context of Scott’s argument: (1) that the jury received insufficient guidance, and (2) that the review by the Florida Supreme Court did not cure the problem because that review was itself inconsistent. The claim of deficient jury guidance was therefore a logically necessary predicate, without which the discussion of the inconsistent appellate review would have been irrelevant. Indeed, the whole point of Scott’s lengthy discussion of the inconsistencies in the Florida Supreme Court review was' to demonstrate that the review function failed to cure the deficient jury guidance.3
Finally, and in addition to the clear language quoted above from Scott’s brief and the fact that the claim of deficient jury guidance was a logically necessary predicate for Scott’s inconsistent review argument, the fact that Scott’s brief relied so heavily upon Maynard v. Cartwright reinforces my conclusion that the deficient jury guidance claim was fairly presented. The context of the analysis in Maynard was identical to the argument presented by Scott in his brief. In Maynard, the Supreme Court’s analysis was two-pronged: (1) deficient jury guidance, and (2) a failure by the Oklahoma Supreme Court to cure the problem. The fact that Maynard was the linchpin of Scott’s argument bolsters my conclusion that Scott’s presentation to this court in the prior appeal must be recognized to have fairly presented a claim that there was an “absence of guidance for the jury at the trial level” and that the review function by the Florida Supreme Court failed to cure the problem because the review was itself inconsistent and arbitrary. For the foregoing reasons,4 I disagree with the failure of the majority to recognize the fact that the issue was fairly presented by Scott.5
*1557IV.
Having determined that Scott has a meritorious claim which was properly preserved in his first federal habeas corpus proceeding and which is not barred by any of the several procedural bars, I turn next to whether this claim warrants relief pursuant to Rule 60(b)(6) and/or recall of our mandate. Scott argues that such relief is justified here because of an error of law. In Ritter v. Smith, 811 F.2d 1398 (11th Cir.), cert. denied, 483 U.S. 1010, 107 S.Ct. 3242, 97 L.Ed.2d 747 (1987), we determined that such relief was warranted when a claim was presented and decided, and then it became apparent that it had been wrongly decided.
A.
I focus first on whether our prior decision, 891 F.2d 800, erroneously decided Scott’s claim. Our prior decision was based on the misapprehension that the determinative factor in resolving a challenge to the vagueness of the HAC aggravating circumstance was the narrowing review function of the sentencing judge and the Florida Supreme Court. We labored under the erroneous assumption that it was not important whether the jury was charged with the narrowing language, because the trial judge (who of course knew the law of State v. Dixon, 283 So.2d 1 (Fla.1973), cert. denied, sub nom. Hunter v. Florida, 416 U.S. 943, 94 S.Ct. 1950, 40 L.Ed.2d 295 (1974), and Proffitt v. Florida, 428 U.S. 242, 96 S.Ct. 2960, 49 L.Ed.2d 913 (1976)) was the ultimate sentencer, and because of the narrowing construction employed by the Florida Supreme Court in its review function.6 However, the Supreme Court in Espinosa v. Florida, — U.S. —, 112 S.Ct. 2926, 120 L.Ed.2d 854 (1992), held otherwise, and we now know we were wrong. In Glock v. Singletary, supra, n. 26, this court held that Espinosa was not new law for the purposes of Teague v. Lane, supra. Thus, we now know not only that we were wrong in deciding Scott, but that it was well established at that time, indeed since Tedder v. State, 322 So.2d 908 (Fla.1975), that a Florida sentencing jury was a eo-sentencer along with the trial judge, and thus must itself receive the narrowing construction of the HAC factor. Accordingly, the ground upon which we decided Scott’s claim in his first federal habeas corpus proceeding was an error of law. Although we applied an erroneous rule of law in our prior resolution of Scott’s claim, I conclude for the reasons set out in Part IV.B. that neither Rule 60(b)(6) relief nor recall of our mandate is appropriate.
B.
A mere error of law is not sufficient in and of itself to obtain relief under Rule '60(b)(6). Ritter v. Smith, 811 F.2d at 1401 (“a change in the law will not always provide the truly extraordinary circumstances necessary to reopen a case.”). Recall of a mandate involves a similarly high standard. Recall of our mandate is appropriate only “to prevent injustice.” Eleventh Circuit Rule 41 — 1(b). I conclude that the circumstances of the instant case do not rise to the level required for recall of our mandate or for Rule 60(b)(6) relief.
It is true that an almost identical instruction to the jury has been held to be deficient in Maynard and Shell. However, it is also true that a portion of the challenged instruction charged Scott’s jury on the meaning of the term cruel:
Cruel means designed to inflict a high degree of pain, utter indifference to or enjoyment the suffering of others, pitilessness.
The concurring opinion in Shell, 498 U.S. at 2-4, 111 S.Ct. at 314, suggests that such a definition of cruel may meet constitutional muster, but that the constitutional vagueness occurred in the alternative grounds of heinous and atrocious. Thus, I cannot conclude *1558that the cruel definition which Scott’s jury heard (aside from the heinous definition and the atrocious definition) constituted a clear error of law. No court has so held. See Maynard, 486 U.S. at 365, 108 S.Ct. at 1859 (“We also do not hold that some kind of torture or serious physical abuse is the only limiting construction.”). In Scott’s case, the particular facts persuade me that there is no reasonable doubt but that the jury would have applied the cruel definition to the exclusion of the vaguer heinous and atrocious definitions. My reason is that the facts of the instant murder constitute overwhelming evidence of “a high degree of pain” and “utter indifference ... to the suffering of others.” In other words the facts clearly focused the jury on the high degree of pain and the utter indifference of Scott to the victim’s suffering. The victim was found in his own home the next morning covered with blood, his hands and feet bound, and brutally beaten to death. In addition to the other blows, the victim sustained six blows to the head with a blunt instrument so severe that it caused a' compressed fracture of the skull. Id. The doctor testified that the victim was still alive when his hands and feet were bound, and that the subsequent blows to the head were fatal. Id. at 869. Thus, the evidence overwhelmingly established torture and indifference to the suffering of the victim. The jury having heard those facts and the instruction which was given, I readily conclude, to a high degree of certainty, that the jury applied the more specific cruel definition, and not the vaguer heinous or atrocious definition. The. cruel definition was specifically oriented to the pain and suffering which the victim in this case suffered to a high degree.7 Although it is clear that the entire heinous, atrocious or cruel instruction was deficient, it is far from clear that the jury relied upon the unconstitutionally vague portion; indeed, I conclude (with a high degree of certainty) quite the contrary. Therefore, Scott has failed to establish a clear constitutional error infecting his death sentence. Ritter v. Smith, 811 F.2d at 1401.
In addition, several other factors contribute to my conclusion that Scott has failed to satisfy the high standard for recall of a mandate or Rule 60(b)(6) relief. As in Ritter v. Smith, 811 F.2d at 1403, considerations of comity point against Scott’s position. Also, it is significant that the overwhelming evidence in this case indicates that the murder was committed in a manner that fell well within the narrowest conceivable definition of the HAC aggravating factor. Grant of the writ in this case would be futile, because the overwhelming evidence indicates clearly that the Florida Supreme Court would exercise its review function to conclude that the deficiency in the instruction in this case was harmless beyond a reasonable doubt.8
*1559All of the foregoing reasons in combination persuade me that Scott has failed to demonstrate the extraordinary circumstances required for Rule 60(b)(6) relief, and has failed to show that it is necessary to recall our mandate “to prevent injustice.”
V.
Accordingly, I concur in the decision to deny Scott’s motion to recall the mandate, and the decision to affirm the district court’s denial of Rule 60(b)(6) relief, and the decision to deny Scott’s application for a certificate of probable cause and stay of his execution.
. And also identical to the language in Shell v. Mississippi, supra.
. It is clear that Scott's use of the term “similar" meant similar to Maynard. The immediately preceding discussion, indeed in the linchpin of Scott's argument, focused on Maynard.
. The majority is misled because most of the space devoted to the HAC claim in Scott's brief addresses the inconsistency of the appellate review function. However, as noted in the text, the consistency of appellate review was relevant only in conjunction with Scott's claim that there was an "absence of guidance for the jury.” Nor did Scott rely merely upon the fact that deficient jury guidance was a logically necessary predicate for his argument; the clear language of his brief framed the argument in that manner.
. I do not believe that Henderson v. Dugger, 925 F.2d 1309 (11th Cir.1991), is inconsistent with my conclusion that on the facts of this case Scott did fairly raise the jury guidance issue. In Henderson, this court stated that Henderson had referred "only to Florida's construction of the aggravating factors.” Id. at 1317. By contrast, in the instant case, as demonstrated above, Scott expressly referred not only to the inconsistent appellate review function, but also to the "absence of guidance for the jury at the trial level.”
.The majority asserts throughout its opinion that none of the claims which Scott presented to the state courts challenged his HAC instruction. I do believe that the issue was not raised at trial. However, unlike the majority, I am not at all sure that the issue was not fairly presented to the state courts, and the State’s mere assertion to that effect rings hollow in the face of their similar assertion (which is erroneous) with respect to Scott's previous appeal to this court. Unfortunately, the record is not available to me to verify whether the State is similarly misreading Scott's presentation to the state courts. In any event, I believe the issue was fairly presented in Scott’s original federal habeas corpus proceeding, and it is clear that the State interposed no procedural bar. Although the State, the district court, and the majority assert that the issue was not fairly presented in Scott’s previous appeal to this court, the language which I quote above and the’ context of Scott’s argument persuade me otherwise. *1557I much prefer to resolve this case on the surer basis which I explain in Part IV of this opinion.
. For a more explicit statement of this erroneous assumption, see Bertolotti v. Dugger, 883 F.2d 1503, 1526-27 (11th Cir.1989), cert. denied, 497 U.S. 1031, 110 S.Ct. 3296, 111 L.Ed.2d 804 (1990); Hargrave v. Wainwright, 804 F.2d 1182, 1192-96 & n. 23; vacated for rehearing en banc, 809 F.2d 1486, rev’d on other grounds, 832 F.2d 1528 (11th Cir.1987), cert. denied, 489 U.S. 1071, 109 S.Ct. 1353, 103 L.Ed.2d 821 (1989).
. In other contexts and under some circumstances, the Supreme Court has condemned post hoc speculation as to which alternative ground informed a jury’s verdict. See Shell, 498 U.S. at 2-4, 111 S.Ct. at 314; Cf. Boyde v. California, 494 U.S. 370, 378-82, 110 S.Ct. 1190, 1197-98, 108 L.Ed.2d 316 (1990). In the very different posture of this case (Rule 60(b)(6) or recall of a mandate), in which Scott must demonstrate injustice or extraordinary circumstances, I am confident that it is appropriate to conduct an analysis to determine whether there is any risk that the jury did in fact rely on the unconstitutionally vague alternative, or whether we can be sufficiently confident that the juty in fact relied only upon the constitutional alternative (i.e., the definition of cruel). No cases suggest otherwise.
. Scott argues that it would be inappropriate for this court to conduct a harmless error review (either following the traditional analysis of Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967), or the new analysis for most federal habeas corpus cases set out in Brecht v. Abrahamson, -U.S.-, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993)). He argues that Supreme Court case law indicates that when an aggravating circumstance has been determined to be unconstitutionally vague, only the state appellate courts can conduct the harmless error review, citing Richmond v. Lewis, - U.S. -, -, 113 S.Ct. 528, 535, 121 L.Ed.2d 411 (1992), Stringer v. Black, -U.S. -, -, 112 S.Ct. 1130, 1140, 117 L.Ed.2d 367 (1992), and Maynard v. Cartwright, 486 U.S. 356, 363-65, 108 S.Ct. 1853, 1859, 100 L.Ed.2d 372 (1988). One argument against Scott's position is that this court could, by application of well-established harmless error review, eliminate any reasonable doubt that "the death sentence has been infected by a vague or otherwise constitutionally invalid aggravating factor.” Richmond v. Lewis, - U.S. at-, 113 S.Ct. at 535. The court could do that by determining beyond any reasonable doubt that the jury relied on the specific cruel definition and not on the unconstitutionally vague portion of the instruction (pursuant to an analysis similar to the one described in the text *1559of this Part IV.B). In other words, through well-established harmless error analysis, this court could determine that the guidance actually applied by the jury was not deficient after all and thus that there is in this particular case no error comparable to the error contemplated in Richmond v. Lewis, Stringer, and Maynard. Those cases contemplated constitutional error infecting the entire HAC factor; thus, those cases contemplated a harmless error review to determine whether other aggravating circumstances, considered in light of the circumstances of the case, were such that an appellate court could conclude that the effect of the vague aggravating factor was harmless beyond a reasonable doubt. The harmless error review contemplated by those Supreme Court cases much more nearly approached a reweighing of the sentencing circum-stancés, which of course would be a sentencing function entirely inappropriate for federal courts to undertake. By contrast the harmless error review which would be called for in this case would leave intact the HAC factor. Although such analysis is clearly distinguishable from the harmless error analysis contemplated in Richmond v. Lewis, Stringer, and Maynard, in this case we need not resolve the question of whether it would be appropriate for this court itself to conduct such a harmless error review. In the peculiar posture of this case, Scott must demonstrate extraordinary circumstances to obtain Rule 60(b)(6) relief or injustice to obtain a recall of the mandate. In this posture, it suffices for this court to satisfy itself that any grant of the writ would be futile because the Florida Supreme Court, in conducting the appropriate harmless error review (and Scott concedes that it would be appropriate for the Florida Supreme Court to do so), would conclude beyond any reasonable doubt that the unconstitutionally vague instruction did not affect the jury's weighing process.