concurring in part and dissenting in part:
I concur in the court’s disposition of Hardwick’s claim of ineffective assistance of counsel at the guilt phase of the trial. I also agree with the majority’s expression of doubt with respect to whether Hard-wick’s counsel rendered effective assistance of counsel at the penalty stage. However, I dissent from the court’s refusal to honor the state court findings of fact and the court’s remand of the case to the district court.
The basis for the court’s remand to the district court for an evidentiary hearing is apparently a conclusion that, under pre-AEDPA law, either
(a) the fact-finding procedure employed by the state 3.850 court was not adequate to afford -a full and fair hearing, 28 U.S.C. § 2254(d)(2); or
(b) its findings of fact were not fairly supported by the record, § 2254(d)(8).
Respectfully, I cannot agree.. The most specific reason indicated by the majority involves the finding of fact of the 3.850 judge relating to whether or not Hard-wick’s trial counsel, Tassone, discussed possible mitigation testimony with Dr. Barnard. In his 3.850 testimony, Dr. Barnard could not recall such conversations with Tassone. On the other hand, .the 3.850 judge found:
Mr. Tassone, on the other hand, had a very clear and vivid recollection of having numerous conversations and discussions with Dr. Barnard regarding possible' mitigation testimony and evidence. This court, having heard and observed both Dr.‘Barnard and Mr. Tassone testify regarding this matter, finds that Mr. Tassone did discuss possible mitigation testimony and evidence with Dr. Barnard.
Supplemental Order, Circuit Court of the Fourth Judicial Circuit, Duval County, Florida, March 21, 1991, at 6-7 (hereafter “3.850 order”).
It is clear from the 3.850 order that the testimony of attorney Tassone was found to be credible in this regard. The findings of fact of the 3.850 court with respect to Dr. Barnard and attorney Tassone are supported by the record and entitled to a presumption of correctness. Indeed, Tas-sone testified that when he questioned Dr. Barnard about a diminished capacity, the doctor indicated that “he could have had diminished capacity ..., but my bottom line is the guy knew exactly what he was doing.” February 22, 1990 hearing at 116. When asked why Dr. Barnard’s written report included nothing with respect to mitigating circumstances, and nothing with respect to diminished capacity, Tassone explained that he thought the report was prepared prior to his talking to Dr. Barnard about those issues.
The majority suggests that the 3.850 proceedings conducted on February 22, 1990, i.e., the examination of Tassone, failed to provide a full and fair hearing because it was conducted with an unprepared, stand-in counsel, principal counsel having been engaged in another previously-scheduled trial, and because .the hearing was conducted beginning in the early afternoon. and not ending until 1 a.m. the next morning. Majority opinion at 1153-54, n. 130 (M/S at 50-54 n. 130), and 1158, n. 141 (M/S at 60-61 n. 141). I disagree. I need not either condone or condemn the *1194procedures employed by the 3.850 judge during the February 22, 1990, hearing (a matter which has not been addressed by the parties), because the Florida Supreme Court granted a stay of execution and remanded the case for a complete eviden-tiary hearing on Hardwick’s claim. On remand, the 3.850 judge conducted hearings on May 3^1, 1990, and August 15-16, 1990, in order to “fully accommodate Mr. Hardwick.” 3.850 order at 1. The state judge expressly 'found: “Mr. Hardwick was allowed to call all desired witnesses.” Id. Thus, to the extent Hardwick might have found fault with the examination or cross-examination of Tassone, the record is clear that there was ample opportunity to recall Tassone either in May or August 1990. Indeed, at the beginning of the evidentiary hearing on remand, on May 3, 1990, the 3.850 judge noted that Tassone’s previous testimony was in the record. However1, he expressly advised the parties that: “I don’t want to preclude anybody from calling Mr. Tassone that wishes to.” May 3, 1990 hearing at 209. Counsel for Hardwick entered no objection, and did not recall Tassone. • I know of no case holding that a deficient hearing by a lower state court forever taints the state proceedings, notwithstanding an appellate remand for a full evidentiary hearing.
The foregoing is the only specific reason apparent to me in the majority’s decision to discount completely the entirety of the 3.850 proceedings, thus discounting other findings of fact by the 3.850 judge. The majority does not seem to challenge the fullness or fairness of the proceedings conducted on remand from the Florida Supreme Court, including the May 1990 and August 1990 evidentiary hearings. Rather, the majority seems to conclude that remaining findings of fact are not fairly supported by the record. 28 U.S.C. § 2254(d)(8). I deduce this from the fact that the majority repeatedly contrasts statements or findings of the 3.850 court with the evidence recited in the majority opinion. Again, I respectfully disagree. After a careful reading of the forceful and comprehensive majority opinion and a careful reading of the 3.850 order and the record, I cannot conclude that there is warrant for disregarding the entirety of the 3.850 proceedings, or for disregarding other crucial findings of fact of the 3.850 judge. In particular, I cannot conclude that the 3.850 judge’s findings of fact with respect to Hardwick’s mother, Nell Lawrence, are not fairly supported by the record. The 3.850 court noted:
Mr. Tassone stated that he was in regular contact with Ms. Lawrence and that she refused to testify, did not get along that well with Mr. Hardwick, and that she felt he deserved a death sentence .... Ms. Lawrence agreed with Mr. Tassone about their regular contact but denied refusing to testify ... [and] agreeing with the result.... This testimony is the clearest conflicting evidence in this case and causes the Court to gauge their respective credibility of these witnesses.
3.850 Order, at 3. The 3.850 court then found as a fact that:
Ms. Lawrence confronted with her child’s execution, has succumbed to internal pressure to save her son’s life by testifying at this late date. As such, her credibility is certainly suspect in comparison to Mr. Tassone. For that reason, the court finds Mr. Tassone’s statement more credible and concludes that Ms. Lawrence was not a willing witness in 1985.
Id. My careful review of the transcript of the 3.850 evidentiary hearing persuades me that the state judge’s fact finding is supported by the record. Tassone testified that he discussed the case with Ms. Lawrence many times, that she “felt great sympathy for her son ... [but] did not want to testify at the penalty phase of the *1195trial.” February 22, 1990, hearing, at 159. The attorney testified that he urged her repeatedly to testify, but that “she finally told me that if, in fact, she did testify she would say that death rather than life was a more appropriate penalty for her son.” Id. Although Ms.. Lawrence’s testimony was directly contrary, the state judge actually heard and saw the testimony of the two witnesses, was in position to evaluate their demeanor, and gave a plausible explanation for finding that Tassone was credible and that Ms. Lawrence was not. I cannot conclude that this credibility finding is lacking in record support.1
The majority justifies its disregard of the state court finding by asserting that the actions of Hardwick’s mother speak louder than Tassone’s recollection of what she said. It is true that the majority lists a number of actions on the part of Hard-wick’s mother indicating a mother’s love and concern for her son, which would seem to be inconsistent with a refusal to testify or with an opinion that her son deserved the death penalty. On the other hand, her son was pretty clearly guilty of a heinous murder and the record does paint a clear picture of her son as leading a very degenerate life involving, inter alia, activities as a drug dealer as well as abuser of alcohol and drugs. Even conceding some considerable probative value in the mother’s actions indicating sympathy, I very respectfully can see no warrant for disregarding the finding of fact by the state judge who observed both Tassone and Nell Lawrence testify, who had an opportunity to observe their demeanor, who made an express credibility finding, and who gave a plausible reason therefore. Such a change of heart on the part of a mother faced with the imminent execution of her son is clearly. plausible.
For the foregoing reasons, I cannot agree with the majority that this court is justified in completely disregarding the 3.850 proceedings and the findings of fact by the state judge. Thus, I doubt our authority to remand for an evidentiary hearing. The majority of course finds authority to remand because it concludes that the state 3.850 proceedings were not full and fair and because it concludes that significant state court findings of fact were not fairly supported by the record. Because I disagree with those two conclusions, and see no other basis for remanding, I must dissent from that decision. However, I agree with the strong doubt which the majority opinion expresses with respect to whether Tassone rendered constitutionally effective assistance of counsel at sentencing. Even accepting the findings of fact of the state judge, my careful review of this record leaves me in considerable doubt as to whether Tassone did render effective assistance of counsel at the penalty stage, and considerable doubt as to whether there is a reasonable probability that effective counsel would have produced a different result.
Even accepting the state court’s finding of fact that Tassone did in fact discuss the possibility of mitigation evidence with Dr. Barnard, the record, as so forcefully described in the majority opinion, does suggest strongly that Tassone was ineffective in failing to provide Dr. Barnard with the information which would have resulted in much different conversations between the doctor and Tassone about mitigating evidence. As the majority points out, it is *1196pretty clear that Tassone did not initially ask Dr. Barnard for an evaluation relating to mitigating evidence, nor did he provide Dr. Barnard with information conducive thereto. My reading of the state judge’s finding of fact is that it finds only that Tassone did discuss possible mitigating testimony with Dr. Barnard. Tassone’s testimony, upon which the finding was based, suggests that this discussion occurred after his report was rendered; in other words, Tassone’s discussion in this regard was more of an afterthought, and what Dr. Barnard told him at that time had to have been influenced by that fact and the fact that Tassone had provided to Dr. Barnard little or none of the evidence which we now know was available. For these reasons, I do not regard the state judge’s finding of fact with respect to Dr. Barnard as an obstacle to a conclusion that Tassone was ineffective.
The other crucial finding of fact by the 3.850 judge involves the testimony of Hardwick’s mother, Nell Lawrence. For the reasons discussed above, I must honor the 3.850 judge’s credibility finding, and presume correct his finding of fact that she refused to testify, and indeed thought at that time that her son deserved the death penalty. Thus, I feel compelled to disregard Nell Lawrence’s testimony in the 3.850 hearing. It is true that Ms. Lawrence was a forceful witness in support of Hardwick in the 3.850 proceedings. However, my review of the record suggests to me that most of the evidence she provided was also provided by one or more other witnesses. In other words, in my review of the record, I discounted the testimony of Nell Lawrence, and evaluated the totality of the circumstances without her 3.850 testimony. My review leaves me with substantial doubt about whether Tassone rendered effective assistance of counsel at sentencing, and whether there is a reasonable probability that the result would have been different.
In other words, after discounting the evidence which we should not consider because of the state judge’s findings of fact (e.g., after discounting the testimony of Hardwick’s mother)2, my review of the record leaves me with considerable doubt about: (a) whether counsel’s performance at sentencing measured up even to that minimum constitutional standard required by Strickland v. Washington, 466 U.S. 668, 698, 104 S.Ct. 2052, 2070, 80 L.Ed.2d 674 (1984), and (b) whether there is a reasonable probability that the result of the penalty phase would have been different had there been effective counsel. The gist of the mother’s testimony — evidence with respect to Hardwick’s alcohol and drug use over that weekend and evidence about his upbringing including his history of alcohol and drug abuse — is duplicated in this record from several sources. Numerous witnesses who were clearly available to Tas-sone at the time testified during the 3.850 proceedings about Hardwick’s extensive use of drugs and alcohol during the relevant weekend.3 With respect to Hard-*1197wick’s harsh upbringing, the other focus of the mother’s testimony, there-was ample other evidence, for example, the. completely objective evidence in the social 'service and juvenile records.
The most plausible explanation for Tas-sone’s failure to pursue the significance of the potential mitigating evidence is either á misunderstanding of how such evidence could be helpful, or a simple lack of investigation and preparation. This hypothesis is not only not inconsistent with the state judge’s finding that- Tassone did discuss possible mitigating evidence with Dr. Barnard, it is actually supported by Tassone’s testimony which the judge credited. Tas-sone essentially acknowledged that he did not discuss mitigating evidence with Dr. Barnard until Dr. Barnard’s report was already tendered. Moreover, the record is clear that Dr. Barnard was not asked in formal fashion to evaluate anything but competence to stand trial and insanity at the time of the offense. Tassone’s own testimony thus supports the inference that Tassone’s discussions with Dr. Barnard about mitigating evidence were'an afterthought. The record is also clear that Tassone provided Dr. Barnard with no background information on Hardwick which would have focused the doctor’s attention on Hardwick’s harsh upbringing or his' history of alcohol and drug abuse. Dr. Barnard’s testimony at the 3.850 hearing and the opinions he expressed there — 'that Hardwick’s capacity to conform his conduct to the requirements of the law was substantially impaired and that his judgment and impulse control were substantially impaired — indicate that it would not have been futile to provide Dr. Barnard with appropriate background information and seek his studied opinion with respect to mitigating evidence. Thus, the state judge’s finding of fact with respect to Dr. Barnard does not ultimately excuse Tas-sone’s failure to investigate and pursue mitigating evidence either with Dr. Barnard or with. some other mental health expert.
In conclusion, although I part company with the majority’s decision to disregard the findings of fact o'f the 3.850 court, and thus must dissent from the decision to remand, I share the majority’s considerable doubt about whether Tassone rendered effective assistance of counsel at the penalty stage of this case. Had the majority concluded on the basis of the current record that counsel was ineffective at the penalty phase, that confidence in that result had been undermined, and thus.that the writ should issue with respect to the penalty phase, I would probably have agreed, even though I would rely on only part of the evidence marshalled so persuasively by the majority opinion. That is, even accepting the above mentioned findings of fact by the state 3.850 judge, I am persuaded that counsel rendered constitutionally ineffective performance at the penalty phase, and my confidence in the result of that phase is undermined. While I disagree with the grounds upon which the majority relies in remanding for an eviden-tiary hearing, and therefore dissent in that regard, I share the concerns of the majority with respect to the penalty phase and concur in much of what the majority says.
. I also cannot agree with the majority’s suggestion that the state court's finding with respect to the mother applied only to the guilt phase. It is true that the language quoted above appeared during the judge’s discussion of ineffective assistance of counsel at the guilt phase. However, in discussing ineffective assistance of .counsel at the sentencing, phase, the judge specifically, incorporated his previous discussion: “Nell Lawrence’s testimony has already been discussed.” 3.850 order at 6.
. The elimination of the testimony of Hard-wick's .mother is clearly the most significant discount we must make. The other findings of the state judge are much less significant. The finding with respect to Dr. Barnard has already been discussed. The judge’s credibility finding with respect to Hardwick’s brother, Jeff Hardwick, related to the guilt phase. Indeed, the 3.850 Order expressly said with respect to the penalty phase: " Jeff Hardwick discussed his testimony with Mr. Tassone and the strategic decision was made not to use it.” 3.850 Order at 6. Similarly, other witnesses discounted by the stale judge, including Connie Wright, were discounted because the state judge honored Tassone's strategic decision to that effect. Of course, it is within our purview to exercise appropriate review with respect to the reasonableness of such strategic decisions.
. Tassone's own testimony at the 3.850 hearing acknowledged awareness of such use in the days leading up to the murder. And *1197strong evidence from wholly disinterested witnesses establishes the drug, alcohol and emotional influences upon him at the time of his arrest. The booking clerk at the jail, Braddy, testified that he was intoxicated when booked late Christmas afternoon. Even the arresting officers testified that shortly after the arrest he became incoherent and out of control.