dissenting.
Law professors write whole books on what the meaning of a “right” is, yet that does not mean that such words cannot be used for ordinary purposes by people of average, or indeed below-average, intellect. To invalidate a waiver of Miranda rights *418because a person of limited IQ cannot give satisfactory definitions of words like “right” is to make it practically impossible for police to rely on objectively reasonable agreements on the part of such persons to talk with police. Nothing in the policies underlying Miranda mandates such an unreasonable obstacle to desirable police procedures.
I am therefore compelled to disagree with the conclusion of the majority opinion in this case that the defendant did not knowingly and intelligently waive his Miranda rights. The district court determined that the waiver was knowing and intelligent, based on the court’s careful analysis of the record and of the evidence of the expert who administered an evaluative test on the defendant. The district court’s factual conclusion in this regard is compelled by the district court’s thoughtful analysis, see Dist. Ct. Op. at 12-26, and is obviously not erroneous, much less clearly erroneous.
To overturn such a factual determination on the basis of our independent appellate review is to create a wholly unwarranted rule of law. To rely essentially on the low score of defendant on a test, applied six years after the relevant waiver — when the test is scored low because the testee does a poor job of explaining the meaning of words such as “rights,” “attorney,” and “interrogation” — is to create a powerful litigation tool. That tool can easily become an engine that will effectively preclude the interrogation by police of criminal suspects in custody who are not articulate enough to convey effectively what they may basically understand. Because it is unrealistic to expect most criminal suspects to be able to explain abstract concepts in an articulate fashion, the rule created will bring into question the bulk of statements by persons in custody, no matter how reasonable and careful the police have been in giving Miranda warnings.
There is no argument that the police in this case were not reasonable or careful in giving the warnings. After virtually each element of the Miranda warning the police asked and obtained assurance that the suspect understood the meaning. Words with a potential for misunderstanding- — such as “attorney” — were, indeed, simplified (e.g., to “lawyer”). Importantly, there is nothing in the record to indicate that the police were made aware that there was a lack of understanding. It is not apparent what more the police could do, short of administering the Grisso test themselves.
Miranda cannot logically be extended to protect the hidden misunderstandings of suspects, where the police have been objectively reasonable in obtaining a waiver. The underlying interest protected is the right of suspects not to talk when they don’t want to, or when they would prefer to have a lawyer. Miranda is a protective rule. That is, Miranda protects the underlying right, in part, by requiring the police to obtain an effective waiver, without which the information cannot be used. But if evidence is excluded notwithstanding proper police conduct, the deterrent aspect of Miranda is simply not applicable. It is the logical equivalent of saying that police violate the knock-and-announce rule for warrant-authorized home entries when the police do knock and announce but the inhabitant, unknown to the police, is deaf.
To succeed with his Miranda claim, Garner needed to prove that, under the totality of the circumstances, he did not knowingly and intelligently waive his rights. Clark v. Mitchell, 425 F.3d 270, 283 (6th Cir.2005). The objective evidence in this case, however, demonstrated that Garner did waive his rights knowingly and intelligently. The undisputed evidence *419shows that Garner appeared “perfectly normal” and “very coherent” when officers read him his Miranda rights and when he confessed to his crimes. JA 944. The evidence also shows that Garner stated that he understood the term “waiver” and that he responded to each Miranda warning by indicating that he understood the warnings. JA 955. The Ohio Supreme Court, moreover, found that Garner signed a waiver of rights form and acknowledged verbally that he had previously executed a waiver. State v. Garner, 74 Ohio St.3d 49, 656 N.E.2d 623, 635 (1995). Finally, at the time of Garner’s interrogation, there were no obvious signs that Garner was mentally disabled, unable to understand the instructions, or under the influence of drugs or alcohol. That is, all objective evidence pointed to Garner’s knowing and intelligent waiver.1
It is a mistake to rely entirely on Garner’s subjective understanding of the Miranda warnings instead of relying on objective signs that Garner’s waiver was knowing and intelligent. A purely subjective approach deviates from the original purpose of the Miranda warnings, namely, “to protect the suspect’s privilege against compulsory self-incrimination.” Young v. Walls, 311 F.3d 846, 850 (7th Cir.2002). As the Supreme Court explained in New York v. Quarles, “[t]he Miranda decision was based in large part on this Court’s view that the warnings which it required police to give to suspects in custody would reduce the likelihood that the suspects would fall victim to constitutionally impermissible practices of police interrogation.” 467 U.S. 649, 656, 104 S.Ct. 2626, 81 L.Ed.2d 550 (1984). Here, there is no evidence that authorities compelled Garner to testify against himself, and the police officers’ objective understanding (of the suspect’s subjective understanding) should be the ultimately determinative factor in the majority’s analysis. As the Seventh Circuit has reasoned, the “relevant constitutional principles are aimed' not at protecting people from themselves but at curbing abusive practices by public officers.” Rice v. Cooper, 148 F.3d 747, 750 (7th Cir.1998). Judge Posner’s analysis in Rice is thoughtful and instructive: Of course if the subject is a small child, or obviously can’t speak English, or is apparently so mentally ill or retarded as not to be able to make a rational choice, that objectively observable lack of subjective understanding invalidates a Miranda waiver. See id. On the other hand,
[o]n this analysis, the knowledge of the police is vital. If they have no reason .... to think that the suspect doesn’t understand them, there is nothing that smacks of abusive behavior. It would seem to follow that the question is not whether if [the subject] were more intelligent, informed, balanced, and so forth he would not have waived his Miranda rights, but whether the police believed he understood their explanation of those rights; more precisely, whether a reasonable state court judge could have found that the police believed this.
Id. at 750-51; see also Taylor v. Rogers, No. 95-3904, 1996 WL 515349, at *3 (6th Cir. Sept.10, 1996) (considering objective factors in determining whether consent was knowing and intelligent); United States v. Turner, 157 F.3d 552, 555 (8th *420Cir.1998) (finding, based only on objective signs, that consent was knowing and intelligent); Starr v. Lockhart, 23 F.3d 1280, 1294 (8th Cir.1994) (same); Derrick v. Peterson, 924 F.2d 813, 824 (9th Cir.1990) (relying, in part, on objective signs to find waiver); United States v. Rojas-Tapia, 446 F.3d 1, 7-8 (1st Cir.2006).
Moreover, as even the majority opinion recognizes, a purely subjective approach will “put the police in the difficult position of having to assess a suspect’s understanding and intellectual capacities at the time of interrogation.” Maj. Op. at 408 n. 5. That is, police departments will never know whether a suspect who confessed will claim years later that, contrary to objective signs at the time, he subjectively failed to consent; and these departments will need to hire mental-health professionals to divine the subjective intent of all defendants. These costs create no discernible benefits.
The district court’s analysis, in short, properly considered evidence that Garner knowingly and intelligently waived, his Miranda rights, and it would be wrong to assign no value to this objective evidence.
Even if it were proper to disregard contemporaneous objective evidence that Garner knowingly and intelligently waived his Miranda rights, the evidence of Garner’s subjective abilities in this case does not require reversal. The majority opinion places great reliance on the expert opinion of Caroline Everington, Ph.D., an educational and forensic psychologist, who stated that Garner lacked the “full comprehension of Miranda warnings [and] his right to remain silent.” Maj. Op. at 413.2 As the district court noted, there are serious concerns with the accuracy of Evering-ton’s assessment.
First, as the district court observed, Ev-erington, who administered the so-called Grisso test, did not claim to be licensed as a clinical psychologist or licensed for psychiatric practice. This is important because the Grisso test requires “mental health professionals who are licensed for clinical psychological or psychiatric practice in their state, and who are qualified by training and experience to perform evaluations for use by courts and attorneys ... in criminal cases” to administer the test. See Thomas Grisso, Instruments for Assessing Understanding & Appreciation of Miranda Rights 2 (1998). Presumably, the Grisso test requires that a professional administer the questions to avoid errors, errors that might have occurred in this case, which result from having a non-professional administer the test.3
Second, the district court expressed concerns over the accuracy of the Grisso test because Everington administered the test almost seven years after the police interrogation and after the imposition of Garner’s death sentence. These are serious concerns because the Grisso test only provides an “index of the person’s capacities for understanding the Miranda warnings at the time of the evaluation!,]” not at the time of the police interrogation, Grisso, supra, at 7, and a defendant who is capable (whether through the insinuation of *421counsel or through other means) to understand the meaning and importance of the Grisso test might feign misunderstanding to avoid a death sentence. This court has no way of knowing whether the test accurately reflected Garner’s abilities at the time that he waived his Miranda rights or whether Garner feigned misunderstanding. One cannot brush aside these serious concerns and accuse the district court of committing plain error by simply noting that, in general, Grisso test results are generally positively correlated with age.
Third, the district court noted that Ev-erington asked Garner whether he understood a Miranda warning with complex terms (ie., “consult,” “attorney,” “interrogation”) when the actual interrogation at issue in this case involved less complicated terms (ie., “talk,” “lawyer,” “questioning”). The district court was correct to find that the manner in which Everington questioned Garner could have skewed the results. See generally Morgan Cloud et al., Words Without Meaning: The Constitution, Confessions, and Mentally Retarded Suspects, 69 U. Chi. L.Rev. 495, 581 (2002) (chart showing that 49% of disabled participants in a survey understood the simplified term “lawyer” while only 40% understood the Miranda term “attorney”).4
The majority opinion minimizes or disregards other evidence that Garner was capable of subjectively waiving his Miranda rights. For example, the district court observed that Garner admitted that he started the fire to create a smokescreen. The confession suggests that, at the time, Garner understood the consequences of committing theft and therefore had the capacity to understand the consequences of waiving his rights. See United States v. Macklin, 900 F.2d 948, 952 (6th Cir.1990). In addition, as the district court noted, the competency report stated that Garner was “near average intelligence” and “able to understand all questions and materials presented to him.”5
The Ohio courts in this case essentially determined that Garner knowingly and intelligently waived his Miranda rights, although their conclusion appears in a slightly different context and without the benefit of Everington’s observations. In rejecting Garner’s claim that his counsel was ineffective in failing to raise this issue, the state trial court, for example,, considered all the evidence that was available at that time and concluded that Garner’s counsel had no essential duty to claim that Garner could not understand the Miranda warnings. JA 188. Because the bulk of the trial court’s analysis deals with the issue of *422whether Garner could have knowingly and intelligently waived his Miranda rights, the trial court’s conclusion that Garner did not suffer ineffective assistance of counsel appears to be the result of the trial court’s conclusion that there was no merit to Garner’s Miranda claim because the evidence established that Garner knowingly and intelligently waived his rights. The Ohio Court of Appeals also rejected Garner’s ineffective assistance of counsel argument, after reviewing the record, State v. Garner, No. C-960995, 1997 WL 778982, at *3 (Ohio App. Dec.19, 1997), suggesting that it too did not believe that Garner lacked the ability to waive his Miranda rights.
If there were any remaining doubt that Garner knowingly and intelligently waived his rights, the fact that the Ohio courts considered and implicitly rejected Garner’s Miranda claim bolsters the conclusion that Garner did not suffer a constitutional violation. Although the issue before this court and the issue before the Ohio Courts are not identical, in that this court must decide whether there was merit to Garner’s Miranda claim and the Ohio courts determined whether counsel was ineffective in not raising Garner’s Miranda claim, the Ohio courts clearly considered Garner’s argument that the evidence demonstrated .that he lacked the capacity to consent. After considering that evidence, the Ohio courts found that counsel was not ineffective, and the courts’ evaluation of that evidence should help guide our analysis of the Miranda claim, see Filiaggi v. Bagley, 445 F.3d 851, 854 (6th Cir.2006), especially considering that the Ohio courts reached the correct result.
In addition, I question the decision to review an issue that Garner procedurally defaulted but as to which the state failed to argue procedural default in the district court. First, it is not clear whether, under AEDPA, a state can forfeit a procedural default defense based on failure to exhaust a.remedy no longer available, absent an express waiver. In cases to which AED-PA applies, such as this one, “[a] State shall not be deemed to have waived the exhaustion requirement or be estopped from reliance upon the requirement unless the State, through counsel, expressly waives the requirement.” 28 U.S.C. § 2254(b)(3). In Banks v. Dretke, 540 U.S. 668, 705, 124 S.Ct. 1256, 157 L.Ed.2d 1166 (2004), the Supreme Court noted that “under pre-AEDPA law, exhaustion and procedural default defenses could be waived based on the State’s litigation conduct,” but that “AEDPA forbids a finding that exhaustion has been waived unless the State expressly waives the requirement.” The close conceptual relationship between the distinct doctrines of procedural default and exhaustion suggests that express waiver should be required for both. The Eleventh Circuit has explicitly held that, although § 2254(b)(3) by its language applies only to exhaustion, the section “applies with full force in cases ... where the procedural bar arises only as a direct result of the petitioner’s failure to exhaust his state law remedies.” McNair v. Campbell, 416 F.3d 1291, 1305 (11th Cir.2005); see also Gonzales v. McKune, 279 F.3d 922, 924 (10th Cir.2002) (en banc). But see Franklin v. Johnson, 290 F.3d 1223, 1231 (9th Cir.2002) (“[Section] 2254(b)(3)’s reference to exhaustion has no bearing on procedural default defenses.”). The Eleventh Circuit reasoned that “[b]e-cause § 2254(b)(3) provides that the State can waive [petitioner’s] failure to properly exhaust his claim only by expressly doing so, it logically follows that the resulting procedural bar, which arises from and is dependent upon the failure to properly exhaust, can only be waived expressly.” McNair, 416 F.3d at 1305.
The majority in this case relies upon our decisions in Sowell v. Bradshaw, 372 F.3d *423821 (6th Cir.2004), and Howard v. Bouchard, 405 F.3d 459 (6th Cir.2005), as permitting discretionary disregard of a procedural default argument not raised in the district court. These cases simply do not address the question of whether, under AEDPA, a state may implicitly waive a procedural default based on failure to exhaust a presently unavailable state remedy. Sowell was a pre-AEDPA case to which § 2254(b)(3) did not apply. It is true that this court in a post-AEDPA case relied on Sowell for the proposition that we are “permitted to consider the procedural default issue even when raised for the first time on appeal if we so choose,” thereby suggesting by negative inference that refusal to consider procedural default is also within our discretion when raised for the first time on appeal. White v. Mitchell, 431 F.3d 517, 524 (6th Cir.2005). But the White opinion did not address the possible applicability of § 2254(b)(3) and did not exercise any discretion that might have been implied to refuse to consider procedural default. Indeed, the White court denied relief on the issue in question on the basis of procedural default. 432 F.3d at 525.
Howard is also very different. In Howard, the state raised the procedural default argument in district court but failed to reassert the argument on appeal. 405 F.3d at 476. First, Howard involved waiver in the court of appeals rather than in the district court, and the discretion involved in Howard was whether to affirm a judgment on grounds presented below but not argued on appeal — a traditionally broad appellate court discretion. Second, and more fundamentally, Howard held that it was within the appellate court’s discretion to invoke procedural default that had arguably been waived. Id. The conclusion was proper regardless of whether any waiver had to be express. This is simply not a holding that it is within the court’s discretion not to invoke procedural default where1 procedural default had arguably been waived. The express waiver requirement of AEDPA, which had no effect on the resolution of the former question, is dispositive of the latter question, under the Eleventh Circuit’s analysis. Thus, the question of our discretion to refuse to consider a procedural default claim not raised below, where the procedural default consists of a failure to exhaust a remedy no longer available, remains open in this circuit. In my view, the reasoning of the Eleventh Circuit in McNair is persuasive, and procedural default accordingly precludes our reaching the Miranda waiver issue in this case.
Second, even if we have the discretion to disregard the procedural default because of the state’s failure to argue procedural default in the district court, it is inconsistent with the guiding principles of AEDPA to exercise that discretion in the context of this case. In Sowell, we exercised the discretion “[i]n light of the resources that have been expended by the district court and the serious consequences facing So-well.” 372 F.3d at 830. However, in White, a post-AEDPA case, we held— without giving particular reasons for not exercising Sowell discretion — -that procedural default should bar a death-row defendant’s claim even though the State did not raise procedural default in the federal district court. 431 F.3d at 524-25. Thus Sowell cannot be read to require the dispensation of the procedural default requirement simply because the stakes are high. And the post-AEDPA White case can be read as at least implicitly taking into account state-comity considerations of the type that drove the enactment of AEDPA.
Such considerations counsel against disregarding procedural default in this case, notwithstanding the state’s failure to raise *424the procedural default of the Miranda waiver competence issue in the district court. The Seventh Circuit in similar circumstances assumed arguendo that it had the discretion post-AEDPA to reach a procedurally defaulted claim because the state failed to raise procedural default in the district court, but that court found it appropriate to reach the state’s procedural default defense for several reasons. Perruquet v. Briley, 390 F.3d 505, 516-19 (7th Cir.2004). First, the procedural default was clear, id. at 518, as it is in this case. Second, “because no [state] court was ever given the opportunity to pass on the merits of [petitioner’s] constitutional claim, comity and federalism principles weigh strongly against permitting [petitioner] to assert the claim in federal court.” Id. This is true in the present case. Indeed, this consideration weighs particularly strongly where — -as here — the state court’s lack of opportunity to pass on the merits was not the result of, for instance, a state court’s erroneous application of some procedural hurdle or the ineffective assistance of counsel appointed by the state courts. Third, in Judge Rovner’s words,
if we were to reach the merits of [petitioner’s] constitutional claim, we necessarily would have to do so de novo, as there is no state-court decision we can look to for an evaluation of this claim. This would be inconsistent with the high level of deference to state-court decisions that Congress mandated when it passed the Antiterrorism and Effective Death Penalty Act of 1996. It would also amount to a windfall for [petitioner], who would win plenary review of a claim that he never presented to the [state] courts, whereas habeas petitioners who properly present their claims to state courts first are entitled only to the extremely narrow review mandated by section 2254(d).
Id. (citations omitted). This consideration directly applies in this case where the majority has rejected modified AEDPA review in favor of de novo review on the theory that the precise issue of voluntary and intelligent waiver was not necessarily determined by the state courts. (Were modified AEDPA review to apply, this factor would weigh less in favor of considering procedural default.)6 All of these considerations strongly counsel in favor of considering the state’s procedural default contention raised for the first time on appeal. Since no real argument is put forward that the Miranda waiver competence issue was not procedurally defaulted, I would affirm in the alternative on that basis alone.
Finally, none of Garner’s other claims requires habeas relief. First, because Garner’s Miranda claim lacks merit, his counsel was not ineffective for failing to investigate the claim or to raise it before the state courts. Second, Garner claims that the state trial court’s denial of expert assistance unfairly kept him from developing evidence that he could not have knowingly and intelligently waived his Miranda rights. This claim lacks merit because the assistance that the experts would have given would not have been sufficient to show that his waiver was intelligent. See Dist. Ct. Op. at 62. Finally, Garner claims that the process for selecting the petit jury venires violated his constitutional rights. Garner admits that he did not present this claim to the state courts. See Appellant’s Supp. Reply Br. at 2. For the reasons given by the district court, this claim was *425procedurally defaulted and in any event is without merit. See Dist. Ct. Op. at 27-34. For the thoughtful and extensive reasons provided by the district court on these issues, I would affirm.
. The majority opinion discusses at some length the affidavit of Dr. Jeffrey Smalldon. See Maj. Op. at 411. Dr. Smalldon, however, did not conclude that Garner did not knowingly and intelligently waive his Miranda . rights. Rather, Dr. Smalldon merely called for a more "focused assessment.” JA 922.
. Gamer argues that Everington is a nationally known expert with extensive background in testing mental retardation and is a certified forensic examiner by the American Board of Forensic Examiners. However, Gamer did not establish Everington’s credentials before the district court and did not &sk for an evi-dentiary hearing to allow Everington to testify-
. The majority opinion appears to suggest that mere inability to explain the two terms that appeared in both Everington's test and the actual warning at issue in this case (''appoint” and "right”) is sufficient to invalidate a waiver of Miranda rights even if the suspect has no difficulty in recognizing the meaning of the Miranda warning when presented in a true-false format, as is the case here. Such a sweeping holding threatens to preclude police from taking a vast number of otherwise proper statements.
. The district court did not take these statements out of context, but quoted a passage from the competency report in full. The court quoted on page 13 of its opinion that:
[Garner] appeared to be of near average intelligence by observation. His memory appeared to be intact. He appeared to be able to understand all questions and material presented to him suggesting that his receptive language is intact. Likewise, his expressive language abilities were intact. He was familiar with the specifics of the allegations against him. Mr. Gamer was able to give a coherent, realistic account of his behavior relevant to the allegations although his account differed in a couple of major respects is [sic] from the statement he made to police.
Dist. Ct. Qp. at 13.
. The Seventh Circuit also relied on its observation that the federal issue in that case required substantial familiarity with elements of state criminal law. Perruquet, 390 F.3d at 518. While that particular consideration does not apply in this case, the issue here, on the other hand, is one of great impact on the conduct of state law enforcement systems.