Roosevelt Arthur Williams v. State of Arizona

FILED BY CLERK
IN THE COURT OF APPEALS MAY 17 2013
STATE OF ARIZONA
DIVISION TWO COURT OF APPEALS
DIVISION TWO

ROOSEVELT ARTHUR WILLIAMS, ) 2 CA-SA 2012-0070
) DEPARTMENT A
Petitioner, )
) OPINION
v. )
)
HON. PETER J. CAHILL, Judge of the )
Superior Court of the State of Arizona, in )
and for the County of Pima, )
)
Respondent, )
)
and )
)
THE STATE OF ARIZONA, )
)
Real Party in Interest. )
)

SPECIAL ACTION PROCEEDING

Pima County Cause No. CR20103630001

JURISDICTION ACCEPTED; RELIEF DENIED

Lori J. Lefferts, Pima County Public Defender
By Sean Bruner and Dawn Priestman Tucson
Attorneys for Petitioner

Barbara LaWall, Pima County Attorney
By Jacob R. Lines Tucson
Attorneys for Real Party in Interest

H O W A R D, Chief Judge.
¶1 In this statutory special action, Roosevelt Williams challenges the

respondent judge’s ruling that he failed to establish, by clear and convincing evidence, an

intellectual disability rendering him ineligible for the death penalty in his pending

prosecution for murder. Our consideration of the merits of Williams’s petition is

mandatory. See A.R.S. § 13-753(I). For the following reasons, we deny relief.

Background

¶2 As a matter of statutory and constitutional law, a person convicted of a

capital offense who suffers from an intellectual disability, previously known as mental

retardation, may not be sentenced to death. § 13-753(H); Atkins v. Virginia, 536 U.S.

304, 321 (2002) (Eighth Amendment prohibits execution of mentally retarded persons).1

Under § 13-753(K)(3), an intellectual disability is defined as follows:

[A] condition based on a mental deficit that involves
significantly subaverage general intellectual functioning,
existing concurrently with significant impairment in adaptive
behavior, where the onset of the foregoing conditions
occurred before the defendant reached the age of eighteen.

1
Since Atkins was decided, the American Association on Intellectual and
Developmental Disabilities (AAIDD), formerly known as the American Association on
Mental Retardation (AAMR), has changed the designation of this disability from “mental
retardation” to “intellectual disability.” See Coleman v. State, 341 S.W.3d 221, 226 n.5
(Tenn. 2011) (noting change; terms “mental retardation” and “intellectual disability”
“interchangeable”); Robert L. Schalock et al., The Renaming of Mental Retardation:
Understanding the Change to the Term Intellectual Disability, 45 Intell. &
Developmental Disabilities 116, 120 (2007) (“intellectual disability” “currently preferred
term” for mental health profession to describe same “population of individuals who were
diagnosed previously with mental retardation in number, kind, level, type, and duration of
the disability”). Consistent with this change in terminology, § 13-753 was amended in
2011 to substitute “intellectual disability” for “mental retardation” with no substantive
changes to the statute. See 2011 Ariz. Sess. Laws, ch. 89, § 5.
2
The statute further defines “[s]ignificantly sub-average general intellectual functioning”

as “a full scale intelligence quotient of seventy or lower,” taking into account “the margin

of error for the test administered.” § 13-753(K)(5). “‘Adaptive behavior’” is defined as

“the effectiveness or degree to which the defendant meets the standards of personal

independence and social responsibility expected of the defendant’s age and cultural

group.” § 13-753(K)(1).

¶3 Williams was indicted for two counts of first-degree murder, and the state

filed a notice of its intent to seek the death penalty. In accordance with § 13-753(B), the

respondent judge appointed a “prescreening psychological expert” to evaluate Williams’s

intelligence quotient (IQ). Upon that expert’s report that Williams’s IQ test score was

less than seventy-five, the respondent appointed additional experts and scheduled an

evidentiary hearing to determine whether Williams suffers from an intellectual disability

and therefore is ineligible for a death sentence. See § 13-753(D).

¶4 After the evidentiary hearing, the respondent judge found Williams had

“met his burden of showing that, at least currently, he presents with significantly sub-

average general intellectual functioning” based on “[t]wo valid IQ test scores” of sixty-

eight and seventy and the neuropsychological assessment performed by defense expert

James Sullivan.2 But the respondent further found Williams had “not met his burden of

showing that the mental impairment existed concurrently with significant adaptive

2
Sullivan administered the Wechsler Adult Intelligence Scale 4 (WAIS-4), and
reported Williams had a full-scale IQ of seventy, with ninety-five percent confidence that
his “true IQ” would fall between sixty-seven and seventy-five. Psychologist Serena
Gorgueiro administered the Stanford-Binet 5 and reported Williams had a full-scale IQ of
sixty-eight, with a range of sixty-five to seventy-three.
3
behavior impairment or that the onset of the conditions occurred before [he] reached the

age of [eighteen].” Accordingly, the respondent denied Williams’s request to dismiss the

state’s notice of its intent to seek the death penalty.

¶5 For the most part, Williams does not dispute the respondent judge’s

thorough summary of the evidence presented at the hearing. Instead, he contends the

respondent abused his discretion in applying § 13-753 “in such a manner that it violated

the Eighth Amendment to the United States Constitution and article II § 15 of the

Arizona Constitution.” But Williams does not articulate clearly the basis for a

constitutional claim; instead, he challenges the respondent’s reliance on certain evidence

and his rejection of other evidence in concluding Williams had failed to sustain his

burden of proof. According to Williams, clear and convincing evidence not only

established that he suffers from impairments in intellectual and adaptive functioning

contemplated by § 13-753, but also established, as required, that the onset of these

conditions occurred before the age of eighteen.

Discussion

¶6 At a hearing conducted in accordance with § 13-753, “the defendant has the

burden of proving intellectual disability by clear and convincing evidence.” § 13-753(G);

see also State v. Grell (Grell II), 212 Ariz. 516, ¶ 29, 135 P.3d 696, 702 (2006) (statute’s

burden of proof requirements constitutionally permissible). We defer to the respondent

judge’s factual findings if they “are supported by the record and not clearly erroneous.”

State v. Rosengren, 199 Ariz. 112, ¶ 9, 14 P.3d 303, 307 (App. 2000). Moreover, a trial

judge “‘has broad discretion in determining the weight and credibility given to mental

4
health evidence’” presented in an Atkins hearing. Grell II, 212 Ariz. 516, ¶ 58, 135 P.3d

at 708, quoting State v. Doerr, 193 Ariz. 56, ¶ 64, 969 P.2d 1168, 1181 (1998). We

review legal questions, including questions of constitutional law, de novo, id. ¶¶ 22, 55,

but we will not upset a legal determination that evidence was insufficient to meet a clear

and convincing standard unless we can say “as a matter of law that no one could

reasonably find that the evidence . . . was less than clear and convincing.” Groth v.

Martel, 126 Ariz. 102, 103, 612 P.2d 1065, 1066 (App. 1979); see also State v. West, 226

Ariz. 559, ¶ 15, 250 P.3d 1188, 1191 (2011) (“[Q]uestion of sufficiency of the evidence

is one of law . . . .”).

¶7 In Atkins, the United States Supreme Court held that executing a mentally

retarded offender violates the Eighth Amendment’s ban on cruel and unusual punishment.

536 U.S. at 321. The Court announced this categorical rule based on a “national

consensus,” evinced by prohibitions enacted by state legislatures, that mentally retarded

persons are “categorically less culpable than the average criminal” and more vulnerable

to wrongful execution. Id. at 315-21. According to the Court, “[t]o the extent there is

serious disagreement about the execution of mentally retarded offenders, it is in

determining which offenders are in fact retarded.” Id. at 317.

¶8 The Court cited clinical definitions of mental retardation found in the

Diagnostic and Statistical Manual of Mental Disorders (4th ed. 2000) (DSM-IV),

published by the American Psychiatric Association (APA), and Mental Retardation:

Definition, Classification, and Systems of Supports (9th ed. 1992), published by the

5
American Association on Mental Retardation (AAMR),3 stating that both definitions

require evidence of “subaverage intellectual functioning . . . [and] significant limitations

in adaptive skills such as communication, self-care, and self-direction that became

manifest before age [eighteen].”4 Id. at 318 & n.3. Adopting the approach it chose when

it prohibited execution of offenders who are insane, the Court left “‘to the State[s] the

task of developing appropriate ways to enforce the constitutional restriction upon [their]

execution of sentences,’” noting that “statutory definitions of mental retardation” in states

that legislatively had prohibited execution “are not identical, but generally conform to the

clinical definitions” promulgated by the APA and the AAMR. Id. at 317 & n.22, quoting

Ford v. Wainwright, 477 U.S. 399, 416-17 (1986) (alterations in Atkins).

¶9 Arizona is among the states that had enacted legislation prohibiting the

execution of mentally retarded offenders before Atkins was decided. Id. at 315; State v.

3
Now the American Association on Intellectual and Developmental Disabilities
(AAIDD), see supra note 1.
4
This description continues to be accurate. The AAIDD since has modified
diagnostic criteria that had required “‘limitations in at least 2 of the 10 specific skill areas
listed in [the AAMR’s] 1992 definition’” and had been “the model for the approach still
used by the APA.” United States v. Hardy, 762 F. Supp. 2d 849, 879 (E.D. La. 2010),
quoting AAMR, Mental Retardation Definition, Classification, and Systems of Supports
73 (10th ed. 2002) [hereinafter AAMR 10th Edition] (alteration in Hardy); see also
Atkins, 536 U.S. at 309 n.3. Since 2002, the AAMR/AAIDD has defined the disability as
“‘characterized by significant limitations both in intellectual functioning and in adaptive
behavior as expressed in conceptual, social, and practical adaptive skills’” that originate
before age eighteen, explaining that using these “‘three broader domains of conceptual,
social, and practical skills’” to identify adaptive limitations is “‘more consistent with the
structure of existing measures and with the body of research on adaptive behavior.’”
Hardy, 762 F. Supp. 2d at 852, 879, quoting AAMR 10th Edition at 1, 73, 78.

6
Grell (Grell I), 205 Ariz. 57, ¶ 38, 66 P.3d 1234, 1240 (2003); see also 2001 Ariz. Sess.

Laws, ch. 260, § 2. As summarized by our supreme court,

[Section 13-7535] involves several steps in which experts
examine a capital defendant “using current community,
nationally and culturally accepted physical, developmental,
psychological and intelligence testing procedures, for the
purpose of determining whether the defendant has mental
retardation.” The experts submit reports and the trial court
holds a hearing at which the defendant bears the burden of
proving mental retardation by clear and convincing evidence.
A finding by the trial court of mental retardation prohibits the
imposition of the death penalty.

Grell I, 205 Ariz. 57, ¶ 39, 66 P.3d at 1240, quoting § 13-753(E) (citations omitted). The

court noted Arizona’s statute “appears to comport substantively and procedurally with the

principles set forth in Atkins,” id. n.4, and remanded the case for a determination of

whether Grell, who had been sentenced before the statute took effect, was “mentally

retarded and therefore ineligible to receive the death penalty” pursuant to constitutional

principles announced by the Supreme Court in Atkins, id. ¶¶ 41-42.

¶10 In an appeal after remand, our supreme court affirmed the trial court’s

finding that Grell had failed to establish mental retardation by clear and convincing

evidence. Grell II, 212 Ariz. 516, ¶ 63, 135 P.3d at 709. The court rejected Grell’s

arguments that § 13-753 is unconstitutional because it places the burden of proving

mental retardation on the defendant; because it requires the defendant to prove mental

retardation by clear and convincing evidence; or because it permits capital punishment in

5
Section 13-703.02, A.R.S., cited in Grell I, has been renumbered as § 13-753.
2008 Ariz. Sess. Laws, ch. 301, § 26. We refer here to the current statute.

7
the absence of a jury finding, beyond a reasonable doubt, that the defendant is not

mentally retarded. Id. ¶¶ 29, 41, 49.

¶11 Relevant to some of the issues Williams raises, Grell also had argued the

trial court erred in finding the evidence insufficient to establish he had a significant

impairment in adaptive behavior and asserted he had “clearly shown that [he] has deficits

in two of the eleven areas listed in the [APA’s] DSM-IV and therefore has mental

retardation.” Id. ¶ 62. Our supreme court explained,

The DSM-IV definition of mental retardation, . . . while
similar in overall meaning, is not the same as the statutory
definition. The statute requires an overall assessment of the
defendant’s ability to meet society’s expectations of him. It
does not require a finding of mental retardation based solely
on proof of specific deficits or deficits in only two areas.

Id. With this standard in mind, the court found the evidence was sufficient to “support a

finding that Grell was able to function at a level higher than that of ‘significant

impairment’” and concluded the trial court did not clearly err in finding Grell had “failed

to prove mental retardation by clear and convincing evidence.” Id. ¶ 63. The court also

found Grell was entitled to be resentenced by a jury, pursuant to the Supreme Court’s

decision in Ring v. Arizona, 536 U.S. 584, 589 (2002), and remanded the case for that

proceeding. Id. ¶¶ 64-67.

¶12 On remand, a jury imposed the death penalty, and Grell’s case was returned

to our supreme court on automatic appeal, pursuant to A.R.S. § 13-755(A). State v. Grell

(Grell III), 231 Ariz. 153, ¶ 1, 291 P.3d 350, 351 (2013). The court “independently

8
review[ed] the propriety of the death sentence,”6 considered whether Grell had

established mental retardation in the penalty phase of his resentencing hearing, and

concluded, “Grell is not subject to the death penalty by reason of mental retardation.” Id.

¶¶ 3-4, 10.

¶13 The court emphasized in Grell III that its “inquiry differ[ed]” from the

questions raised in Grell II, in which it had deferred to the trial court’s determination

“that Grell had not proved by clear and convincing evidence that he had significant

deficits in adaptive behavior” because “‘[r]easonable minds [could] differ as to how to

interpret the evidence presented.’” Id. ¶¶ 9-10, quoting Grell II, 212 Ariz. 516, ¶ 63, 135

P.3d at 709 (alteration in Grell III). In contrast to its inquiry in Grell II, the court

explained it was required in Grell III to “independently review the evidence presented in

the 2009 resentencing,” without deference to the jury’s findings or decisions, “to

6
The court noted Grell’s death sentence was subject to independent review because
he had committed the murder before August 1, 2002. Grell III, 231 Ariz. 153, ¶ 3, 291
P.3d at 351; see also § 13-755(A). For capital murders committed on or after August 1,
2002, our supreme court does not conduct an independent review of the propriety of a
death sentence, but “determine[s] whether the trier of fact abused its discretion in finding
aggravating circumstances and imposing a sentence of death.” See A.R.S. § 13-756(A);
State v. Cota, 229 Ariz. 136, ¶¶ 90-92, 272 P.3d 1027, 1044 (2012) (section 13-756 sets
appellate standard of review; review of death sentence for abuse of discretion under § 13-
756 constitutionally permissible; Constitution does not require independent review of
whether “appellate court itself would have imposed a death sentence”). It appears
Williams’s claim of mental retardation would be subject to the court’s automatic review
of the propriety of any death sentence that may be imposed in the future, although the
evidence may be reviewed for an abuse of discretion, rather than independently. Cf.
A.R.S. § 13-752 (“The trier of fact shall make all factual determinations required by this
section or the Constitution of the United States or this state to impose a death sentence. If
the defendant bears the burden of proof, the issue shall be determined in the penalty
phase.”); § 13-756(A).

9
determine whether Grell proved mental retardation by a preponderance of the evidence.”

Id. ¶ 10. The court further noted its determination in Grell III had been based on a

preponderance of the evidence—a “less demanding” standard of proof “than the clear and

convincing evidence standard” employed in Grell II and “required for a pre-trial finding,”

pursuant to § 13-753(G), “that mental retardation barred imposition of the death penalty.”

Id. Finally, the court observed that Grell had presented “substantially more—and more

convincing—evidence of adaptive skill deficits” at his 2009 resentencing hearing than he

had presented at the 2005 hearing held on the sole issue of whether he was mentally

retarded and therefore ineligible for a death sentence. Id. ¶ 11.

¶14 Like the court in Grell II, our review is limited to whether the respondent

judge “clearly err[ed]” in concluding Williams failed to prove an intellectual disability by

clear and convincing evidence, and we must defer to the respondent’s determination if

“reasonably supported by evidence.” Grell II, 212 Ariz. 516, ¶ 63, 135 P.3d at 709; see

also Book Cellar, Inc. v. City of Phx., 139 Ariz. 332, 335, 678 P.2d 517, 520 (App. 1983)

(in review by special action, court does not weigh evidence but determines whether

sufficient evidence supported decision and whether trial court properly exercised

discretion).

¶15 Here, three experts testified at the Atkins evidentiary hearing:

Neuropsychologist James Sullivan testified for the defense on the issue of Williams’s

cognitive function, and, on the issue of Williams’s adaptive behavior, psychologist

Ricardo Weinstein testified for the defense and psychologist Sergio Martinez testified for

the state. Each of these experts also provided a written report summarizing his evaluation

10
and conclusions. Sullivan concluded “beyond a reasonable psychological certainty” that

Williams presently “has pronounced and authentic neuropsychological and psychiatric

impairment.” He offered no similarly confident conclusion regarding the etiology of

these impairments or whether Williams met other clinical or statutory criteria for an

intellectual disability, but found it “quite likely” that Williams’s “documented cerebral

dysfunction . . . is a direct result of prenatal insult,” such as his mother’s substance abuse

during pregnancy, “as well as exposure to severe childhood psychological trauma in the

form of physical and sexual abuse.” Weinstein concluded Williams “fulfills the

definitions of Intellectual Disabilities (Mental Retardation) contained in . . . § 13-753(K)

as well as the ones contained in the DSM-IV . . . and [promulgated by] the AAIDD.”

Martinez concluded Williams “does not meet the criteria” in § 13-753(K)(3), finding “no

evidence” to demonstrate that “he experienced significantly subaverage general

intellectual functioning along with significant impairment in adaptive functioning” before

the age of eighteen.

¶16 In the order under review, the respondent judge correctly identified the

“determinative issues” as “whether there is sufficient evidence that [Williams’s]

intellectual functioning was significantly sub-average and whether his ‘adaptive

behavior’ was significantly impaired, and if so, whether the evidence is clear and

convincing that he suffered from [these conditions] before the age of [eighteen].”

¶17 With respect to intellectual functioning, the respondent judge found

“[Williams’s] current intelligence quotient is below [seventy], ‘significantly sub-

average.’ However, his IQ likely was higher [twenty] years ago, at age [eighteen]. The

11
current level ‘absolutely,’ as Dr. Sullivan put it, is lower because of [Williams’s] drug

and alcohol abuse and worsening mental illness.” Similarly, the respondent found,

“[Williams] does present now with ‘significantly impaired function’ as shown by Dr.

Sullivan’s recent neuropsychological testing,” but “the evidence does not support a

finding that, as the governing statute mandates, there is ‘clear and convincing evidence’

that [Williams’s] adaptive behavior was affected” by this intellectual deficiency, either

before or after the age of eighteen.

¶18 To establish an intellectual disability under § 13-753, Williams was

required to show by clear and convincing evidence that he suffered “the onset of an IQ

[of seventy or] below . . . before the age of [eighteen].” Moormann v. Schriro, 672 F.3d

644, 648-49 (9th Cir. 2012) (Arizona Supreme Court’s denial of stay of execution, where

defendant’s offer of proof pertained to declining IQ as adult post-conviction, not contrary

to or unreasonable application of clearly established federal law); see also State v.

Arellano, 213 Ariz. 474, ¶ 21, 143 P.3d 1015, 1021 (2006) (“The provision [in § 13-753]

requiring that symptoms of mental retardation occur before age eighteen applies to both

elements of mental retardation: significantly subaverage intelligence and significantly

impaired adaptive behavior.”). Sullivan reported Williams’s full scale IQ currently is

seventy, “consistent with borderline overall function” and on the cusp of Arizona’s

definition of the “[s]ignificantly subaverage general intellectual functioning” required to

establish an intellectual disability pursuant to § 13-753(K)(3) and (5).

¶19 Although Sullivan stated Williams’s reported substance abuse and mental

illness as an adult, considered either alone or in combination, could not account for the

12
full extent of impairments Sullivan found during neuropsychological tests, he also

testified there was “no question that these two things combined together [have] lessen[ed]

his over all [sic] level of function” over time, to an extent Sullivan could not quantify.7

Similarly, Sullivan reported Williams’s performance on recent achievement tests is “quite

likely . . . lower than [his] level of function when he was able to graduate from [high

school] and attend college,” with “[t]he source of this decline . . . most probably a

combination of ongoing substance abuse and worsening mental illness.” And, when

asked directly about the effect of Williams’s “extensive drug and alcohol abuse” on his

current test scores, Sullivan testified Williams’s “testing scores [absolutely would] have

been better than they are right now” had he “never pick[ed] up a drug or a drink.” Thus,

Sullivan, an expert in assessing cognitive functioning, disagreed with Weinstein’s

opinion that Williams’s use of drugs and alcohol as an adult “would not have

significantly reduced his overall cognitive abilities” as they existed before the age of

eighteen.

¶20 As the respondent judge observed, Sullivan’s opinion that Williams’s test

scores would have been higher, by some incalculable amount, when he graduated from

high school supports a conclusion that Williams’s full scale IQ was higher than seventy at

the age of eighteen, and only declined to its current level during Williams’s adult years,

7
When asked if he could quantify the relative effects of substance abuse and
mental illness in accounting for Williams’s current level of intellectual functioning,
Sullivan said he could not, because there were “too many [unknown] variables.” He
added that “the history and the issues of the sexual abuse, the physical abuse, [and] the
potential in utero insults from the mother” would likely “carry more of the variance” with
respect to the impairments, due to the greater “plasticity” of the brain during childhood.
13
due to his “drug and alcohol abuse and worsening mental illness.” Based on this

evidence alone, the respondent reasonably could have found the evidence less than clear

and convincing—and thus insufficient—to establish that Williams suffered from

significantly subaverage general intellectual functioning with an onset that occurred

“before [he] reached the age of eighteen.” § 13-753(K)(3).

¶21 The respondent judge’s determination finds further support in Martinez’s

opinions that Williams “was able to perform at higher levels of adaptive and intellectual

functioning in adolescence and adulthood” when he abstained from alcohol and drugs and

that “no evidence” demonstrates Williams had the intellectual and adaptive behavior

deficits required by § 13-753 “prior to the age of eighteen.” The absence of clear and

convincing evidence establishing that Williams had an IQ of seventy or below before the

age of eighteen provided a sufficient basis to deny Williams’s pre-trial claim under § 13-

753(G). See Moormann, 672 F.3d at 648-49; Arellano, 213 Ariz. 474, ¶ 21, 143 P.3d at

1020.

¶22 Williams challenges the respondent judge’s rejection of the “retrospective

analysis” performed by Weinstein, whose opinion is based in part on interviews he

conducted with seventeen people who knew Williams before and after the age of

eighteen, including family and extended family members, former classmates and friends,

work supervisors, and a high school administrator. In addition to reporting their

memories of Williams at different ages, Williams’s mother told Weinstein she had used

alcohol and drugs and had tripped on some stairs while pregnant with Williams, and other

14
interviewees reported Williams had landed on a hard floor when dropped as an infant and

had been physically and sexually abused as a child.

¶23 The respondent judge found, “Dr. Weinstein’s . . . methodology . . . is not

persuasive . . . . In particular, Dr. Weinstein’s ‘history’ is suspect. It is largely supported

by biased reports. No independent or verifiable evidence supports it.” According to

Williams, “[T]o find that such an analysis cannot meet a clear and convincing standard is

tantamount to denying mentally retarded individuals who are substantially older than

[eighteen]” the protection afforded by Atkins, “because [their intellectual disability]

would be impossible to prove.”8

¶24 But a trial court is required to evaluate the weight and credibility of the

evidence produced to support a finding of mental retardation. Grell II, 212 Ariz. 516,

¶ 58, 135 P.3d at 708. And, on appeal, a finding of fact cannot be clearly erroneous if

substantial evidence supports it, even though substantial conflicting evidence also exists.

State v. Berryman, 178 Ariz. 617, 623, 875 P.2d 850, 856 (App. 1994), citing Moore v.

Title Ins. Co. of Minn., 148 Ariz. 408, 413, 714 P.2d 1303, 1308 (App. 1985). “We do

not reweigh the evidence to decide if we would reach the same conclusions as the trier of

fact.” State v. Carey, 186 Ariz. 121, 124, 920 P.2d 1, 4 (App. 1995).

¶25 Williams’s argument essentially begs the question our supreme court

answered in Grell II, when it concluded that “requiring the defendant to prove mental

retardation by clear and convincing evidence in the initial retardation hearing does not

violate constitutional standards.” Grell II, 212 Ariz. 516, ¶ 41, 135 P.3d at 705. In a case

8
Williams presently is thirty-nine years old.
15
such as this, where a defendant’s intellectual and adaptive functioning were not tested

when he was a child, or where such test results are unavailable, a trial court must consider

whether inferences from other evidence establish that the onset of concurrent deficits in

these areas, sufficient to meet statutory requirements, occurred before the defendant

reached the age of eighteen. See Arellano, 213 Ariz. 474, ¶ 21, 143 P.3d at 1020-21

(noting “pre-age-eighteen IQ test results are not always available”). That appears to be

what the respondent judge did in this case. Nothing in the record suggests the respondent

categorically excluded evidence that might be offered in support of an older defendant’s

Atkins claim; to the contrary, in his order, the respondent painstakingly summarized the

evidence presented and set forth the reasons he found certain evidence more or less

persuasive.

¶26 As the respondent judge pointed out, both Weinstein and Sullivan identified

problems with the reliability of retrospective analyses. Moreover, Sullivan explained that

he regarded in utero insult and childhood abuse as “risk factors,” rather than “causative

factors” responsible for Williams’s present condition, because he had no documentation

to corroborate reports that Williams’s mother had used drugs and alcohol while pregnant

or that Williams had been abused as a child. Sullivan testified, “[I]f we . . . had birth

records that show he was born a crack baby and we had his mother laying it out on the

line being honest about what actually happened then[, when Williams was born,] . . . they

would become causative factors, right now I just don’t know.” In light of Sullivan’s

doubts about the veracity of these reports, we cannot say the respondent clearly erred in

finding the history of sexual and physical abuse and in utero insults, obtained from

16
Williams and his family, less than clear and convincing evidence that Williams suffered

the impairments required by § 13-753(K) before the age of eighteen.

¶27 Williams nonetheless challenges—as “illogical and contrary to the

evidence”—the respondent judge’s conclusion that Williams failed to show the onset of

the required impairments occurred before the age of eighteen, and he asserts “the only

plausible explanation for [Williams’s] condition is pre-natal insult combined with

childhood physical and sexual abuse.” Specifically, Williams cites the testimony of

Sullivan and Weinstein to support the proposition that “unless there is a credible

explanation for how a person could become functionally retarded after age [eighteen], . . .

it is assumed that the condition occurred prior to” that age.

¶28 Williams is referring to Sullivan’s testimony that “neuropsychological

impairments are presumed to be neurodevelopmental in nature” and occurring “early on

in the life span until proven otherwise” by an adequate medical explanation for brain

damage occurring as an adult, such as a traumatic brain injury, stroke, or certain kinds of

substance abuse. But Sullivan also testified Williams’s substance abuse was a

“contributing factor” to his intellectual impairment and accounted for some portion of the

deficits currently reflected in his full-scale IQ test score of seventy; thus, as already

discussed, according to Sullivan, Williams’s IQ test scores would have been higher than

seventy in late adolescence and early adulthood. Similarly, Martinez concluded

Williams’s current full-scale IQ score of seventy did not evince an onset of that condition

prior to the age of eighteen, particularly where there was “no indication” Williams

routinely had “performed in the significantly impaired range of intellectual functioning”

17
while in public school. These expert opinions constitute reasonable evidence that

supports the respondent judge’s findings, even if other, contradictory evidence may be

found in the record. See State v. McCurdy, 216 Ariz. 567, 573, 169 P.3d 931, 937 (App.

2007) (“substantial evidence” supports finding “[i]f reasonable persons could differ as to

whether the evidence establishes a fact in issue”); State v. Mercer, 13 Ariz. App. 1, 2, 473

P.2d 803, 804 (1970) (conflicts in testimony do not render evidence insufficient); see also

State v. Cañez, 202 Ariz. 133, ¶ 114, 42 P.3d 564, 594 (2002) (according “great

deference” to trial court’s resolution of conflicting psychological evidence offered by

experts).

¶29 Quoting Nicholson v. Branker, 739 F. Supp. 2d 839, 857 (E.D.N.C. 2010),

Williams maintains there is “no requirement that [a defendant] show he had scored

[seventy] or below on a[n IQ] test given prior to the age of [eighteen]” to establish his

mental deficiency “manifested before [eighteen] years of age.” We do not suggest

otherwise. See Arellano, 213 Ariz. 474, ¶ 21, 143 P.3d at 1020-21. We agree with those

authorities concluding that, when no childhood IQ tests were performed, subaverage

intellectual functioning before the age of eighteen properly may be inferred from other

evidence of intellectual functioning, such as school performance. See, e.g., Rivera v.

Quarterman, 505 F.3d 349, 363 (5th Cir. 2007). But the respondent judge was not

required to infer that Williams had suffered, as a child, from the level of intellectual

impairment required to establish an intellectual disability under Arizona law, particularly

in light of Sullivan’s and Martinez’s opinions to the contrary. See Pizzuto v. State, 202

P.3d 642, 651 (Idaho 2008) (trial court considering Atkins claim not required to infer

18
defendant’s IQ “had not decreased during the eleven-year period from his eighteenth

birthday to the date of his IQ test[,] . . . especially in light of the opinions of [his own]

experts that his long history of drug abuse and his epilepsy would have negatively

impacted his mental functioning”); Commonwealth v. Vandivner, 962 A.2d 1170, 1186-

87 (Pa. 2009) (insufficient evidence of onset before age eighteen where expert testified

head injuries and other conditions occurring in adulthood “led to a decline in cognitive

capacity”; refusing to extend categorical prohibition against execution of mentally

retarded to other capital defendants with mental deficiencies absent evidence of “national

consensus” found in Atkins).

¶30 Because we conclude the respondent judge did not err in finding Williams

failed to establish by clear and convincing evidence he had significantly subaverage

general intellectual functioning before the age of eighteen, we need not review his ruling

that “[i]t certainly is not ‘clear’” that Williams suffered from significant impairments in

adaptive behavior, either before or after the age of eighteen. See State v. Snelling, 225

Ariz. 182, n.8, 236 P.3d 409, 417 n.8 (2010) (when one issue dispositive, court need not

reach other issues). However, we will address briefly the issues Williams raises, because

many of his arguments are related to this determination.

¶31 In finding the evidence of adaptive impairments less than clear and

convincing, the respondent judge noted that Williams “went to school, traveled by bus,

stayed out of trouble, graduated from school on time, went to college, and was employed,

at one job for several years.” The respondent appears to have relied on Martinez’s

opinion that any difficulties Williams’s experienced in adaptive behavior did not rise to

19
the level of the “significant impairment” required to support a finding of intellectual

disability under Arizona law. See § 13-753(K)(3).

¶32 Williams first seems to suggest the respondent judge abused his discretion

in relying on Martinez’s report and testimony; he contends Martinez conducted only a

“cursory assessment” of Williams’s intellectual and adaptive functioning and is only

“marginally qualified” when compared to Sullivan and Weinstein. But Williams does not

contend Martinez was unqualified or his testimony was incompetent under the

requirements of § 13-753. Although Williams asserts Martinez had “no foundation” for

his opinion that, had Williams suffered from an intellectual disability, his deficits in

intellectual functioning and adaptive behavior would have been identified while he was in

public school, Martinez testified he had once been employed as a school psychologist.

Williams is correct that Martinez professed no knowledge of the policies employed by

the particular public schools Williams had attended, but although this limitation might

affect the weight of the evidence, it does not render it inadmissible. See State v. Davolt,

207 Ariz. 191, ¶ 70, 84 P.3d 456, 475 (2004) (degree of expert’s qualifications goes to

weight of testimony, not its admissibility).

¶33 Williams had a full opportunity to examine Martinez on his knowledge and

experience, and we see no clear error in the respondent judge’s consideration of this

testimony. Carey, 186 Ariz. at 124, 920 P.2d at 4 (reviewing court does not reweigh

expert testimony); State v. Wassenaar, 215 Ariz. 565, ¶ 2, 161 P.3d 608, 612 (App. 2007)

(reviewing court resolves conflicts in evidence in favor of sustaining judgment).

20
¶34 Citing AAIDD assessment guidelines, Williams also asserts Martinez and

the respondent judge “relied upon [Williams’s] adaptive strengths, as opposed to [his]

deficits, which is not a proper consideration in an adaptive behavior analysis,” because

“mentally retarded people are capable of many adaptive functions.” But in Grell II, our

supreme court emphasized that Arizona’s statutory definition, “while similar in overall

meaning, is not the same as” clinical definitions, and “requires an overall assessment of

the defendant’s ability to meet society’s expectations of him,” not “proof of specific

deficits.” Grell II, 212 Ariz. 516, ¶ 62, 135 P.3d at 709. As in this case, the decision

under review in Grell II “was based largely on expert testimony,” and “the trial court

determined that the State’s expert was more credible.” Id. ¶ 58. Deferring to the court’s

broad discretion in determining the weight and credibility afforded expert testimony, our

supreme court affirmed the court’s ruling. Id. ¶ 63.

¶35 The respondent judge’s reasoning here appears consistent with the ruling

affirmed in Grell II, and we decline to adopt a standard for analysis inconsistent with that

decision or the deference it requires. See id. ¶¶ 61-62 (state countered evidence of Grell’s

poor academic and social behavior “with three main themes: no doctor before defense

expert . . . had ever diagnosed Grell as having mental retardation; behaving badly does

not necessarily indicate adaptive deficits; and Grell can behave himself when he wants to

do so”); Grand v. Nacchio, 214 Ariz. 9, ¶ 19, 147 P.3d 763, 771 (App. 2006) (declining

to apply doctrines that might “effectively nullify” test adopted by supreme court); see

also Hooks v. Workman, 689 F.3d 1148, 1172 (10th Cir. 2012) (AAIDD’s “clinical

standard” of considering deficits rather than strengths “not a constitutional command”

21
nor required by Atkins); Ortiz v. United States, 664 F.3d 1151, 1168-69 (8th Cir. 2011)

(Atkins did not “delegate[] to the scientific community the finding of whether an

individual is mentally retarded”; rejecting standard limiting assessment to adaptive

deficits because “[c]onsideration of . . . strengths” may “provide context and definition

for consideration of reported deficits”); Clark v. Quarterman, 457 F.3d 441, 445 (5th Cir.

2006) (Atkins “did not dictate that the approach and the analysis of the State inquiry must

track the approach of the AAMR or the APA exactly”).

¶36 In other arguments that the respondent judge erred in finding insufficient

evidence of adaptive behavior impairment, Williams challenges the manner in which the

respondent weighed the evidence and the inferences he drew from undisputed facts. But

in our review of the respondent’s ruling pursuant to § 13-753, we defer to his resolution

of conflicting evidence, and we do not substitute our independent judgment for his. Grell

III, 231 Ariz. 153, ¶¶ 9-10, 291 P.3d at 353; Grell II, 212 Ariz. 516, ¶ 63, 135 P.3d at

709; Gen. Elec. Capital Corp. v. Osterkamp, 172 Ariz. 185, 188, 836 P.2d 398, 401 (App.

1992) (reviewing court will not “second-guess or substitute [its] judgment” for trial

court’s resolution of disputed fact).

The Dissent

¶37 As an initial matter, it appears our dissenting colleague has reached an issue

that Williams did not raise below, a practice we ordinarily avoid. See, e.g., Cornerstone

Hosp. of Se. Ariz., L.L.C. v. Marner, 231 Ariz. 67, n.4, 290 P.3d 460, 472 n.4 (App.

2012). The limited record before us contains no indication that Williams had argued the

definition of intellectual disability in § 13-753 is unconstitutional or that the respondent

22
judge considered such a claim. Although the respondent referred informally to “an Atkins

hearing” in his order, he clearly conducted the hearing and entered his ruling in

accordance with § 13-753, applying that statute’s definitions in the proceeding it

mandated. See § 13-753(G), (K). The dissent maintains the respondent erred precisely

because he ruled in “conformity with the statutory standard,” when he instead should

have rejected the statutory definition in favor of a “clinical definition” promulgated by

the APA or AAIDD. See infra ¶¶ 53, 61-62.

¶38 First, as addressed above, we conclude our supreme court considered and

rejected a nearly identical argument in Grell II, 212 Ariz. 516, ¶¶ 60-63, 135 P.3d at 709.

According to the dissent, the court in Grell II did not expressly consider whether the

definition of intellectual disability in § 13-753 is consistent with the holding in Atkins,

and this remains an open issue. See infra ¶¶ 48, 78-79. But the fact remains that in Grell

II, after resolving other “issues of constitutional law and statutory construction” and

recognizing state procedures to identify mentally retarded capital defendants “must

comport with the Constitution,” 212 Ariz. 516, ¶¶ 22, 24, 135 P.3d at 701, 702, the court

rejected the defendant’s argument that he had established mental retardation pursuant to

the DSM-IV definition, explaining that, “while similar in overall meaning, [that standard]

is not the same as the statutory definition.” Id. ¶ 62. The court then relied on this

distinction in finding, based on an “overall assessment” of adaptive functioning permitted

by Arizona’s statute, substantial evidence supported the trial court’s ruling that Grell had

failed to sustain his burden of proving mental retardation. Id. ¶¶ 62-63. When later

conducting its own independent review of Grell’s death sentence, the court again relied

23
on the statutory definition to guide its analysis, albeit with a different result. Grell III,

231 Ariz. 153, ¶¶ 5, 7, 15, 34, 291 P.3d at 351-52, 353, 356-57; see also State v. Boyston,

231 Ariz. 539, ¶¶ 34-36, 298 P.3d 887, (2013) (“[a]lthough the DSM–IV defines

impairments in adaptive functioning based on deficits in two areas,” that definition “is

not the same as the statutory definition,” which, “by contrast, ‘requires an overall

assessment of the defendant’s ability to meet society’s expectations of him’”), quoting

Grell II, 212 Ariz. 518, ¶ 62, 135 P.3d at 709.

¶39 Thus, Arizona’s supreme court appears to have interpreted Atkins as

“leaving the definition of mental retardation to the states,” such that the prohibition

against executing the mentally retarded “depends on the state’s definition of mental

retardation.” State v. Roque, 213 Ariz. 193, ¶ 149, 141 P.3d 368, 402 (2006); Grell II,

212 Ariz. 516, ¶ 37, 135 P.3d at 705 (knowing that “states had already begun to develop

their own procedures, and had drawn in different places the line for establishing the

mental retardation that would bar execution,” Supreme Court “left states . . . free to craft

their laws for determining which defendants meet the [national] consensus standard”).

¶40 Second, although the dissent asserts “the Court [in Atkins] defined mental

retardation by the diagnostic criteria universally accepted by clinicians,” infra ¶ 51, this

position is contrary to pronouncements from the United States Supreme Court and every

other state court that has addressed the issue. The Supreme Court itself has stated that

Atkins “did not provide definitive procedural or substantive guides for determining when

a person who claims mental retardation ‘will be so impaired as to fall [within Atkins’

compass]’” but left enforcement of the constitutional restriction to the states. Bobby v.

24
Bies, 556 U.S. 825, 831 (2009), quoting Atkins, 536 U.S. at 317 (emphasis added);

Moormann v. Schriro, 672 F.3d 644, 648 (9th Cir. 2012) (Atkins “did not define mental

retardation as a matter of federal law”); see also Julie C. Duvall & Richard J. Morris,

Assessing Mental Retardation in Death Penalty Cases: Critical Issues for Psychology

and Psychological Practice, 37 Prof. Psychol.: Res. & Prac. 658, 658 (2006) (Atkins “did

not define mental retardation” or identify “any terms or procedures that could guide

legislatures or judges in determining” those defendants “‘so impaired as to fall within the

range of mentally retarded offenders about whom there is a national consensus’”),

quoting Atkins, 536 U.S. at 317. Additionally, every other state court that has addressed

the issue has determined or implied that Atkins allows the states to define mental

retardation without strict adherence to the clinical standards. Of the thirty-three states

that still permit use of the death penalty, courts in twenty-three have stated or implied that

Atkins did not define mental retardation, but instead left that task to individual states. See

infra Appendix 1. The other ten states have not spoken on the issue. Many state statutes

also define mental retardation or intellectual disability in ways that vary from the AAMR,

AAIDD, and DSM-IV definitions. See infra Appendix 2. As noted above, several courts

have expressly rejected the proposition that states must strictly adhere to clinical

definitions or assessment standards. See supra ¶ 35.

¶41 Third, Arizona’s statute was among those enacted before Atkins was

decided and was found to be evidence of the “national consensus” that mentally retarded

offenders should not be executed. See Atkins, 536 U.S. at 314-15. The dissent

acknowledges the Supreme Court’s statement that existing “statutory definitions of

25
mental retardation are not identical, but generally conform to the clinical definitions [of

the AAMR and APA],” Atkins, 536 U.S. at 317 n.22, and maintains this observation

served only “to reinforce the conclusion that a true national consensus had coalesced

around” those clinical definitions. See infra ¶ 74. But the placement of this footnote

suggests otherwise. It appears at the end of a paragraph in which the Court addressed

“disagreement . . . in determining which offenders are in fact retarded” and immediately

after the Court’s statement that it would leave to the states “the task of developing

appropriate ways to enforce the constitutional restriction.” Atkins, 536 U.S. at 317.

Thus, in context, the Supreme Court signaled that those states that had already enacted

statutes prohibiting execution of the mentally retarded, including Arizona, had met this

task by adopting definitions of mental retardation that “generally conform” to clinical

standards. Id. at 317 & n.22.

¶42 Fourth, we agree with the state that, as a policy matter, requiring strict

adherence to clinical standards could create some instability in this area of the law. As

we have seen, clinical standards may change over time, see supra note 4; it seems

unlikely the Supreme Court would have delegated the interpretation of the Eighth

Amendment to clinicians. Rather, deference to state legislation that generally conforms

to clinical standards is consistent with the Court’s recognition that “‘[c]ourts are not

representative bodies’” and do not act “as legislators,” and that, “under our federal

system,” the deference owed “state legislatures . . . is enhanced where the specification of

punishments is concerned, for ‘these are peculiarly questions of legislative policy.’”

Gregg v. Georgia, 428 U.S. 153, 174-76 (1976), quoting Dennis v. United States, 341

26
U.S. 494, 525 (1951) (Frankfurter, J., concurring) (alteration added) and Gore v. United

States, 357 U.S. 386, 393 (1958); see also Panetti v. Quarterman, 551 U.S. 930, 957

(2007) (Supreme Court has yet to “set forth a precise standard for competency” to be

executed under Ford v. Wainwright); Addington v. Texas, 441 U.S. 418, 431 (1979)

(“The essence of federalism is that states must be free to develop a variety of solutions to

problems and not be forced into a common, uniform mold”; noting substantive and

procedural standards for civil commitment “may vary from state to state, . . . so long as

they meet the constitutional minimum”).

¶43 Finally, we note that our dissenting colleague focuses on the respondent

judge’s analysis of Williams’s adaptive functioning and does not address his

determination that Williams also failed to present clear and convincing evidence of a

substantial impairment in intellectual functioning before the age of eighteen. Nor does

the dissent address our conclusion that this finding was consistent with the evidence and

independently sufficient to support the respondent’s ruling. See supra ¶¶ 29-30; see also

Pruitt v. State, 834 N.E.2d 90, 110 (Ind. 2005) (affirming death penalty, notwithstanding

court’s erroneous application of “too restrictive” standard to determine “substantial

impairment of adaptive behavior,” where record supported conclusion defendant failed to

prove significantly subaverage intellectual functioning).

¶44 Accordingly, we conclude we are bound by our supreme court’s

pronouncements on this issue. The conclusion that states may enforce the prohibition in

Atkins by implementing definitions that “generally conform” to clinical standards, 536

U.S. at 317 n.22, as Arizona’s statutory definition does, id. at 314-15, 317 n.22, finds

27
further support in a later statement by the United States Supreme Court and in each state

court case that has addressed the issue. Therefore, we cannot accept the dissent’s

position.

¶45 We do not mean, as the dissent suggests, that the Supreme Court “intended

to . . . preclude any future conceivable specific Eighth Amendment challenge” to statutes

enacted before Atkins, see infra note 22; rather, the Court has stated clearly that as-

applied challenges might be expected. See Schriro v. Smith, 546 U.S. 6, 7-8 (2005) (per

curiam) (noting states’ “measures for adjudicating claims of mental retardation” by

capital defendants “might, in their application, be subject to constitutional challenge”;

Ninth Circuit exceeded authority in “pre-emptively” ordering jury determination). But in

light of our own supreme court’s repeated statements that the statute is based on and

generally conforms to clinical standards, as well as its rejection of the nearly identical

position taken by the defendant in Grell II, we must leave any changes in that position to

the supreme court. See Roque, 213 Ariz. 193, ¶ 149, 141 P.3d at 402 (Arizona statute

based on “accepted medical definitions”); Grell II, 212 Ariz. 516, n.4, 135 P.3d at 699

n.4 (statutory definition “substantially consistent” with criteria cited in Atkins); Grell I,

205 Ariz. 57, n.4, 66 P.3d at 1241 n.4 (Arizona’s statute “appears to comport

substantively and procedurally with the principles set forth in Atkins”); State v. Cañez,

205 Ariz. 620, n.2, 74 P.3d 932, 938 n.2 (2003) (Arizona’s statutory definition of mental

retardation “very similar to that set forth in the DSM-IV”); see also Ybarra v. State, 247

P.3d 269, 274-75 (Nev. 2011) (adaptive functioning “‘refers to how effectively

individuals cope with common life demands and how well they meet the standards of

28
personal independence expected of someone in their particular age group, sociocultural

background, and community setting’”), quoting DSM-IV 42; State v. Long, 207 Ariz.

140, ¶ 23, 83 P.3d 618, 623 (App. 2004) (“This court is bound by decisions of the

Arizona Supreme Court and has no authority to overturn or refuse to follow its

decisions.”).

Disposition

¶46 As our supreme court has observed, “[r]easonable minds may differ as to

how to interpret the evidence presented,” but this does not provide a basis for relief from

the respondent judge’s ruling. Grell II, 212 Ariz. 516, ¶ 63, 135 P.3d at 709. On this

record, we cannot say the respondent clearly erred in finding Williams failed to prove by

clear and convincing evidence that he is mentally retarded or has an intellectual

disability. See § 13-753(G). Accordingly, we deny relief.

/s/ Joseph W. Howard
JOSEPH W. HOWARD, Chief Judge

CONCURRING:

/s/ J. William Brammer, Jr.
J. WILLIAM BRAMMER, JR., Judge*

*A retired judge of the Arizona Court of Appeals authorized and assigned to sit as a
judge on the Court of Appeals, Division Two, pursuant to Arizona Supreme Court Order
filed December 12, 2012.

29
E C K E R S T R O M, Presiding Judge, dissenting.

¶47 In Atkins v. Virginia, the Court set forth a “substantive restriction” on the

states, prohibiting the execution of the “mentally retarded.” 536 U.S. at 321. In so doing,

the Court made clear, both expressly and implicitly, that the category of those protected

by its holding would be those classified as mentally retarded by clinical standards. 9

Although that holding also provides the respective states some substantive and procedural

leeway in enforcing this constitutional restriction, no language in Atkins invites states to

conjure their own definitions of the condition independent of clinical standards. Indeed,

both the state and majority appear to concede that substantive statutory definitions of

mental retardation must, at minimum, generally conform to the prevailing clinical

definition of the condition to properly enforce the Eighth Amendment restriction on the

execution of the mentally retarded.

¶48 Our own supreme court has observed that Arizona’s statutory standard for

evaluating impairment of adaptive functioning deviates from the clinical standard.

Grell II, 212 Ariz. 516, ¶ 62, 135 P.3d at 709. It did so in the context of addressing an

evidentiary claim raised and addressed exclusively under state law. See id. ¶¶ 58-63.

Our highest court has neither considered, nor purported to address, the considerably more

complex question of federal constitutional law raised by this case: whether, in spite of

9
As the majority correctly points out, “intellectual disability” is now the preferred
term for “mental retardation.” For the sake of analytical clarity as to the specific issue
presented here, I have elected to use the term employed in Atkins.

30
our statute’s varying standard for evaluating adaptive functioning, that section of our

statute substantially conforms to clinical diagnostic criteria and, therefore, Eighth

Amendment standards.10

¶49 As shall be demonstrated below, Arizona’s standard for evaluating

impairment in adaptive functioning departs, in practical application, so markedly from

the parallel clinical standard that it precludes the diagnosis of mental retardation for most

of those who would be so defined by clinical standards. For that reason, I can only

conclude that Arizona’s standard for determining mental retardation fails to enforce the

constitutional restriction set forth in Atkins, and therefore runs afoul of the Eighth

Amendment to the United States Constitution.

In prohibiting the execution of the mentally retarded, Atkins employed the clinical
definition of that condition.

¶50 In Atkins, the Court determined that “mentally retarded” offenders are

“categorically excluded from execution” by the Eighth Amendment. 536 U.S. at 318;

accord id. at 320-21. Because the exclusion is categorical, all those who fall within the

Court’s definition of mental retardation are exempted from capital punishment.

10
Although I agree with the majority that the defendant failed to squarely raise this
claim in his initial petition to this court, he did maintain generally that the trial court’s
consideration of the evidence violated the Eighth Amendment requirements set forth in
Atkins. He thereafter more specifically complained that the respondent judge’s
evaluation of the evidence of adaptive functioning departed from clinical standards.
Moreover, both parties were provided an opportunity to address the specific question we
address here in supplemental briefing. At any rate, we do not ignore fundamental error
when we find it in the record. State v. Musgrove, 223 Ariz. 164, ¶ 4, 221 P.3d 43, 45
(App. 2009). Because application of the correct definition of mental retardation goes to
the core of this proceeding designed to enforce the Eighth Amendment restriction
described in Atkins, the constitutional error here is fundamental and prejudicial.
31
¶51 There can be little doubt that the Court defined mental retardation by the

diagnostic criteria universally accepted by clinicians. In framing the issue presented by

Atkins’s case, the Court noted, and comprehensively recited, the definitions for mental

retardation adopted, respectively, by the American Psychiatric Association (APA) and the

American Association on Mental Retardation (AAMR) (now the American Association

on Intellectual and Developmental Disabilities (AAIDD)). Id. at 308 n.3. Thereafter, the

Court expressly anchored its reasoning in the clinical diagnostic criteria for that condition

and clinical understandings of the capacities of persons so diagnosed. See id. at 317-21.

In that reasoning, the Court itemized the disabilities that the “clinical definitions of

mental retardation require,” it specified the effects of those deficits on criminal

culpability, and it concluded that those deficiencies justify categorical exclusion of the

mentally retarded from eligibility for the death penalty. Id. at 318-21.

¶52 Thus, the United States Supreme Court has defined the class of mentally

retarded persons ineligible for capital punishment with express reference to the

deficiencies displayed by those diagnosed with the condition under clinical criteria.

Those criteria, as specifically articulated by both the AAMR and the APA are:

(1) significantly subaverage intellectual functioning; (2) significant limitations in

adaptive functioning in two or more specific skill areas, namely “communication, self-

care, home living, social/interpersonal skills, use of community resources, self-direction,

functional academic skills, work, leisure, health, and safety”; and (3) onset before age

eighteen. Id. at 308 n.3, quoting DSM-IV at 41. As legal commentators have observed,

“under Atkins v. Virginia, the Eighth Amendment protects individuals who meet the

32
AAIDD/AAMR criteria for mental retardation or the virtually identical criteria of the

DSM-IV-TR.” John H. Blume et al., Of Atkins and Men: Deviations from Clinical

Definitions of Mental Retardation in Death Penalty Cases, 18 Cornell J.L. & Pub. Pol’y

689, 691 (2009).

Arizona’s statute deviates substantially from the clinical definition.

¶53 Here, the trial court did not apply the clinical definition of mental

retardation in concluding that Williams was not intellectually disabled. Rather, the court

applied Arizona’s statutory definition of mental retardation set forth in § 13-753(K)(3).

As the court acknowledged in its order, that definition deviates from the clinical standard

in its criteria for determining whether an offender suffers from a significant impairment

in adaptive functioning. In Grell II, the Arizona Supreme Court described this variance

when addressing a state law challenge to a trial court’s findings:

The defense claims to have clearly shown . . . deficits
in two of the eleven areas listed in the DSM-IV and therefore
[the defendant] has mental retardation. The DSM-IV
definition of mental retardation, however, while similar in
overall meaning, is not the same as the statutory definition.
The statute requires an overall assessment of the defendant’s
ability to meet society’s expectations of him. It does not
require a finding of mental retardation based solely on proof
of specific deficits or deficits in only two areas.

212 Ariz. 516, ¶ 62, 135 P.3d at 709 (citation omitted).11

11
As the majority indicates, the former A.R.S. § 13-703.02(K) discussed in Grell II
was renumbered § 13-753(K), 2008 Ariz. Sess. Laws, ch. 301, § 26, and then amended to
substitute the term “intellectual disability” for the term “mental retardation.” 2011 Ariz.
Sess. Laws, ch. 89, § 5. The statute has remained the same in material part.

33
¶54 Specifically, Arizona’s statute requires a court to evaluate a defendant’s

“[a]daptive behavior” by a global assessment of “the effectiveness or degree to which the

defendant meets the standards of personal independence and social responsibility

expected of the defendant’s age and cultural group.” § 13-753(K)(1). By contrast,

clinical definitions of mental retardation, as adopted by Atkins, require a clinician to

consider seriatim how a defendant performs on each of an array of specific life skills.

See DSM-IV at 41 (itemizing eleven specific skill areas); AAMR, Mental Retardation:

Definition, Classification, and Systems of Supports 5 (9th ed. 1992) (itemizing ten). By

those clinical standards, individuals are diagnosed as mentally retarded if they show

significant limitations in a couple of those skill areas—even if they exhibit strengths in

most others. See DSM-IV at 41 (definition for mental retardation met if person has

limited functioning in two of eleven areas).12

¶55 For example, a defendant in Arizona must show he is significantly impaired

overall in the adaptive category of “social responsibility” to be classified as mentally

12
The AAIDD has since revised its diagnostic criteria for intellectual disability to
require significant limitations in “adaptive behavior,” which is defined as “the collection
of conceptual, social, and practical skills that have been learned and are performed by
people in their everyday lives.” AAIDD, Intellectual Disability: Definition,
Classification, and Systems of Supports 43 (11th ed. 2010) (hereinafter AAIDD Manual).
Behavior in each of those three categories is evaluated with reference to specific itemized
life skills which incorporate the eleven skill areas listed in the DSM-IV criteria. See
AAIDD Manual at 44. And deficits in all areas need not be established to classify an
individual as mentally retarded. Indeed, an individual is deemed to have significant
limitations in adaptive behavior if he or she performs on standardized measures
“approximately two standard deviations below the mean” in any of the three types of
adaptive behavior: conceptual, social, or practical. Id. at 43. Thus, the revised AAIDD
definition is not a meaningful departure from either the DSM-IV criteria or the AAMR’s
prior definitional standard set forth in Atkins.

34
retarded under our statutory definition. See § 13-753(K)(1), (3) (requiring showing of

significant impairment as to both “personal independence and social responsibility”). But

no such showing is required under clinical definitions of mental retardation. Indeed,

under the current AAIDD diagnostic standard, “social responsibility” is identified as but

one of eight nonexclusive sub-factors to consider in evaluating “social skills,” a category

which, in turn, comprises a nonexclusive criterion for evaluating adaptive behavior.

AAIDD, Intellectual Disability: Definition, Classification, and Systems of Supports 44

(11th ed. 2010) (hereinafter AAIDD Manual).

¶56 In essence, then, Arizona law precludes the diagnosis of mental retardation

for any individual who fails to show significant impairment in both broad categories of

“personal independence” and “social responsibility.” And the determination required by

statute involves an “overall assessment,” Grell II, 212 Ariz. 516, ¶ 62, 135 P.3d at 709,

that implicitly allows strengths in some adaptive skills to outweigh significant

impairments in others.

¶57 Clinical definitions of mental retardation include no parallel requirement

that a person with significantly subaverage intelligence also demonstrate significant

deficits in both “personal independence” and “social responsibility.” Indeed, a person

need not exhibit overall impairment in either broad category. Rather, evidence of

significant impairment in any two of many discrete skill areas is considered sufficient—

together with a finding of significantly subaverage intelligence with the onset before age

eighteen—to classify an individual as mentally retarded. And, by clinical standards,

35
individuals may receive such a classification regardless of how well they may perform in

other skill areas pertinent to adaptive functioning.

¶58 Not surprisingly, the diagnostic approach adopted by clinicians conforms to

clinical understandings of the behavior of most mentally retarded persons. As the

definitional manual published by the AAIDD instructs: “Individuals with an [intellectual

disability] typically demonstrate both strengths and limitations in adaptive behavior.

Thus, in the process of diagnosing [intellectual disability], significant limitations in

conceptual, social, or practical adaptive skills is not outweighed by the potential strengths

in some adaptive skills.” AAIDD Manual at 47 (emphasis added). According to the

DSM-IV, mildly mentally retarded persons “usually achieve social and vocational skills

adequate for minimum self-support” and, with appropriate supports, “usually live

successfully in the community, either independently or in supervised settings.” DSM-IV

at 43 (emphases added).13 Indeed, in this case, Williams presented a longitudinal study,

through Dr. Weinstein’s testimony, which found that thirty-six percent of those identified

as mentally retarded in high school later “lived independently,” and seventy-nine percent

13
The DSM-IV also observes that people with “mild mental retardation,” those
with an IQ of fifty-five to seventy, make up about eighty-five percent of those classified
as mentally retarded. DSM-IV at 42-43. Given that there is little practical difficulty in
diagnosing those with an IQ below fifty-five, the correct identification of those with
“mild mental retardation” is, for all practical purposes, the central task of an Atkins
hearing.

36
either became employed, received technical training for employment, or attended higher

education.14

¶59 In application, therefore, the difference between Arizona’s statutory

definition of mental retardation and the accepted clinical definition is neither semantic

nor trivial. Rather, Arizona’s statutory definition produces categorically different results

in identifying the class of persons diagnosed as mentally retarded. While both definitions

require evidence of significant impairment in adaptive behavior, Arizona requires that

such impairment be considered globally and be manifested as to personal independence

and social responsibility. But, by clinical standards, persons suffering from mild mental

retardation typically can demonstrate either personal independence or social

responsibility, and most possess notable strengths in some skill areas pertinent to those

general statutory categories. Thus, Arizona’s statutory requirements substantially narrow

the class of persons who are defined as mentally retarded when compared with the class

of those who would be clinically defined as such.

¶60 Although our legislature may well have drafted § 13-753(K) with the

intention of correctly defining mental retardation in conformity with clinical standards,15

and thereby complying with the holding of Atkins, the language of § 13-753(K) falls short

of achieving that goal. In my view, Arizona’s statutory definition cannot exclude the

14
Dr. Weinstein so testified with reference to an exhibit that was not made part of
the record. Accordingly, no citation to the study is available.
15
The text of § 13-753(E), which requires expert witnesses to use accepted
psychological testing procedures to evaluate a defendant for mental retardation, suggests
such a legislative intent.

37
“typical” person suffering from mental retardation and yet comply with Atkins’s

command that those suffering from mental retardation be categorically ineligible for

execution. See Pruitt v. State, 834 N.E.2d 90, 110 (Ind. 2005) (permissible variation in

defining adaptive functioning cannot constitutionally exclude “a majority of those who fit

clinical definitions”).

The improper standard affected the outcome of the hearing.

¶61 The record before us shows that the above flaws in Arizona’s statute

defining intellectual disability had a demonstrable effect on the trial court’s ultimate

conclusion that Williams does not suffer from mental retardation. In its exhaustive and

conscientious minute entry, the trial court specifically noted the variation between the

clinical standard for mental retardation and that set forth by § 13-753(K)(1), and the court

adopted the statutory one.

¶62 In conformity with the statutory standard, the trial court evaluated

Williams’s adaptive behavior “overall” by considering whether he displayed a “wide

variety of difficulties” in activities of daily living and whether he had difficulties living

“independently in the world.” In the court’s view, these inquiries essentially captured the

statutory definition for adaptive behavior. The court then expressly found Williams’s

display of one discrete adaptive skill—the ability to cook for a group of people—to be

dispositive evidence that he could live independently. According to the court, “[a]

finding that [Williams] was unable to live independently is not possible when he could

cook for ten people for a year.”

38
¶63 Given that the trial court understood Arizona’s statutory definition of

mental retardation to require a showing that the defendant could not live independently,

this finding essentially resolved the question of the defendant’s mental retardation under

§ 13-753(K)(1) and (3).16 But, as discussed above, no clinician would exclude the

diagnosis of mental retardation based on a person’s display of a single adaptive strength.

Indeed, as the above clinical definitions make clear, most mentally retarded individuals

possess some useful life and vocational skills, and most mildly mentally retarded persons

can be taught such skills.17 As commentators have noted, when state standards allow

fact-finders to place dispositive weight on individual adaptive strengths, those standards

encourage diagnostic conclusions anchored in fundamental misunderstandings of the

condition. Blume, supra, at 707-08. That appears to be what occurred in the case before

us.18

16
The trial court’s remaining analysis of Williams’s adaptive functioning
repeatedly dismissed evidence of adaptive impairment with reference to competing
evidence of adaptive strengths, an approach arguably compelled by Arizona’s “overall
assessment” standard. Grell II, 212 Ariz. 516, ¶ 62, 135 P.3d at 709.
17
Indeed, Williams’s supervisor stated she had to direct him how to properly shop
for those meals “until he learn[ed] how to do it.” Thus, Williams’s adult ability to cook
for a group was, in part, taught behavior. See DSM-IV at 43 (defining those with “mild
mental retardation” as “educable” and observing they usually achieve vocational skills).
Dr. Weinstein also directly testified that the ability to cook does not foreclose a finding of
mental retardation, noting that his own son, who suffers from Down syndrome, is a very
good cook.
18
As our supreme court implied in Grell III, 231 Ariz. 153, ¶¶ 5, 10, 291 P.3d at
351-52, Arizona’s procedural scheme necessarily requires a jury to independently
determine, during the mitigation phase of a capital trial, whether a defendant is mentally
retarded. Yet jurors are even less equipped than experienced trial judges to avoid lay
misunderstandings of mental retardation when faced with the provocative facts of a
39
¶64 The majority suggests that any such defect in the court’s analysis would be

harmless because the court also found that Williams had failed to present clear and

convincing evidence of onset before age eighteen. But here, where the trial court lacked

evidence of an IQ test conducted on Williams before age eighteen, evidence of his

adaptive functioning as a youth necessarily had heightened importance in the court’s

assessment of whether his mental retardation had a juvenile onset. Any conclusion that

Williams had demonstrated significant impairment in his adaptive functioning as a child

would have countered the testimony credited by the trial court that Williams’s

intelligence level was higher as a juvenile than as an adult.

Atkins cannot be construed as inviting states, without limitation, to redefine the
nature of the substantive constitutional restriction set forth in that case.

¶65 The state suggests we may overlook any statutory deviation from prevailing

clinical standards because the United States Supreme Court has, in Atkins, invited the

states to develop their own substantive definitions for mental retardation. It contends this

invitation occurred in the following portion of that opinion:

To the extent there is serious disagreement about the
execution of mentally retarded offenders, it is in determining
which offenders are in fact retarded. In this case, for
instance, the Commonwealth of Virginia disputes that Atkins
suffers from mental retardation. Not all people who claim to
be mentally retarded will be so impaired as to fall within the
range of mentally retarded offenders about whom there is a
national consensus. As was our approach in Ford v.
Wainwright, 477 U.S. 399, 106 S.Ct. 2595, 91 L.Ed.2d 335
(1986), with regard to insanity, “we leave to the State[s] the
task of developing appropriate ways to enforce the

capital case. Instructing those jurors with correct clinical standards thus becomes
especially important in that context.
40
 constitutional restriction upon [their] execution of sentences.”
Id., at 405, 416-417, 106 S.Ct. 2595.

Atkins, 536 U.S. at 317 (alterations in Atkins). But this paragraph, by its terms, only

invites states to develop “ways to enforce” the constitutional restriction imposed in

Atkins. No part of that language suggests the states are likewise entrusted with the power

to redefine the substance of the constitutional restriction itself. As Chief Justice

Rehnquist correctly observed in his dissent, the very point of the majority opinion was to

limit Virginia’s power, and that of its judges and sentencing juries, to determine who has

presented “an insufficiently compelling reason to lessen their individual responsibility for

the crime.” Id. at 321-22 (Rehnquist, C.J., dissenting). That intention can likewise be

found in the unambiguous words of the majority’s ultimate holding that “the Constitution

‘places a substantive restriction on the State’s power to take the life’ of a mentally

retarded offender.” Id. at 321, quoting Ford, 477 U.S. at 405 (emphasis added).

¶66 I would therefore decline the state’s suggestion that we construe a portion

of the Court’s reasoning in a fashion so at odds with the clear intent expressed in its

ultimate holding. Indeed, if courts accept the state’s interpretation of the language,

nothing would prevent individual states from requiring an IQ below fifty-five to

demonstrate mental retardation or erecting a presumption that a defendant’s ability to

commit first-degree murder itself demonstrates sufficient intelligence to qualify for

execution. Such potential legislative definitions of that condition would render the Atkins

41
holding irrelevant in practical effect and would eviscerate any meaningful “substantive

restriction on the State’s power” to take the life of a mentally retarded offender.19 Id.

¶67 Rather, the above language relied upon by the state is correctly construed as

reserving in the individual states the power to erect a procedural framework by which

those suffering from mental retardation are identified. Notably, in the pivotal sentence of

that paragraph, the Court invites states to “‘develop[] appropriate ways to enforce’”

constitutional restrictions “[a]s was our approach in Ford v. Wainwright.” Atkins, 536

U.S. at 317, quoting Ford, 477 U.S. at 416-17. To avoid any confusion, the Atkins Court

provided specific references to the portions of Ford that exemplified that approach. See

Atkins, 536 U.S. at 317, 321. In those portions of Ford, the Court exclusively analyzes

the procedural mechanisms by which Florida identified insane offenders exempt from

execution. 477 U.S. at 405, 416-17. Indeed, the specific quotation that the Atkins

majority lifts from Ford—the “develop[] . . . ways to enforce” language—was used in the

Ford opinion itself to refer to the development of such “procedural safeguards.” 477

U.S. at 416-17. Not surprisingly, therefore, our own supreme court has read that

paragraph as providing procedural discretion to the states. See Grell II, 212 Ariz. 516,

¶¶ 24, 26, 41, 135 P.3d at 701-02, 705 (interpreting Atkins as providing states authority

19
This concern is not far-fetched. As Justice Scalia notes in his dissent, Kansas’s
statutory definition of “mental retardation” does not include the mildly mentally retarded,
Atkins, 536 U.S. at 343 n.2 (Scalia, J., dissenting), a group which, under clinical
diagnostic criteria, makes up eighty-five percent of those classified as mentally retarded.
DSM-IV at 43. And, as demonstrated above, Arizona’s statutory definition of adaptive
functioning has the practical effect of excluding most of those clinically classified as
mildly mentally retarded.

42
“to develop appropriate procedures” to identify those who are mentally retarded,

specifically the power to set forth the burden and standard of proof for mental retardation

at pretrial hearing); accord People v. Vasquez, 84 P.3d 1019, 1022 (Colo. 2004).

¶68 The majority correctly observes that the supreme courts of both the United

States and Arizona have also understood Atkins as permitting states to set forth

substantive definitions of mental retardation that are not facsimiles of clinical standards.

See Bies, 556 U.S. at 831; Roque, 213 Ariz. 193, ¶ 149, 141 P.3d at 402. But neither

court has suggested that states are entitled to create wholly novel definitions of that

condition, independent of the accepted clinical understandings of mental retardation and

therefore unlimited by how Atkins employed that term. Rather, these cases are properly

read as providing states enough definitional leeway to avoid hypertechnical challenges to

statutes that properly enforce the substantive constitutional restriction set forth in Atkins,

a restriction unquestionably anchored in clinical definitions of mental retardation. See

Bies, 556 U.S. at 831 (acknowledging state procedural and substantive flexibility in

“‘developing appropriate ways to enforce the constitutional restriction’” by determining

who is mentally retarded “within Atkins’ compass”), quoting Atkins, 536 U.S. at 317

(emphasis added); Atkins, 536 U.S. at 317 & n.22 (allowing states variability to the extent

necessary “‘to enforce the constitutional restriction upon [their] execution of sentences’”

and noting in footnote thereafter that non-identical statutes defining mental retardation

“generally conform” to clinical definitions), quoting Ford, 477 U.S. at 416-17 (alteration

in Atkins); Roque, 213 Ariz. 193, ¶ 149, 141 P.3d at 402 (articulating Arizona’s authority

to define mental retardation in context of challenge to Arizona’s IQ threshold, but

43
emphasizing Arizona’s statutory threshold arose from “accepted medical definitions” of

mental retardation).

¶69 As Appendix 1 aptly demonstrates, those states that have addressed the

question of variation have almost universally concluded that state definitions of mental

retardation need not strictly adhere to clinical ones. See Appendix 1.20 But none of those

opinions suggest that states may deviate from clinical understandings of mental

retardation without limitation. To the contrary, those cases which have squarely

contemplated the limits of state flexibility have concluded that conformity with clinical

standards is still required. See Pruitt, 834 N.E.2d at 108 (observing “[t]he Eighth

Amendment must have the same content in all United States jurisdictions” and

concluding state definitions of mental retardation “at odds with” clinical definitions

“would not satisfy” Atkins); Bowling v. Commonwealth, 163 S.W.3d 361, 375 (Ky. 2005)

(recognizing Supreme Court “left it to the states to formulate their own definitions, so

long as they ‘generally conform[ed] to the clinical definitions’ established by the AAMR

and the American Psychiatric Association as approved in Atkins”), quoting Atkins, 536

U.S. at 317 n.22 (alteration in Bowling); Chase v. State, 873 So. 2d 1013, ¶¶ 61-66, 68-72

(Miss. 2004) (noting ambiguity of Atkins’s “national consensus” language, but

20
Several of the cases cited in the appendix concern only procedural questions
such as the timing of an Atkins determination, the appropriate fact-finder, and the burden
and standard of proof. See, e.g., Vasquez, 84 P.3d at 1023; State v. Turner, 936 So. 2d
89, 94, 99 (La. 2006); State v. Jimenez, 908 A.2d 181, 188-89, 191-92 (N.J. 2006); State
v. Poindexter, 608 S.E.2d 761, 767 (N.C. 2005); State v. Laney, 627 S.E.2d 726, 731-32
(S.C. 2006).
44
concluding constitution prohibits execution of all mentally retarded offenders, not subset

thereof, and adopting clinical definitions as standard).

¶70 In Pruitt, the Indiana Supreme Court addressed a very similar problem to

the one presented here. There, the trial court assessed the defendant’s adaptive

functioning by adopting an analytical approach that would predictably characterize only

those with an IQ below sixty as mentally retarded. 834 N.E.2d at 108-09. Observing that

such an approach would “eliminate approximately 75 to 89 percent of all individuals

clinically diagnosed as mentally retarded under the standard medical definitions,” the

supreme court reversed the trial court’s finding and concluded that “[a]lthough variation

is permissible, it cannot go to the point of excluding a majority of those who fit clinical

definitions.” Id. at 110. As discussed above, the analytical approach to adaptive

functioning compelled by Arizona’s statutory standard likewise precludes the diagnosis

of mental retardation for most of those who would be so defined by clinical standards.

¶71 In short, Atkins expressly reserves in the states the power to install varying

procedural frameworks for engaging in the factual determination as to whether a

defendant is clinically mentally retarded. Presumably to preclude hypertechnical

challenges to statutory definitions of clinical mental retardation, the language from Atkins

has also been understood to allow some variation in statutory efforts to define the

condition itself. But, the Court authorized that flexibility exclusively to “‘enforce the

constitutional restriction’” it had imposed upon the states. Atkins, 536 U.S. at 317,

quoting Ford, 477 U.S. at 416. That restriction, as Atkins makes unambiguously clear, is

45
that the mentally retarded, as that condition has been clinically understood, are not

eligible for execution.

In discussing the national consensus against the execution of the “mentally
retarded,” the Court did not intend to alter the clinical definition of that condition.

¶72 In the same vein, the state contends the Atkins majority had no intention to

“adopt any particular definition of ‘mental retardation’ as a constitutional metric.”

Instead, the state posits that “the nature of the legal question, whether a national

consensus existed, defines the scope of the holding.” In so arguing, the state directs us to

language in Atkins leaving individual states the task of identifying those who “fall within

the range of mentally retarded offenders about whom there is a national consensus.” 536

U.S. at 317.

¶73 The state is correct that the core legal question resolved in Atkins was

whether sufficient national “consensus” had developed for the Court to conclude that

“‘evolving standards of decency’” forbade the execution of the mentally retarded. Id. at

311-12, quoting Trop v. Dulles, 356 U.S. 86, 101 (1958). But the presence of a

consensus implies a shared understanding of the topic of consensus. For this reason, the

Court could not have plausibly suggested a national consensus about mental retardation if

it believed the respective states could understand that term to mean markedly different

things.21 Not surprisingly, then, no text in Atkins suggests any significant dispute

21
Given that “mental retardation” is in fact a clinical term, rather than a politically
defined term of art, the Court’s assumption that each state shared that understanding of
the term was not unreasonable. Although to the majority “it seems unlikely the Supreme
Court would have delegated the interpretation of the Eighth Amendment to clinicians,”
supra ¶ 42, I find it more unlikely that the Court would have invited the states to
46
regarding the definition of “mental retardation” around which the Court found a national

consensus had developed. To the contrary, at the outset of its opinion, the Court frames

the issue presented in the trial court by quoting verbatim the clinical standards for

diagnosing that condition. Id. at 308 n.3. And, as discussed, in the paragraphs

immediately leading to the Court’s holding, it expressly describes the logic and content of

the national consensus with express reference to “clinical definitions” of mental

retardation. Id. at 317-18. In so doing, the Court recites, for a second time in the

opinion, the clinical definition itself. Id. Thus, there is little ambiguity as to the

definition of mental retardation employed by the Court when describing the national

consensus of attitudes toward the execution of the mentally retarded.

¶74 Nor does the Court’s particular discussion of a “national consensus,” id. at

316, suggest any other definition of mental retardation. In section III of the opinion, the

Court exhaustively marshals the evidence of that consensus. Id. at 313-17. Therein, the

Court separately itemizes (1) those states that had prohibited the practice of executing the

mentally retarded, (2) those states that allowed the practice but had never done so,

(3) those states that had done so rarely, and (4) the variety of religious and health

organizations that had opposed the practice. Id. Notably absent in that discussion is any

effort to qualify, modify, or redefine the meaning of the term “mental retardation” in any

substantially redefine the nature of the substantive restriction on their own authority to
determine punishment. In context, it seems especially unlikely that the Court would have
endowed lay legislators with the authority to substantially redefine a medical condition,
when the Court’s holding is itself anchored in the nature of the condition as clinically
described.

47
of those contexts. Rather, the Court uses the term generically and in conformity with its

preceding reference to the clinical definition. And, within the Court’s discussion of

national consensus, it applies clinical understandings of mental retardation when it

presumes that those few individuals executed since Penry v. Lynaugh, 492 U.S. 302

(1989), with IQs below seventy, were mentally retarded. Atkins, 536 U.S. at 316.

¶75 In a lone footnote, the Atkins majority acknowledges some minor variation

in statutory language used by states to define mental retardation. See id. at 317 n.22

(“The statutory definitions of mental retardation are not identical, but generally conform

to the clinical definitions set forth in n. 3, supra.”). But, the Court notes that variation not

to suggest any disagreement among the states as to the consensus understanding of what

mental retardation entails, but rather to reinforce the conclusion that a true national

consensus had coalesced around a shared definition of the condition—the “clinical

definitions set forth in n. 3.”22 Id.

22
As discussed above, the language of then § 13-703.02 suggests Arizona’s
legislature intended to express clinical understandings of mental retardation in statutorily
defining the condition. And, the Atkins majority correctly identified that legislation as
evidence of a trend against the execution of the mentally retarded. 536 U.S. at 315 n.15.
But I cannot agree that, in so acknowledging the efforts of the respective states to
statutorily prohibit the execution of the mentally retarded as evidence of evolving social
norms, the Court intended to thereafter preclude any future conceivable specific Eighth
Amendment challenge to those referenced statutes. Notably, the Atkins majority cited
Kansas’s statute as an example of the growing national consensus against executing the
mentally retarded. 536 U.S. at 314 & n.12. But, as Justice Scalia emphasized in his
dissent, Kansas’s statutory definition of mental retardation would exclude the mildly
mentally retarded, a group that Scalia acknowledges are now unambiguously exempted
from execution by the holding of Atkins. Id. at 342-43 & 343 n.2 (Scalia, J., dissenting).
Moreover, the majority clearly contemplates that the states would, as a response to its
holding, be generating legislation designed to “enforce the constitutional restriction upon
48
¶76 Thus, there is no evidence in the Court’s actual discussion of the national

consensus that it intended to define “mental retardation” as a legal term of art, varying

from state to state based on the political views of each state legislature, untethered to the

accepted clinical definitions of that condition. Rather, within its discussion of national

consensus, the Court logically employs the consensus definition of that term (and indeed

its only recognized definition): a specific scientific classification of mental disability. In

context, then, when the Court refers to a “range of mentally retarded offenders about

whom there is a national consensus,” id. at 317, it is referring to the spectrum of

offenders that would be clinically defined as mentally retarded. It is not referring to a

range of varying definitions of the condition.

¶77 Ultimately, when the Atkins Court held that evolving standards of decency

placed “‘a substantive restriction on the State’s power to take the life’ of a mentally

retarded offender,” id. at 321, quoting Ford, 477 U.S. at 405, implicit in that holding was

the definition of mental retardation that the Court itself had set forth and repeatedly used

in formulating its opinion, the only definition in existence around which a national

consensus could develop.23 Put another way, the Court’s restriction on the states would

[their] execution of sentences.” Id. at 317, quoting Ford, 477 U.S. at 416-17 (alteration
in Atkins).
23
The state and the majority observe that clinical definitions may evolve. The state
therefore suggests it would have made little sense for the Court to adopt any particular
clinical definition. Indeed, as noted, one of the two organizations providing a clinical
definition at the time of Atkins, the AAMR, has since changed its own name, the name of
the condition formerly called mental retardation, and the wording (but not the substance)
of the definition of that condition. But the fact that the prevailing clinical definition of
mental retardation could conceivably evolve does not demonstrate that the Court in
49
neither be effectively “categorical” nor “substantive” in the absence of a concrete

definition of mental retardation.

The Arizona Supreme Court has neither addressed nor resolved the issue.

¶78 Lastly, my colleagues maintain that the Arizona Supreme Court has already

addressed and resolved the question here and that we are consequently bound by that

decision. Specifically, they observe that in Grell II, our supreme court identified the

primary differences between the statutory definition for mental retardation and the

DSM-IV standard and nonetheless found no fault with the trial court’s application of the

statutory standard. See Grell II, 212 Ariz. 516, ¶¶ 62-63, 135 P.3d at 709.

¶79 That passage, quoted at length in ¶ 53 supra, indeed identifies and

articulates differences between the statutory definition of mental retardation and the

clinical one. But it does not purport to address the question we must answer here:

whether that deviation constitutes a violation of Eighth Amendment standards set forth by

the United States Supreme Court in Atkins. The passage from Grell II, which addresses a

challenge to the sufficiency of a trial court’s findings under Arizona law, makes no

mention of Atkins, the United States Constitution, or any other federal case law. See 212

Ariz. 516, ¶¶ 58-63, 135 P.3d at 708-09. It undertakes none of the reasoning necessary to

consider whether Atkins requires substantial conformity to the clinical definition or

Atkins intended to eschew clinical definitions in favor of political ones. Unlike clinical
definitions, which represent a widely accepted scientific consensus, the respective states
have no duty to form any consensus definition of the condition. For this reason, state
definitions of mental retardation would provide no jurisprudential stability at all except to
the extent they incorporate recognizable clinical standards.
50
whether Arizona’s statute, notwithstanding the noted deviations from the clinical

definition, so conforms.

¶80 The majority speculates that the Grell II passage implicitly resolved the

federal constitutional issue we face here because the court had addressed other analogous

federal constitutional challenges to our statute in the same opinion. See supra ¶¶ 38-39.

But, the discussion of the requirements of Atkins occurred in a separately headed section

of the opinion and addressed only the authority of our state to erect its own procedural

framework to enforce Atkins. See Grell II, 212 Ariz. 516, ¶¶ 21-49, 135 P.3d at 701-07.

Grell II did not address the extent of the state’s authority to redefine the class of persons

eligible for execution under Atkins.

¶81 Finally, the majority suggests our supreme court resolved the question here

when it observed in a footnote in a prior case that Arizona’s statute “appears to comport

substantively and procedurally with the principles set forth in Atkins.” Grell I, 205 Ariz.

57, n.4, 66 P.3d at 1241 n.4. But that observation did not address any specific claim

raised in the case before it, and was therefore dicta. Moreover, the court employed the

phrase “appears to,” suggesting it was not purporting to have definitively considered or

resolved all future Atkins-based challenges to the statute. Indeed, when Grell later raised

a host of claims that Arizona’s procedural framework violated Atkins, the court

exhaustively addressed those claims on their merits, and it declined to cite its previous

footnote as authority for rejecting them. See Grell II, 212 Ariz. 516, ¶¶ 21-49, 135 P.3d

at 701-07. Thus, the supreme court manifestly did not view itself as resolving any

specific Atkins-based challenges to the statute when it authored the Grell I footnote.

51
¶82 Although we are bound by the holdings of our highest state court, I cannot

assume our supreme court has resolved an issue of such gravity and legal complexity

without engaging in any of the reasoning necessary to do so, without identifying the

constitutional issue it was intending to address, and without citing the pivotal case

controlling that issue. See Calnimptewa v. Flagstaff Police Dep’t, 200 Ariz. 567, ¶ 24, 30

P.3d 634, 639 (App. 2001) (appellate opinions should not be read as authority for matters

not “specifically presented and discussed”). Nor would such an assumption be consistent

with our own rules for determining which claims have been properly presented by

litigants and resolved by our courts. Claims based on state statute, state constitution,

state common law, and federal constitutional law are not identical, even in the context of

similar underlying facts, and it is not our courts’ custom to decide distinct legal claims

not raised by a party. See, e.g., State v. Ovante, 231 Ariz. 180, ¶ 18 & n.1, 291 P.3d 974,

979 & n.1 (2013) (supreme court will review properly developed constitutional claims,

but court “does not consider or address unsupported constitutional claims”); State v.

Tison, 129 Ariz. 526, 535, 633 P.2d 335, 344 (1981) (constitutional issues precluded

when not raised below, even though defendant objected to admission of evidence on

other grounds); State v. Alvarez, 213 Ariz. 467, ¶ 7, 143 P.3d 668, 670 (App. 2006)

(hearsay objection insufficient to preserve claim based on Sixth Amendment

Confrontation Clause); see also Ariz. State Bank v. Crystal Ice & Cold Storage Co., 26

Ariz. 205, 211, 224 P. 622, 623 (1924) (statutes presumed constitutional until question of

constitutionality properly raised before court).

52
¶83 In determining what issues a court’s holding has resolved, we must

necessarily be guided, as all trial courts and attorneys must, by the court’s own

description of the issues it purports to be addressing and resolving. Thus considered, I

cannot find any indication in either Grell I or Grell II that our supreme court intended to

address or resolve the question now before us.

Conclusion

¶84 My colleagues’ scholarly rebuttal suggests some non-trivial explanations

for why the Court in Atkins appeared to allow the respective states a measure of

flexibility in defining mental retardation: respect for the salutary effects of federalism

and the traditional deference owed the states in formulating punishments. But in my

view, we must also recognize that the Atkins majority invited flexibility to enforce its

holding, not eviscerate it. Nor can we overlook that the Atkins holding itself expressly

places a substantive restriction on traditional state authority to determine punishment.

Because Arizona’s statutory standard for determining adaptive functioning would

disqualify most mentally retarded persons from the protection of Atkins, it exceeds the

bounds of any flexibility the Court intended to provide states in enforcing the

constitutional restriction. I therefore respectfully dissent.

/s/ Peter J. Eckerstrom
PETER J. ECKERSTROM, Presiding Judge

53
 Appendix 1

State Court Decisions Addressing Definitions of Mental Retardation after Atkins
State Did Atkins leave Source
substantive definition
of MR to states?
AL Yes Morris v. State, 60 So. 3d 326, 339 (Ala. Crim. App. 2010)

AR Yes Anderson v. State, 163 S.W.3d 333, 354-55 (Ark. 2004)

AZ Yes State v. Grell, 212 Ariz. 516, ¶¶ 24-25, 62, 135 P.3d 696, 701,
709 (2006)
CA Yes People v. Jackson, 199 P.3d 1098, 1107, 1109 (Cal. 2009)

CO Yes People v. Vasquez, 84 P.3d 1019, 1022 (Colo. 2004)

FL Yes State v. Herring, 76 So. 3d 891, 894 (Fla. 2011)

GA Yes Stripling v. State, 711 S.E.2d 665, 668 (Ga. 2011)

ID Yes Pizzuto v. State, 202 P.3d 642, 649 (Idaho 2008)

IL Yes People v. Pulliam, 794 N.E.2d 214, 236-37 (Ill. 2002) (but
death penalty no longer available in Illinois)
IN Yes Pruitt v. State, 834 N.E.2d 90, 109-10 (Ind. 2005)

KY Yes Bowling v. Commonwealth, 163 S.W.3d 361, 376 (Ky. 2005)

LA Yes State v. Turner, 936 So. 2d 89, 92 (La. 2006)

MO Yes State v. Johnson, 244 S.W.3d 144, 150 (Mo. 2008)

MS Yes Chase v. State, 873 So. 2d 1013, 1027-28 (Miss. 2004)

NC Yes State v. Poindexter, 608 S.E.2d 761, 765 (N.C. 2005)

NJ Yes State v. Jimenez, 908 A.2d 181, 183, 189 (N.J. 2006) (but
death penalty no longer available in New Jersey)
NM Yes State v. Trujillo, 160 P.3d 577, 581-82 (N.M. Ct. App. 2007)
(but death penalty no longer available in New Mexico)
NV Yes Ybarra v. State, 247 P.3d 269, 273 (Nev. 2011)

OH Yes State v. Were, 890 N.E.2d 263, ¶¶ 175-76 (Ohio 2008)

OK Yes Smith v. State, 245 P.3d 1233, ¶ 3 (Okla. Crim. App. 2010)

54
State Did Atkins leave Source
substantive definition
of MR to states?
PA Yes Commonwealth v. DeJesus, 58 A.3d 62, 81 (Pa. 2012)

SC Yes State v. Laney, 627 S.E.2d 726, 730 (S.C. 2006)

TN Yes Coleman v. State, 341 S.W.3d 221, 234 (Tenn. 2011); Howell
v. State, 151 S.W.3d 450, 457 (Tenn. 2004)
TX Yes Ex parte Briseno, 135 S.W.3d 1, 4-8 (Tex. Crim. App. 2004)
(Supreme Court left procedure to states, absent Texas
legislative enactment, adopting AAMR definition or
definition in Texas Health and Safety Code)
UT Yes State v. Maestas, 2012 WL 3176383, ¶¶ 187-88 (Utah July
27, 2012)
VA Yes, by implication Burns v. Commonwealth, 688 S.E.2d 263, 264 (Va. 2010)

55
 Appendix 2

State Statutory Standards for Determining Mental Retardation
State Requires Requires Requires 2+ IQ threshold? Prior to Source
diminished diminished diminished Age
intellectual adaptive adaptive
ability? behavior? behavior?
AL Yes Yes No None “Develop- Ala. Code § 15-24-2(3) (defines
mental MR for defendants generally,
period” cited in Morris v. State, 60 So. 3d
326, 339 (Ala. Crim. App. 2010)
for definition in capital
prosecution)
AR Yes Yes No Presumption 18 Ark. Code Ann. § 5-4-618
of MR at IQ of
65
AZ Yes Yes No 70 18 A.R.S. § 13-753(K)(3), (5)
CA Yes Yes No No 18 Cal. Penal Code § 1376
CO Yes Yes No No "Develop- Colo. Rev. Stat. § 18-1.3-1101
mental
period"
CT Yes Yes No two standard 18 Conn. Gen. Stat. § 1-1g (death
deviations penalty no longer available in
below mean Connecticut)
DE Yes Yes Yes 70 18 Del. Code Ann. tit. 11,
§ 4209(d)(3)
FL Yes Yes No two standard 18 Fla. Stat. § 921.137
deviations
below mean
GA Yes Yes No No "Develop- Ga. Code Ann. § 17-7-131
mental
period"
ID Yes Yes Yes 70 18 Idaho Code Ann. § 19-2515A(1)
IN Yes Yes No No 22 Ind. Code § 35-36-9-2
KS Yes Yes No two standard 18 Kan. Stat. Ann. §§ 21-6622(h);
deviations 76-12b01(d)
below mean
KY Yes Yes No 70 "Develop- Ky. Rev. Stat. Ann. § 532.130
mental
period"
LA Yes Yes No No 18 La. Code Crim. Proc. Ann. art.
905.5.1(H)(1)
MD Yes Yes No 70 22 Md. Code Ann., Crim. Law § 2-
202(b)(1)(i)-(ii)

56
State Requires Requires Requires 2+ IQ threshold? Prior to Source
diminished diminished diminished Age
intellectual adaptive adaptive
ability? behavior? behavior?
MO Yes Yes Yes No 18 Mo. Rev. Stat. § 565.030(6),
found unconstitutional on other
grounds by State v. Whitfield, 107
S.W.3d 253, 261 (Mo. 2003)
NC Yes Yes Yes 70 18 N.C. Gen. Stat. Ann. § 15A-2005
NE Yes Yes No Presumption None Neb. Rev. Stat. § 28-105.01(3)
of MR at IQ of
70
NM Yes Yes No Presumption None N.M. Stat. Ann. § 31-9-1.6(E)
of MR at IQ of (death penalty no longer
70 available in New Mexico)
NV Yes Yes No No “Develop- Nev. Rev. Stat. § 174.098(7)
mental
period”
NY Yes Yes No No 18 N.Y. Crim. Proc. Law
§ 400.27(12)(e) (death penalty no
longer available in New York)
OK Yes Yes Yes 70 18 Okla. Stat. tit. 21, § 701.10b(A)-
(B)
SC Yes Yes No No “Develop- S.C. Code Ann. § 44-20-30(12)
mental (defines MR in health context,
period” cited in death penalty case State
v. Stanko, 2013 WL 696816, *16
(S.C. Feb. 27, 2013))
SD Yes Yes No Presumption 18 S.D. Codified Laws §§ 23A-27A-
of no MR if IQ 26.1, 23A-27A-26.2
greater than
70
TN Yes Yes No 70 18 Tenn. Code Ann. § 39-13-203
TX Yes Yes No two standard “Develop- Tex. Code Ann. § 591.003
deviations mental (defines MR in health and safety
below mean period” code, adopted as part of
standard in death penalty cases,
Ex parte Hearn, 310 S.W.3d 424,
427-28 (Tex. Crim. App. 2010))
UT Yes Yes No No 22 Utah Code Ann. § 77-15a-102
VA Yes Yes No two standard 18 Va. Code Ann. § 19.2-264.3:1.1(A)
deviations
below mean
WA Yes Yes No 70 18 Wash. Rev. Code § 10.95.030(2)

57