PUBLISHED


UNITED STATES COURT OF APPEALS
             FOR THE FOURTH CIRCUIT


UNITED STATES OF AMERICA,            
             Petitioner-Appellant,
               v.                          No. 12-6832
PATRICK CAPORALE,
            Respondent-Appellee.
                                     
        Appeal from the United States District Court
   for the Eastern District of North Carolina, at Raleigh.
            Terrence W. Boyle, District Judge.
                    (5:08-hc-02037-BO)

               Argued: September 20, 2012

                Decided: December 6, 2012

  Before KING, GREGORY, and WYNN, Circuit Judges.



Affirmed by published opinion. Judge King wrote the opin-
ion, in which Judge Gregory and Judge Wynn joined.
2                 UNITED STATES v. CAPORALE
                         COUNSEL

ARGUED: Dana Lydia Kaersvang, UNITED STATES
DEPARTMENT OF JUSTICE, Washington, D.C., for Appel-
lant. Lewis A. Thompson, III, BANZET, THOMPSON &
STYERS PLLC, Warrenton, North Carolina, for Appellee.
ON BRIEF: Stuart F. Delery, Acting Assistant Attorney Gen-
eral, Mark B. Stern, UNITED STATES DEPARTMENT OF
JUSTICE, Washington, D.C.; Thomas G. Walker, United
States Attorney, Raleigh, North Carolina, for Appellant.


                          OPINION

KING, Circuit Judge:

   The government appeals the judgment of the district court
directing that Patrick Caporale be freed from the custody of
the Bureau of Prisons and granted supervised release. Capor-
ale finished serving his prison sentence for child molestation
in 2008, but he has remained incarcerated while the govern-
ment seeks to have him declared a "sexually dangerous per-
son" pursuant to the civil-commitment provisions of the
Adam Walsh Child Protection and Safety Act of 2006 (the
"Walsh Act"), Pub. L. No. 109-248, 120 Stat. 587, as specifi-
cally set forth in 18 U.S.C. § 4248.

   A sexually dangerous person under the Walsh Act means
one "who has engaged or attempted to engage in sexually vio-
lent conduct or child molestation and who is sexually danger-
ous to others." 18 U.S.C. § 4247(a)(5). A person is sexually
dangerous to others insofar as he or she "suffers from a seri-
ous mental illness, abnormality, or disorder," and, as a result,
"would have serious difficulty in refraining from sexually vio-
lent conduct or child molestation if released." 

parties have never disputed that, as evi-
denced by his several convictions, discussed infra, Caporale
                  UNITED STATES v. CAPORALE                   3
satisfies the first, prior-conduct element of § 4247(a)(5) by
having engaged in child molestation.

   Following an evidentiary hearing whose scope was thereby
limited to the second element of § 4247(a)(5), the district
court ruled that, as a matter of law, the government had not
proved that Caporale suffers from a serious mental illness,
abnormality, or disorder. The court perceived in the alterna-
tive that even if Caporale were so afflicted, his commitment
was not required because the government had also failed to
sufficiently show that Caporale will experience serious diffi-
culty in refraining from sexually violent conduct or child
molestation if released.

   We conclude that, contrary to the district court’s legal
determination and as established by the evidence, Caporale
indeed suffers from a qualifying mental impairment. We nev-
ertheless affirm the judgment below, discerning no clear error
in the court’s alternative rationale that the government fell
short of carrying its burden to demonstrate a relative likeli-
hood that Caporale will reoffend.

                               I.

                              A.

   Caporale, fifty-nine, has a history of sexual offenses
involving minors. In 1980, he pleaded guilty in New York to
fourth-degree criminal facilitation for recruiting about twenty
boys, aged thirteen to sixteen, to have sex with an adult
female acquaintance. Then, in 1984, Caporale pleaded guilty
to state charges of second-degree sexual abuse for subjecting
a thirteen-year-old boy to sexual contact and for masturbating
another boy, twelve years old. In 1986, Caporale again
pleaded guilty in state court to acting in a manner injurious to
a child, after having persuaded three boys (age fifteen to six-
teen) and a girl (age sixteen) to have sex in his apartment
while he watched.
4                 UNITED STATES v. CAPORALE
   Similar conduct in Maine involving a seventeen-year-old
boy and the boy’s underage female acquaintances — this time
evidenced by photographs and videotapes — resulted in
Caporale’s 1992 federal indictment for child molestation and
for possession of child pornography. Caporale pleaded guilty
to six counts of using a minor to engage in sexually explicit
conduct, for which he was sentenced to eighty-four months in
prison; he served most of that sentence before being granted
supervised release in August 1998. On December 14, 1999,
Caporale was arrested and charged with endangering the wel-
fare of a minor following the discovery by local authorities
that he had discussed masturbation with a fourteen-year-old
male neighbor. The charge was a parole violation, for which
Caporale was returned to prison in September 2000.

   Caporale was again released to supervision in June 2001,
and, two years later, was once more found to have violated his
conditions of release by associating with a felon, a man who
Caporale described as a pedophile. On September 12, 2003,
federal probation officers arrived unannounced at Caporale’s
residence and arrested him upon finding numerous videotapes
and photographic depictions of underage boys engaged in sex-
ual activity. Caporale subsequently pleaded guilty in New
York to a single count of possessing photographs of an
obscene performance of a child.

   Caporale’s myriad violations ultimately resulted in the
revocation of his supervised release. Consequently, Caporale
returned to federal prison in 2003 to serve out the sentence
imposed on his 1992 conviction. Caporale completed his sen-
tence on March 21, 2008, and he would have been released
but for the government’s filing that day of a certification pur-
suant to the Walsh Act, which provides, in pertinent part:

    In relation to a person who is in the custody of the
    Bureau of Prisons, . . . the Attorney General . . . may
    certify that the person is a sexually dangerous per-
    son, and transmit the certificate to the clerk of the
                   UNITED STATES v. CAPORALE                     5
    court for the district in which the person is confined
    . . . . The court shall order a hearing to determine
    whether the person is a sexually dangerous person.
    A certificate filed under this subsection shall stay the
    release of the person pending completion of proce-
    dures contained in this section.

18 U.S.C. § 4248(a).

   The district court was thus required to convene a hearing to
afford the government the opportunity to prove the ultimate
truth of its certification, i.e., that Caporale is sexually danger-
ous. An inmate so declared must be kept in federal custody
until the danger has passed, or until a state agrees to assume
responsibility for the inmate’s custody, care, and treatment.
See 18 U.S.C. § 4248(d). As it was uncontested that Capor-
ale’s criminal history satisfied the first element toward a
determination of sexual dangerousness, it was the govern-
ment’s burden at hearing to establish, by clear and convincing
evidence, both prongs of the second element: (1) that Capor-
ale is impaired by a serious mental illness, abnormality, or
disorder, such that (2) he would, if released, have serious dif-
ficulty refraining from sexually violent conduct or child
molestation. See 

   All proceedings having been stayed during the pendency of
constitutional challenges to the commitment proceedings
wrought by the Walsh Act, see United States v. Hall, 

, 461 n.2 (4th Cir. 2012) (reciting litigation history begin-
ning with Supreme Court’s ruling in United States v. Com-
stock, 

(2010), and progressing through our
decision on remand, United States v. Comstock, 

(4th Cir. 2010), and its companion case, Timms v. Johns, 

(4th Cir. 2010)); see also United States v. Timms,

(4th Cir. 2012), Caporale’s hearing at last took
place on March 5, 2012. The key witnesses were Dr. Gary
6                    UNITED STATES v. CAPORALE
Zinik and Dr. Lela Demby for the government, and Dr.
Joseph Plaud for the defense, all of whom are licensed psy-
chologists and acknowledged experts. All three reviewed the
available records, and Dr. Plaud personally interviewed
Caporale on November 28, 2011.

                                    1.

   According to Dr. Plaud, Caporale is sexually interested in
pubescent to post-pubescent males, "12, 13 up to 16, 17 years
old." J.A. 253.1 Dr. Plaud explained that, inasmuch as the rate
of physical maturity may differ depending on the person,
stages of development defy rigid definition by age, and that
it was more accurate to say that Caporale’s sexual interests
correspond approximately to Stages III-V on the five-stage
Tanner scale of physical development. See 

Zinik essentially concurred with Dr. Plaud, noting that Capor-
ale is attracted to "[p]ubescent boys, ages approximately 11 to
14." 

contrast to her two colleagues, Dr. Demby opined that
Caporale prefers "pre-pubescent boys . . . who ha[ve] not yet
developed the majority of their secondary sexual characteris-
tics . . . as young as 11 and 12." J.A. 207. Asked to elaborate,
Dr. Demby noted that, as set forth in the American Psychiatric
Association’s Diagnostic and Statistical Manual of Mental
Disorders ("DSM"), age thirteen is the presumed ceiling for
a diagnosis of pedophilia, a condition in which the afflicted
individual is attracted to pre-pubescent children. See 

not confronted directly with Dr. Demby’s testi-
mony, Dr. Plaud flatly rejected her diagnosis:
    1
   Citations herein to "J.A. ___" refer to the contents of the Joint Appen-
dix filed by the parties to this appeal.
                  UNITED STATES v. CAPORALE                   7
    [Caporale’s] not a pedophile . . . . [T]he DSM in the
    diagnostic criteria for pedophilia does in parenthesis
    say in terms of an age range, generally age 13 years
    and younger. That’s worthless information. That
    doesn’t inform a professional about anything . . . .
    We have to look at what the data are because the
    critical feature is whether or not the person has
    attained evidence of secondary sexual characteris-
    tics.

J.A. 253. Dr. Zinik agreed with the defense on this point, as
confirmed by the district court’s questioning:

      JUDGE BOYLE: This patient or subject is not a
    pedophile. You don’t describe him — you don’t
    define him a pedophilic with children who are below
    11, say?

       [DR. ZINIK]: That’s correct. I —

      JUDGE BOYLE: You think his target group is the
    pubescent group?

      THE WITNESS: Definitely those who have —
    boys who have reached puberty.

       JUDGE BOYLE: Yeah. Yeah.

      THE WITNESS: He’s not interested in pre-
    pubescent children.

Dr. Demby acknowledged that "individu-
ally, [puberty] can happen at different ages for different chil-
dren," and she frankly admitted that it would "be possible for
somebody who is say, 12 years old to be pubescent." 

Demby nonetheless diagnosed Caporale with "pe-
dophilia, sexually attracted to males exclusive type, and per-
8                  UNITED STATES v. CAPORALE
sonality disorder not otherwise specified of dependent traits."

Zinik, having ruled out pedophilia, concluded that
Caporale suffers from "paraphilia . . . not otherwise specified"
("paraphilia/NOS"). J.A. 161. Dr. Zinik testified that
paraphilia is characterized by "recurrent, intense sexual fanta-
sies, urges or behaviors that . . . last at least six months and
. . . cause a significant impairment or disability in the individ-
ual’s life." 

Zinik occasionally referred to
Caporale’s affliction as "hebephilia," which is not a listed
diagnosis in the DSM, but a colloquial term understood by
mental health professionals to describe the proclivity of "men
whose erotic interest centers on pubescents." Ray Blanchard
et al., Pedophilia, Hebephilia, and the DSM-V, 38 Archives of
Sexual Behavior 335, 336 (2009).

   Only eight specific manifestations of paraphilia —
pedophilia, transvestic fetishism, exhibitionism, fetishism
with respect to inanimate objects, voyeurism, sexual masoch-
ism, sexual sadism, and frotteurism (sexual rubbing against
non-consenting persons, often in crowded areas) — are docu-
mented in the Fourth Edition of the DSM, a treatise accepted
as authoritative by all the testifying experts. More generally,
however, the DSM lists paraphilia/NOS, which, Dr. Zinik
explained, "is a residual category in which really all — any
other paraphilia could be included." J.A. 167. According to
Dr. Zinik, a paraphilia’s inclusion within the catch-all should
not depend on whether it is diagnosed with a frequency simi-
lar to those specifically listed, but should merely consider
whether it impairs normal functioning to a similar extent. See

opinion, Caporale’s paraphilia "has caused
serious impairment and dysfunction in his life, because he
keeps getting arrested and . . . he’s never really done much
with his life, because of this obsession." J.A. 161. Dr. Demby
agreed with Dr. Zinik that impairment is the hallmark of a
                   UNITED STATES v. CAPORALE                    9
mental infirmity, explaining that "[t]he mere presence of sex-
ual attraction to pubescent children in and of itself does not
qualify as a mental disorder. The disorder comes when it
causes one to have significant impairment in major spheres of
functioning in one’s life." See 

opposition to the government’s experts, Dr. Plaud opined
that Caporale suffers from no qualifying illness, abnormality,
or disorder under § 4247(a). Dr. Plaud observed that Capor-
ale’s sexual liaisons with pubescent males "is illegal to a cer-
tain extent . . . upwards, depending on the jurisdiction. But it
is not disordered." J.A. 251. Emphasizing the lack of a dis-
crete DSM listing for Caporale’s affliction, Dr. Plaud rea-
soned that "if it’s . . . truly a disorder, it should have its own
diagnostic criteria. [I]f you have something that’s high fre-
quency and it’s truly considered by the professional commu-
nity to be a disorder, then you’ve got to have a diagnosis for
it." 

Plaud took issue with Dr. Zinik’s conclusion that, for
purposes of the Walsh Act, hebephilia can be shoehorned into
the DSM-listed diagnosis of paraphilia/NOS, explaining that
"[i]t’s only more recently that this has become something, I
think, as a consequence of what’s going on in the larger civil
commitment world . . . . And I do not believe a paraphilia
NOS is . . . for that purpose. It’s there for low frequency sex-
ual aberration." J.A. 253. Cross-examination served to further
develop Dr. Plaud’s position. Asked by the government’s
counsel whether sexual acts with a thirteen-year-old boy are
deviant, Dr. Plaud responded:

    They’re sexually deviant in the legal sense of the
    term, and in the moral sense . . . they can be, of
    course. But diagnostically, is — does he have an
    underlying sexual disorder? That’s the issue . . . .
    The issue with sexual deviance is, is an individual
    attracted underneath, sexually motivated by a range
    of experiences that lie outside of what most human
10                 UNITED STATES v. CAPORALE
     beings experience . . . . That’s what makes it not sex-
     ually deviant. It may be socially deviant. It may be
     morally deviant. Put any word you want on it, but
     I’m here to talk about sexual deviance in that diag-
     nostic sense.

   The experts also differed on whether, upon release, Capor-
ale would have serious difficulty in refraining from child
molestation. Dr. Zinik, referring to pornographic stories about
teenage boys composed in 2008 by Caporale and other
inmates using a prison computer, along with suggestive pho-
tographs seized from Caporale’s cell in May 2011, opined that
Caporale is "still pretty charged up about pubescent boys, and
that he would have, as a result of that, serious difficulty con-
trolling himself in the community if he were in contact with
pubescent boys as he has done in the past." J.A. 179. Using
three actuarial scales to compute the risk of recidivism for
offenders having similar characteristics, Dr. Zinik recited that
two of the scales scored Caporale’s profile "high risk" and the
third "moderate high risk." See 

the stories and photos "evidence of
sexual preoccupation, attitudes that condone sexual contact
with children as recently as 2011," which, in her opinion, "ex-
acerbates Mr. Caporale’s risk." J.A. 227. Noting that Caporale
had, in the past, personally taken photographs of illegal activ-
ity, Dr. Demby also took into account Caporale’s more recent
writings to differentiate his behavior from that typically
exhibited by a passive consumer of pornography. See 

Demby conceded that writing such stories was not
itself "sexually dangerous" in the sense that it did not equate
to molestation or violent conduct, but she offered that "if
someone knows that they are at risk . . . of molesting children
and reengaging in past behavior . . . to continue to create and
                     UNITED STATES v. CAPORALE                11
read . . . and reiterate these themes . . . it’s leading him down
that same path of a hands-on offense, of re-doing what . . . we
already know he’s done in the past." 

Plaud, on the other hand, thought it more significant
that, since 1992, Caporale had committed only noncontact
offenses. Caporale had further demonstrated volitional capac-
ity, according to Dr. Plaud, by having ceased writing stories
since May 2011, almost a year prior to the hearing. See J.A.
255, 261, 265-66. Dr. Plaud was reluctant to use actuarial
tools to predict Caporale’s probability of recidivism, testify-
ing that such tools more accurately serve to exclude test sub-
jects as candidates to engage in molestation, rather than to
pinpoint them as likely offenders. See 

Zinik
also acknowledged the tools’ limitations. See 

Plaud emphasized, moreover, that even those tools indicating
that Caporale posed a "high risk" of recidivism more particu-
larly defined that risk as amounting to less than a twenty-five
percent chance over eight years. Indeed, as Dr. Plaud pointed
out, the actuarial risk decreases to less than ten percent for
men sixty and over, an age that Caporale has nearly attained.
See 

   On April 20, 2012, the district court issued a written deci-
sion, designated its "Findings of Fact and Conclusions of
Law" (the "Findings"), in which the court set forth its ruling
against the government.2 Most significantly, the court con-
cluded that hebephilia — whether elevated to a titular condi-
tion or deemed a shorthand way to describe the behavioral
predicate underlying a more formal diagnosis of
paraphilia/NOS — "is not an appropriate basis for civil com-
mitment under the [Walsh] Act." Findings 6. This is so, the
court reasoned, because whether hebephilia constitutes a seri-
ous mental illness, abnormality, or disorder "is a contested
  2
   The district court’s Findings appear at J.A. 338-47.
12                UNITED STATES v. CAPORALE
issue in the mental health community." 

the
court found "more credible the testimony of Dr. Plaud, who
opined . . . that Mr. Caporale currently suffers from no sexual
disorder." 

district court also regarded Dr. Plaud’s testimony more
credible concerning the issue of whether Caporale would have
serious difficulty refraining in the future from sexually violent
conduct or child molestation. The court stressed that Dr.
Plaud had focused pertinently on evidence of Caporale’s cur-
rent condition, more so than Dr. Zinik and Dr. Demby, who
based their evaluations in substantial part on Caporale’s his-
torical misconduct. See Findings 9-10.

  Consistently with its written decision, the district court
entered judgment for Caporale and ordered his release. The
government timely appealed, and, contemporaneously there-
with, filed an emergency motion for a stay of the district
court’s order, which we granted. The parties’ respective argu-
ments having thereafter been fully briefed and argued on an
expedited basis, the matter is ripe for our disposition.

                               II.

   To the extent that the district court’s judgment in this
Walsh Act proceeding rests on the accuracy of its findings of
fact, our review is for clear error. See United States v. Hall,

, 462 (4th Cir. 2012). The court’s legal conclu-
sions, however, are paid no deference. See 

our
limited role as an appellate court to disturb the judgment
below solely on the ground that we would have decided to the
contrary. See 

Bessemer City, 

, 573 (1985)). Precedent dictates that "‘[i]f the district
court’s account of the evidence is plausible in light of the
record viewed in its entirety, the court of appeals may not
reverse it even though convinced that had it been sitting as the
trier of fact, it would have weighed the evidence differently.’"
United States v. Wooden, 

, 451 (4th Cir. 2012)
                   UNITED STATES v. CAPORALE                   13
(quoting Bessemer City at 573-74). We must therefore permit
the court’s judgment to stand unless, with respect to its factual
underpinnings, we are "‘left with the definite and firm convic-
tion that a mistake has been committed.’" 

(quoting Easley v. Cromartie, 

, 242
(2001)).

                               III.

                               A.

                                1.

   The experts at hearing were asked whether Caporale suffers
from a serious mental illness, abnormality, or disorder, and
they each strove to answer that question within the strictures
of the DSM, which is familiar territory to mental health pro-
fessionals. Dr. Demby perceived Caporale’s sexual aberra-
tions as suggesting a specific DSM diagnosis, while Dr. Zinik
thought the aberrant behavior corresponded more closely to a
less rigorous, residual category of diagnoses yet within the
DSM. Dr. Plaud was of the opinion that Caporale’s condition
conformed to no listed diagnosis, which in turn precluded the
possibility that Caporale suffers from any serious mental
impairment.

   The district court’s resolution of the issue in reliance on Dr.
Plaud’s testimony cannot be squared with United States v.
Carta, 

(1st Cir. 2010), another Walsh Act pro-
ceeding involving a diagnosis of hebephilia, in which the
lower court’s judgment on behalf of the respondent was prem-
ised on a legal analysis similar to that under review here. In
Carta, the respondent’s expert "asserted that hebephilia was
not a generally accepted diagnosis in the mental health com-
munity, did not fit within the DSM definition of paraphilia,
lacked diagnostic criteria and could not be consistently
defined." 

with that testimony, the dis-
trict court entered judgment in favor of the respondent, ruling
14                   UNITED STATES v. CAPORALE
that hebephilia is not a serious mental illness, abnormality, or
disorder.

   The First Circuit vacated and remanded, observing that the
lower court "may have assumed that the statutory concept is
delimited by the consensus of the medical community, but
this is not so. Further, a mental disorder or defect need not
necessarily be one so identified in the DSM in order to meet
the statutory requirement." 

. The
Carta court’s rationale appears sound, as one will search
§ 4247(a)(6) in vain for any language purporting to confine
the universe of qualifying mental impairments within clinical
or pedagogical parameters. The statute could have been
drafted to comport with clinical norms, but inasmuch as Con-
gress chose not to do so, it has been left to the courts to
develop the meaning of "serious mental illness, abnormality,
or disorder" as a legal term of art. See Kansas v. Hendricks,

, 359 (1997) (explaining that "the term ‘mental
illness’ is devoid of any talismanic significance," and instruct-
ing that legislative development of specialized "[l]egal defini-
tions . . . need not mirror those advanced by the medical
profession").

   Such an approach is by no means unprecedented. Under the
Black Lung Act, for example, we have acknowledged that
legal pneumoconiosis encompasses a more inclusive category
of ailments than the purely clinical diagnosis, with the result
that a claimant’s entitlement to benefits is evaluated under the
broader definition. See Clinchfield Coal Co. v. Fuller, 

, 625 (4th Cir. 1999). When the presence of legal —
as opposed to clinical or medical — pneumoconiosis is at
issue in a black lung case, eligibility for benefits ultimately
depends on whether the disease has caused a worsening of
symptoms to the extent that the miner’s respiratory or pulmo-
nary functioning has been impaired. See id.3
  3
   In a similar vein, Congress has codified the common law rule that
insanity is an affirmative defense to federal criminal prosecution. See 18
                      UNITED STATES v. CAPORALE                           15
   The analogy to the Black Lung Act is especially apt here,
in light of the government experts having testified without
contradiction that mental infirmities in general (and paraphilia
in particular) have little meaning except to the extent they are
manifested by an impairment in the subject’s ability to func-
tion as a normal human being. In large part for that reason, the
court in Carta had no difficulty concluding that hebephilia is
a qualifying illness, abnormality, or disorder within the mean-
ing of § 4247(a)(6), observing that the respondent "has a
decades-long sexual fixation on minors that plainly has
caused significant distress or impairment in his 

(internal quotation marks omitted).

   The First Circuit in Carta made short shrift of the lower
court’s rejection of hebephilia as fulfilling the serious-
impairment prong, reasoning that the condition also fits
within the rubric of paraphilia/NOS, which "is expressly a
DSM-listed 

. In the case at bar, Dr.
Plaud’s testimony cast some doubt that hebephilia may so
qualify, but the scope of "illness, abnormality, or disorder" in
§ 4247(a)(6) is certainly broad enough to include hebephilia,
by its own or any other name. The district court’s determina-
tion to the contrary was thus incorrect as a matter of law.
Though his condition may elude definitive labeling, the evi-
dence at hearing established beyond question that Caporale’s
ability to function normally in society has been preempted by
his sexual fixation on underage, pubescent boys, such fixation
having heretofore so dominated his psyche as to substantially
impair and disrupt his life.4

U.S.C. § 17. The accused proves his insanity "at the time of the commis-
sion of the acts constituting the offense" by demonstrating that he "was
unable to appreciate the nature and quality or the wrongfulness of his
acts," in that he suffered from "a severe mental disease or defect." 

defense remains available even though "insanity" is "a legal
rather than a medical term." Dorland’s Illustrated Medical Dictionary 957
(31st ed. 2007).
   4
     In Carta, the court of appeals avoided the issue of whether hebephilia
per se is within the ambit of § 4247(a)(6), instead holding that the underly-
16                    UNITED STATES v. CAPORALE
                                    2.

   The substantial delays occasioned by the constitutional liti-
gation concerning the Walsh Act, 

Part I.B, resulted
in a backlog of commitment proceedings in the district courts.
Those proceedings now appear to be moving forward in ear-
nest, and, as a consequence, Caporale’s is already the fourth
merits-based appeal of a § 4245 judgment that we have con-
sidered, all within the past year. The three previous appeals
having been disposed of via published opinions, and each
having required an analysis of the hearing evidence under the
serious-difficulty prong of § 4247(a)(6), a review of those
prior decisions is in order.

                                    a.

   In United States v. Hall, 

(4th Cir. 2012), we
affirmed the district court’s ruling in favor of the respondent
on the serious-difficulty prong. The diagnosis in Hall was
pedophilia, and it was undisputed. The government’s two
experts, including Dr. Demby, based their serious-difficulty
conclusions primarily on actuarial tools, the respondent’s
ongoing interest in collecting pornographic drawings and pho-
tographs, and his general noncompliance with rules and direc-
tions while on supervised release and even following his
certification. See 

condition qualifies under the statute as a DSM-listed paraphilia/NOS.
Here, with Dr. Zinik and Dr. Plaud being sharply at odds concerning the
proper scope of a paraphilia/NOS diagnosis, 

at 8-10, we think
the better course is to avoid that question and simply hold that hebephilia,
as colloquially understood, is a § 4247(a)(6) "illness, abnormality, or dis-
order." Our holding does not mean, however, that a district court may find
the first prong of the second element of § 4247(a)(5) to be satisfied simply
because an expert labels the respondent a hebephile. A diagnosis of
hebephilia (or pedophilia, or other mental disorder) is merely the starting
point for the court to consider the true thrust of the § 4247(a)(6) inquiry
— whether, on a case-specific basis, the respondent’s underlying condi-
tion constitutes a serious functional impairment.
                  UNITED STATES v. CAPORALE                   17
   The defense expert, in turn, stressed the respondent’s lim-
ited history of offenses and victims, his participation in a
treatment program, and his abstention from hands-on offenses
during his twenty-eight months of release. See 

. The respondent’s expert pointed to recent research
linking possession of pornography with a lower recidivism
rate, adding that the stringent release provisions and possibil-
ity of a life sentence for further offenses would serve as pow-
erful deterrents. See 

district court found the defense expert to be the most
credible of the three, discerning that he had more completely
discussed the actuarial tools and psychological tests adminis-
tered to evaluate the respondent. See 

.
These tools assessed the respondent’s recidivism risk as being
between ten to thirty-six percent within ten years following
release. See 

court relied on the defense
expert’s testimony regarding the respondent’s control on his
prior release and the deterrent effect of the conditions
attached to the proposed release. See 

ment took issue with the determinative weight afforded the
respondent’s lack of offense conduct when released previ-
ously, we declined to assign clear error, deferring to the dis-
trict court’s analysis and its credibility assessments, including
that of the respondent himself, who indicated remorse and an
understanding of the risks of recidivism. See 

long thereafter, in United States v. Francis, 

(4th Cir. 2012), we again affirmed the district court’s
judgment in favor of the respondent. In Francis, a govern-
ment expert diagnosed the respondent with paraphilia/NOS
(telephone scatalogia), i.e., the propensity to make obscene
and threatening phone calls; another diagnosed sexual sadism
and antisocial personality disorder. 

supporting
the government’s case on the serious-difficulty prong, these
two experts relied heavily on actuarial tools and, to a lesser
extent, on the respondent’s perceived hostility toward women
and his noncompliance with supervised release. 

STATES v. CAPORALE
respondent’s risk of recidivism over a five-year period was
actuarially projected to be between approximately thirty and
sixty percent. 

for the respondent, including Dr. Plaud,
either agreed with or did not seriously dispute the govern-
ment’s diagnosis of paraphilia/NOS (telephone scatalogia).
The defense instead primarily contested the serious-difficulty
prong, urging the court to consider the respondent’s absten-
tion from reoffending during the seven months of his prior
release and contending that the actuarial tools employed by
the government’s experts were of no use to predict non-
physical behavior. See 

. Employing
his own methodology, one of the defense experts testified that
the respondent actually posed a low risk of recidivism. See 

evidence, the district court assumed with-
out deciding that the government had established the first,
prior-conduct element. See 

-73. No
clear finding was made with respect to the serious-impairment
prong of the second element, but the circumstances attendant
to the court’s discussion required us to assume that it had
implicitly resolved the issue in favor of the respondent. See 

that as it may, the court explicitly ruled in favor of
the respondent on the serious-difficulty prong, according
greater weight to the testimony of the defense experts. The
court noted particularly those experts’ evaluation of the
respondent’s behavior on prior release, and it credited their
opinions that the government’s actuarial evidence was of
scant probative value. 

appeal, we concluded
that the court’s analysis in that regard was sufficiently cogent
and thorough. 

recently, in United States v. Wooden, 

(4th Cir. 2012), we reversed the district court’s determination
in favor of the respondent on both second-element prongs,
and we remanded the matter for reconsideration. Like Hall,
Wooden involved a diagnosis of pedophilia, but on this latter
                  UNITED STATES v. CAPORALE                  19
occasion we observed that the court had improperly relied on
the testimony of the defense expert, in that the expert could
not square his opinions with the undisputed facts, particularly
that the respondent: (1) was continuing to have intense fanta-
sies and urges involving prepubescent children; (2) had
offered testimony that was littered with rationalizations, deni-
als, and other cognitive distortions or "thinking errors"; (3)
was having difficulty distinguishing between dreams and real-
ity with respect to an incident of molestation that was alleged
— but not proved — to have occurred in 2005 during his
parole period; (4) had admitted to attempting to molest
another child perhaps a year before that; (5) had sent a Christ-
mas card to the child in the alleged 2005 incident, in which
he promised that, if he were out of prison, "I would do a
whole lot for you"; and (6) had, while on parole, hired himself
out as a babysitter for young children. See 

, 452-56.

   We observed that the evidence set forth above was relevant
to both the serious-impairment and serious-difficulty prongs,

, and with especial regard to the latter, did
not support the defense expert’s fanciful theory that the
respondent’s tested reluctance to engage generally in impul-
sive behavior necessarily meant that he could exercise voli-
tional control of his deviant sexual urges. The expert’s
opinion, we noted, did not account for the documented exis-
tence of patient, calculating child molesters who groom their
victims. See 

the record revealed that,
beyond the laboratory context in which the expert had con-
ducted his evaluation, the respondent had in fact acted impul-
sively during his prior crime, and had also evidenced his
impulsiveness through his inability to adhere to institutional
rules while incarcerated. 

reasons offered by the defense expert to support his
opinions suffered from similar logical flaws. For example, the
expert cited the respondent’s disinterest in child pornography
as evidence of docility, yet neglected to acknowledge the utter
20                UNITED STATES v. CAPORALE
lack of evidence that the respondent had ever been so inter-
ested. See 

, 455. In addition, the
expert failed to account for the respondent’s attempted moles-
tation of one child in 2004, his correspondence thereafter with
another child via Christmas card, and his decision to place
himself in proximity to other children through babysitting, all
of which we explained "speaks directly to the serious-
difficulty prong." 

Other relevant consider-
ations included the respondent’s criminal history, see 

testimonial admissions, see 

any cogni-
tive limitations, see 

light of all the circumstances, we concluded that the dis-
trict court had clearly erred in relying upon the testimony of
an expert who "largely ignored all contradictory evidence"
and whose "analysis was internally inconsistent." 

. We also determined that the court committed
an error of law as to the government’s burden of proof by
demanding that it demonstrate, through its actuarial models,
not merely that the respondent was at high or very high risk
of reoffending within five years, but that the likelihood could
be quantified in excess of fifty percent. See 

456-57, 460-62.

                               b.

   The foregoing recitation illustrates the principle that "while
clear-error review is deferential, it is not toothless." 

(citation and internal quotation marks omit-
ted). We may discern clear error when a court "makes find-
ings without properly taking into account substantial evidence
to the contrary." 

).
In Wooden, we reiterated the familiar principle that such error
"occurs when a district court’s factual findings are against the
clear weight of the evidence considered as a whole." 

omitted).

  In support of reversing the judgment below in favor of
Caporale on the serious-difficulty prong, the government con-
                     UNITED STATES v. CAPORALE                          21
tends that the district court misplaced its reliance on the
expert testimony proffered by the defense, in that Dr. Plaud
failed to adequately consider: (1) one of the boys victimized
by Caporale’s 1992 offense, though seventeen at the time, had
been sexually involved with Caporale for four years; (2)
Caporale’s 1999 discussion of masturbation with a fourteen-
year-old boy constituted grooming behavior; and (3) Caporale
associated with a pedophile in 2003. The government also
suggests that the court’s resolution of the serious-difficulty
prong was fatally colored by its misanalysis of the serious-
impairment prong. Cf. 

(noting gov-
ernment’s argument "that the district court’s error in conclud-
ing that Wooden does not suffer from pedophilia prevented
the court from properly assessing the serious-difficulty ques-
tion").

   The defense case was by no means irrefutable, but, after
reviewing the entire record, it would be inaccurate for us to
say that Dr. Plaud was anything but thoroughly familiar with
Caporale’s history of molestation. Dr. Plaud readily acknowl-
edged that Caporale was interested in boys as young as twelve
or thirteen, but regarded Caporale’s involvement with older
minors in 1992 — though similarly unlawful and perhaps no
less reprehensible — as perhaps indicating an evolving inter-
est in those who have reached "an adult type of sexual matu-
rity," J.A. 266. On cross-examination, the government
specifically asked Dr. Plaud whether the 1992 incident justi-
fied enhancing his previous estimate of Caporale’s recidivism
risk. Dr. Plaud responded, "[I]t argues quite to the contrary."

     Insofar as the government contends that Dr. Plaud may have initially
misperceived Caporale’s latest "contact offense" against younger males to
have occurred in 1984, rather than in 1988 or 1989 at the outset of Capor-
ale’s relationship with the male involved in the 1992 incident, Dr. Plaud’s
testimony on cross-examination reveals that he had become fully apprised,
yet had not changed his opinion. See J.A. 265-66. Indeed, Dr. Plaud had
testified with reasonable accuracy on direct examination that it was "up to
22                    UNITED STATES v. CAPORALE
    Although the government is correct that the district court
did not address Caporale’s supposed grooming conduct in
1999 or his prohibited association with a pedophile in 2003,
it is also the case that neither of those episodes featured prom-
inently in the experts’ testimony. Dr. Zinik characterized the
1999 incident as grooming in his answer to a single question
on the second round of cross-examination conducted by
Caporale’s counsel, see J.A. 194-95, and, apart from the most
ephemeral of allusions as part of a question posed to Dr.
Demby on the subject of treatment, see 

2003
incident was not touched upon at the hearing, but discussed
only in a single paragraph of Dr. Zinik’s written report, see

the court’s analysis would have been more
comprehensive had it included a discussion of each secondary
point, these particular omissions amount to little more than
No. 8 needles in a stack of alfalfa.

   The battleground at hearing on the serious-difficulty prong
instead proceeded on two fronts: the probative value of the
government’s actuarial evidence, and the meaning to be
accorded the pornographic materials seized from Caporale’s
cell during May 2011. The parties agreed that it was impossi-
ble to predict with exactitude whether Caporale would reof-
fend, but that his profile indicated a risk ceiling of about
twenty-five percent, which would tend to decrease as he grew
older.6 Under the circumstances, it was not unreasonable for
the district court to credit Dr. Plaud’s opinion that the actuar-
ial evidence militated toward a conclusion in favor of Capor-
ale. Given our holding in Wooden that a recidivism risk

maybe about 1990 or so that [Caporale] was engaged in any type of
contact-based offending. That’s 20 or more years ago." 

court clearly understood the testimony, relating that Dr. Plaud "relied on
the fact that Mr. Caporale has not had a contact offense in roughly two
decades." Findings 8.
   6
     Of course, just as milk does not necessarily go sour at the date printed
on the carton, it cannot be assumed that "every sex offender magically
becomes safe at midnight on his 60th birthday." J.A. 184.
                  UNITED STATES v. CAPORALE                   23
ceiling of fifty percent did not rule out the respondent’s com-
mitment, 

, the court may well have rea-
sonably reached the opposite conclusion. But it was not
required to.

   Similarly, the evidence of Caporale’s recent activities cut
both ways. It is hardly surprising that Caporale’s continued
predilection for written pornography and suggestive photo-
graphs depicting teenage boys would generate some skepti-
cism among the government’s experts regarding his volitional
control, though it might be argued as more relevant to the
serious-impairment prong. At the same time, that no such
materials had been found in Caporale’s possession during the
ten months prior to hearing plausibly supports — albeit to a
limited degree — an inference of control. See 

(relating with approval district court’s character-
ization of respondent’s conformance with the law for more
than seven months on release as "compelling evidence" of his
volitional control); 

-66 (rejecting gov-
ernment’s contention that district court clearly erred by credit-
ing respondent’s lawful behavior on release for twenty-eight
months as probative of his capacity for control). The failure
to discover pornographic contraband in Caporale’s cell after
May 2011 might also have supported the inference that he
became better at concealing it, but it was manifestly the task
of the factfinder to choose between competing reasonable
inferences.

   The district court’s analysis is not immune from attack,
however, most notably its preference for Dr. Plaud’s testi-
mony based on the perception that his opinion on the serious-
difficulty prong considered Caporale’s current state more
thoroughly than the opposing experts. See Findings 9-10. The
court’s emphasis on current information was somewhat of a
straw man, inasmuch as the government’s experts were not
permitted access to Caporale and were therefore constrained
to rely solely on their forensic reviews of the record. More-
over, we do not share the court’s apparent disdain for histori-
24                   UNITED STATES v. CAPORALE
cal evidence, having observed that a respondent’s criminal
record "may well be a historical factor, but it is by no means
a stale or irrelevant one. When the question is whether an
inmate . . . will have serious difficulty refraining from re-
offending if released, consideration of the nature of his prior
crimes provides a critical part of the answer." 

   Lastly, although the district court did not specifically indi-
cate any reliance upon them, other aspects of the case support
the underlying judgment. For instance, Caporale’s deposition
testimony evidenced a normal intelligence, thus allaying con-
cerns that cognitive impairments could potentially "affect [the
respondent’s] ability to refrain from re-offending if released."

. Caporale expressed remorse for his
past behavior, and he demonstrated the ability to differentiate
between right and wrong. See 

. Caporale
also testified that the counseling and treatment he had
received for his condition had prepared him for the rigors of
release, enabling him to recognize and avoid high-risk cir-
cumstances that might otherwise tempt him to engage in
wrongdoing. See 

strength of the parties’ evidence and the
logical force of the district court’s analysis paint a picture that
is more reminiscent of Hall and Francis than it is of Wooden.
Though the government attempts to nitpick Dr. Plaud’s opin-
ions, his testimony, by and large, told "‘a coherent and
facially plausible story that is not contradicted by extrinsic
evidence.’" 

(quoting Anderson v.
   7
     For a court assessing the likelihood of recidivism upon a respondent’s
proposed release among the general public, an examination of the respon-
dent’s criminal conduct undertaken in the same setting — though having
occurred a number of years past — may be the most probative predictor.
It should be noted that Caporale has been incarcerated since 1992, except
for about four years between 1998 and 2000, and again between 2001 and
2003.
                      UNITED STATES v. CAPORALE                           25
Bessemer City, 

, 573 (1985)). It is therefore
unlikely that the court clearly erred in relying upon it.8

                                     B.

   Although the district court incorrectly ruled that impair-
ment through hebephilia cannot be, as a matter of law, a seri-
ous mental illness, abnormality, or disorder within the
meaning of the Walsh Act, it did not commit clear error by
alternatively crediting Dr. Plaud’s testimony that Caporale
possesses sufficient volitional control to prevent him from
having serious difficulty refraining from sexually violent con-
duct or child molestation in the future. Under the circum-
stances, we are not left with the definite and firm conviction
that the court was mistaken in concluding that the government
had failed to carry its burden of proving Caporale a sexually
dangerous person by clear and convincing evidence.

                                    IV.

   Pursuant to the foregoing, we affirm the judgment of the
district court.

                                                             AFFIRMED
   8
     We attribute little significance to the government’s contention that the
district court’s legal misstep with regard to the serious-impairment prong
fatally poisoned its analysis of the serious-difficulty prong. We hinted at
such a possibility in Wooden, but in that case, we traced the court’s legal
error to the permissible inferences and conclusions drawn from the evi-
dence, rather than to the proper construction of statutory language. In
addition, there was a great deal of evidentiary overlap in Wooden between
the § 4247(a)(6) prongs, and less so in Caporale’s case. Courts routinely
offer alternative holdings in support of their judgments, a practice that
facilitates review and promotes judicial efficiency. In order to encourage
that beneficial practice, we should tend to refrain from placing disincen-
tives in its path and be content, in the usual instance, to compartmentalize
our review.