RECOMMENDED FOR FULL-TEXT PUBLICATION
                        Pursuant to Sixth Circuit I.O.P. 32.1(b)
                               File Name: 14a0040p.06

            UNITED STATES COURT OF APPEALS
                           FOR THE SIXTH CIRCUIT
                             _________________


UNITED STATES OF AMERICA,                       ┐
                        Plaintiff-Appellee,     │
      v.                                        │
                                                │       No. 13-5288
                                                │
DARNELL MITCHELL,                               >
                                                │
                        Defendant-Appellant.    │
                                                ┘

                  Appeal from the United States District Court
           No. 2:10-cr-20299-1—Jon Phipps McCalla, District Judge.
                            Argued: January 22, 2014
                      Decided and Filed: February 28, 2014

     Before: SUHRHEINRICH, SILER, and KETHLEDGE, Circuit Judges.


                               _________________
                                   COUNSEL

ARGUED: David M. Bell, FEDERAL PUBLIC DEFENDER, Memphis,
Tennessee, for Appellant. Daniel T. French, UNITED STATES ATTORNEY’S
OFFICE, Memphis, Tennessee, for Appellee. ON BRIEF: David M. Bell,
FEDERAL PUBLIC DEFENDER, Memphis, Tennessee, for Appellant. Daniel T.
French, UNITED STATES ATTORNEY’S OFFICE, Memphis, Tennessee, for
Appellee.




                                         1
13-5288        United States v. Mitchell                                     Page 2


                               _________________

                                    OPINION
                               _________________



       SILER, Circuit Judge.     Defendant Darnell Mitchell appeals the district
court’s determination that his prior Tennessee state robbery convictions are violent
felonies under the Armed Career Criminal Act (“ACCA”), 18 U.S.C. § 924(e). For
the following reasons, we AFFIRM.

            I. FACTUAL AND PROCEDURAL BACKGROUND

       Mitchell was convicted in 2012 on one count of being a felon in possession
of a firearm in violation of 18 U.S.C. § 922(g)(1). A Presentence Investigation
Report (“PSR”) was completed that classified Mitchell as an “armed career
criminal” on account of having committed three violent felonies:        (1) a 1988
conviction for robbery under now-repealed Tenn. Code Ann. § 39-2-501(a); (2) a
2003 conviction for robbery under Tenn. Code Ann. § 39-13-401; and (3) a 2003
conviction for intentionally evading arrest in an automobile. The PSR assigned
Mitchell a total offense level of 34 and a criminal history category of VI, which
resulted in an applicable Guidelines range of 262 to 327 months’ imprisonment.
Mitchell objected to his armed career criminal classification, arguing that his two
robbery convictions did not qualify as “violent felonies.”       The district court
disagreed and sentenced Mitchell to 300 months’ imprisonment.

                            II. STANDARD OF REVIEW

       We review de novo a district court’s determination that an offense
constitutes a “violent felony” under the ACCA. United States v. Benton, 

, 729 (6th Cir. 2011).
13-5288         United States v. Mitchell                                       Page 3


                                 III. DISCUSSION

       Under the ACCA, a defendant who violates 18 U.S.C. § 922(g) and has
three prior convictions for serious drug offenses or violent felonies must receive a
fifteen-year mandatory minimum sentence. United States v. Johnson, 

, 658 (6th Cir. 2013) (citing 18 U.S.C. § 924(e)(1)). The ACCA defines
“violent felony” as:

       any crime punishable by imprisonment for a term exceeding one
       year, or any act of juvenile delinquency involving the use or
       carrying of a firearm, knife, or destructive device that would be
       punishable by imprisonment for such term if committed by an adult,
       that—
       (i) has as an element the use, attempted use, or threatened use of
       physical force against the person of another; or
       (ii) is burglary, arson, or extortion, involves use of explosives, or
       otherwise involves conduct that presents a serious potential risk of
       physical injury to another[.]


18 U.S.C. § 924(e)(2)(B). We refer to § 924(e)(2)(B)(i) as the “use of physical
force” clause and the portion of § 924(e)(2)(B)(ii) following the enumerated
offenses as the “residual clause.”

1. Application of the Categorical Approach

       In determining whether a particular offense qualifies as a violent felony,
courts must use the “categorical approach.” Taylor v. United States, 

,
600, 602 (1990). They must look only to the fact of conviction and the statutory
definition of the prior offense and not the particular facts underlying that
conviction. James v. United States, 

, 202 (2007); see also United
States v. Bartee, 

, 359 (6th Cir. 2008). This approach “avoid[s] the
practical difficulties and potential unfairness of permitting a sentencing court to
relitigate facts and delve into the details of a prior conviction.” 

(citing United States v. Armstead, 

, 947 (6th Cir. 2006)).
13-5288          United States v. Mitchell                                     Page 4


       At issue are Mitchell’s 1988 and 2003 robbery convictions under Tennessee
law. In 1988, Tennessee law defined robbery as “the felonious and forcible taking
from the person of another, goods or money of any value, by violence or putting the
person in fear.” Tenn. Code Ann. § 39-2-501(a) (1982) (repealed). The version in
effect in 2003 differed only slightly, defining robbery as the “intentional or
knowing theft of property from the person of another by violence or putting the
person in fear.” Tenn. Code Ann. § 39-13-401.

       A. Violent Felony under the “Use of Physical Force” Clause

       As    a   threshold   matter,   “[t]he   meaning   of   ‘physical   force’      in
§ 924(e)(2)(B)(i) is a question of federal law, not state law.” Johnson v. United
States, 

, 138 (2010). “Physical force” in turn is “force capable of
causing physical pain or injury to another person.” 

(citing Flores v.
Ashcroft, 

, 672 (7th Cir. 2003)).          However, in determining the
meaning or scope of robbery, we are bound by the Tennessee Supreme Court’s
interpretation, including its guidance on the elements of the crime. See 

       Robbery in Tennessee involves the “felonious” (under the former language)
or “intentional” (under the present language) taking of property from the person of
another “by violence or putting the person in fear.” The Tennessee Supreme Court
has applied the plain meaning of “violence,” as expressed in both versions of the
statute, as “physical force unlawfully exercised so as to injure, damage or abuse.”
State v. Fitz, 

, 214 (Tenn. 2000). The element of violence contained
in the robbery statute thus satisfies § 924(e)(2)(B)(i)’s requirement of the “use,
attempted use, or threatened use of physical force.”

       Mitchell relies on United States v. Fraker for the proposition that robbery
“includes criminal activity outside the scope of the ACCA because a defendant can
violate the statute by employing only fear, rather than physical violence or force.”
458 F. App’x 461, 463 (6th Cir. 2012) (unpublished). The panel in Fraker turned
to the colloquial understanding of fear, as opposed to the meaning of fear as
interpreted by the Tennessee Supreme Court. When evaluated in the latter context,
13-5288         United States v. Mitchell                                      Page 5


there is no doubt that a taking of property carried out through fear, for purposes of
Tenn. Code Ann. §§ 39-2-501(a) and 39-13-401, qualifies as a “violent felony”
under the ACCA. This is because, as the Tennessee Supreme Court has held, the
“fear constituting an element of robbery is a fear of bodily injury and of present
personal peril from violence offered or impending.” State v. Taylor, 

, 398 (Tenn. 1989).       Applying the term “violence” derived in Fitz, the
Tennessee robbery statutes refer to a “fear of bodily injury and of present personal
peril from physical force offered or impending.” The word “impending” itself
means “imminent; threatening.” Webster’s Revised Unabridged Dictionary 734
(1913). Therefore, the commission of a robbery through fear, which in Tennessee
reduces to the fear of bodily injury from physical force offered or impending,
directly corresponds to § 924(e)(2)(B)(i)’s “use . . . or threatened use of physical
force.” Moreover, it is clear that the harm characterized in the Tennessee statutes is
of a physical nature, rather than emotional harm or distress.

       Our finding is consistent with the manner in which the Seventh Circuit has
interpreted an analogous Indiana statute in United States v. Lewis, 

(7th Cir. 2005). There, the court was confronted with whether a conviction for
robbery, in violation of Ind. Code 35-42-5-1, is a “crime of violence” (as defined in
USSG § 4B1.2) for purposes of a sentencing enhancement received under
§ 2K2.1(a)(4)(A) of the Guidelines. 

“Whether a conviction is a
‘violent felony’ under § 924(e) is analyzed the same way as whether a conviction is
a ‘crime of violence’ under USSG § 4B1.2.” United States v. Perry, 

,
910 (6th Cir. 2013). In Indiana, robbery entails taking property from the person of
another by “using or threatening the use of force on any person” or “putting any
person in fear.” 

. The court went on to explain that the fear
in Indiana’s formulation “is fear of physical injury rather than of defamation” and
had no difficulty in reaching the conclusion that robbery “is readily classified as a
crime of violence.” 

panel neglected to assign the meaning of each element of
robbery as construed by the Tennessee Supreme Court. In his reply brief, Mitchell
13-5288         United States v. Mitchell                                       Page 6


cites to United States v. Johnson, 530 F. App’x 528 (6th Cir. 2013) (unpublished), a
case in which the panel adopted the reasoning of Fraker and declined to hold that a
conviction under Tenn. Code Ann. § 39-2-501 categorically is a violent felony.
Like our criticism of Fraker, the Johnson panel failed to apply the definition of
“fear” enunciated by the Tennessee Supreme Court. See 

Neither this
court “nor any other federal tribunal has any authority to place a construction on a
state statute different from the one rendered by the highest court of the State.”
Johnson v. Fankell, 

, 916 (1997). Contrary to Fraker and Johnson,
robbery in violation of either Tenn. Code Ann. § 39-2-501 or § 39-13-401 is
categorically a “violent felony” under § 924(e)(2)(B)(i) of the ACCA.

       B. Violent Felony under the Residual Clause

       We also evaluate whether Mitchell’s convictions for robbery under
Tennessee law categorically qualify as “violent felonies” under the residual clause
of the ACCA. Prior to Begay v. United States, 

(2008), “the only
relevant question was whether the offense posed a serious risk of physical injury to
others.” United States v. Vanhook, 

, 710 (6th Cir. 2011) (citing

). In Begay, the Supreme Court recognized this question
as only the initial inquiry; “[f]or a crime to qualify as a violent felony, it must also
be similar ‘in kind as well as in degree of risk posed’ to the crimes Congress
specifically [enumerated in § 924(e)(2)(B)(ii)].” 

553 U.S. at
143). In holding that the crime of driving under the influence (“DUI”) is not an
ACCA predicate, the Begay Court stated that DUI is not “purposeful, violent, and

. However, the Supreme Court’s decision in Sykes
v. United States distinguished Begay as involving “a crime akin to strict liability,
negligence, and recklessness crimes.” 

, 2276 (2011). Moreover,
Sykes clarified that a comparison of “the risk posed by [the crime in question] . . .
to that posed by its closest analog among the enumerated offenses” provides “a
categorical and manageable standard” relative to the “purposeful, violent, and
aggressive” phrase, which bears “no precise textual link to the residual clause.” 

2275-76. Nevertheless, “[i]n many cases the purposeful, violent, and
13-5288         United States v. Mitchell                                       Page 7


aggressive inquiry will be redundant with the inquiry into risk, for crimes that fall
within the former formulation and those that present serious potential risks of
physical injury to others tend to be one and the same.” 

       The first layer of inquiry is directed at whether robbery, as characterized by
the Tennessee statutes at issue here, presents a serious potential risk of physical
injury to another. In the context of larceny from the person, which is a lesser
included offense of robbery, we have concluded that “stealing from the person of
another” poses a serious risk of injury to others, thus satisfying the initial inquiry.
See United States v. Taylor, 

, 631-32 (6th Cir. 2012) (citing United
States v. Payne, 

, 375 (6th Cir. 1998)). In arriving at this conclusion,
we observed that taking property “from the possession of the victim” is “clearly the
type of situation that could result in violence”; victims of “such an invasion of
personal space would likely resist or defend in a manner that could lead to
immediate violence”; violence or harm need not actually result from the theft
because the residual clause requires only “a serious potential risk of physical
injury,” and “no situation [can be anticipated] in which larceny from the person
could occur without presenting” such a risk. 

163 F.3d at 375).
As explained by the Tennessee Supreme Court, “[r]obbery is a species of larceny,
involving the same elements and turpitude, aggravated by taking from the person
by open violence, or putting in fear.” Hammond v. State, 43 Tenn. (3 Cold.) 129,
134 (1866). Like larceny, robbery requires theft from the person of another, but
because robbery additionally requires violence actually dispensed, or putting the
victim in fear from violence offered or impending, the crime presents a risk of
physical injury that is even greater than that associated with larceny. As the
Tennessee Supreme Court emphasized in State ex rel Anderson v. Winsett,

, 743 (Tenn. 1965), “[i]t is violence that makes robbery an offense
of greater atrocity than larceny.”

       Before proceeding to the second layer of the comparative risk inquiry, in
which we consider whether robbery is “roughly similar, in kind as well as in degree
of risk posed” to the enumerated offenses, 

, we pause to
13-5288         United States v. Mitchell                                       Page 8


revisit the clarification provided in Sykes concerning usage of the “purposeful,
violent, and aggressive” standard. As recognized by the dissenting opinions in
Sykes, “[t]he Court now suggests” that “Begay’s ‘purposeful, violent, and
aggressive’ test . . . applies only to strict liability, negligence, and recklessness
crimes.” 

(Scalia, J., dissenting); see also 

n.1
(Kagan, J., dissenting) (“understand[ing] the majority to retain the ‘purposeful,
violent, and aggressive’ test, but to conclude that it is ‘redundant’ in this case.”).
Assuming the “purposeful, violent, and aggressive” standard is relevant only where
the prior felony lacks “a stringent mens rea requirement,” 

both versions
of the Tennessee robbery statute would be ineligible for review under the standard.
It is clear that Tennessee’s current robbery statute, which was in place at the time of
Mitchell’s 2003 conviction, involves an “intentional or knowing theft.”
Tennessee’s prior formulation, however, specified a “felonious and forcible
taking.” Despite the slight difference in language, this earlier version of robbery
nonetheless is a crime of intent. See Young v. State, 

, 307 (Tenn.
1972) (finding that robbery requires the specific intent of permanently depriving
the owner of the property taken; therefore, it is in the classification of specific
intent crimes); Dillard v. State, 50 Tenn. (3 Heisk.) 260, 261 (1871) (indicating that
the word “feloniously” means “with intent to commit a felony”); Hammond,
43 Tenn. (3 Cold.) at 134 (“The same design and malicious purpose or intent
necessary in [larceny] must be present in [robbery].”).

       Separately, Justice Kagan’s dissent in Sykes raised the point that to limit the
“purposeful, violent, and aggressive” test to strict liability, negligence, and
recklessness crimes “would collide with Chambers v. United States, 

(2009)—a decision the majority cites approvingly—which applied the test to an
intentional crime.” 

n.1 (Kagan, J., dissenting). Even if
the test extended to robbery, the convictions at issue here for violations of Tenn.
Code Ann. §§ 39-2-501(a) and 39-13-401 are sufficiently “purposeful, violent, and
aggressive” to categorically qualify as “violent felonies” under the residual clause
of the ACCA. Purposeful conduct “require[s] that the defendant have acted with []
intent to cause such injury.” 

(citing 

        United States v. Mitchell                                       Page 9


145). As previously discussed, robbery in Tennessee, whether under the current or
prior statutory formulation, is a crime of intent and echoes the “deliberate kind of
behavior” reflected in the violent felonies enumerated in § 924(e)(2)(B)(ii) of the
ACCA. 

. Furthermore, robbery is the type of violent and
aggressive crime generally characterized as a violent felony.              Robbery in
Tennessee, by definition, is a violent form of theft. See State v. Swift, 

, 830 (Tenn. 2010) (“The use of violence or fear elevates theft to robbery.”);
State v. Owens, 

, 638 (Tenn. 2000) (“[W]hether a taking is properly
characterized as a theft or a robbery is contingent upon whether and when violence
or fear is imposed.”); Hammond, 43 Tenn. (3 Cold.) at 134 (“Robbery is a species
of larceny . . . aggravated by taking from the person by open violence, or putting in
fear.”). Aggressive, violent acts are “aimed at other persons or property where
persons might be located and thereby injured.” 

(quoting
United States v. Archer, 

, 1351 (11th Cir. 2008)). Because both
versions of the Tennessee robbery statute specify theft “from the person” as an
element of the crime, robbery necessarily is a crime that occurs not only “where [a]
person[] might be located,” but where the presence of another is assured with
absolute certainty.   Although strict adherence to the “purposeful, violent, and
aggressive” rubric is not definitively required, and although its future is in question,
robbery nonetheless categorically qualifies as a “violent felony” under this test.

       Continuing to the second part of the comparative risk analysis, we evaluate
whether robbery is “roughly similar, in kind as well as in degree of risk posed,” to
the enumerated offenses in § 924(e)(2)(B)(ii). As for similarity with respect to the
degree of risk posed, the offense compares favorably with generic burglary. The
Supreme Court in Sykes explained that burglary is risky for the reason that it can
end in confrontation leading to 

; see also 

(“The fact that an offender enters a building to commit a crime
often creates the possibility of a violent confrontation between the offender and an
occupant, caretaker, or some other person who comes to investigate.”). Robbery
presents an even greater degree of risk of physical injury relative to burglary, for
burglary does not require that the building or structure be occupied when burgled,
13-5288            United States v. Mitchell                                 Page 10


see 

, 597, while robbery requires another’s presence.
Whereas burglary raises the mere possibility of confrontation, robbery is defined as
to guarantee it.

        Robbery is similar also “in kind,” or in the “way or manner” it produces the
risk of injury, to the enumerated offenses for the following reason: it is not a crime
“akin to strict liability, negligence, and recklessness crimes.” 

(quoting 

). It requires intentional conduct. See 

. The Tennessee robberies for which Mitchell was convicted
categorically qualify as “violent felonies” under the residual clause of the ACCA.
Our view is consistent with decisions from other circuits that have addressed
robbery statutes post-Begay. See United States v. Welch, 

, 1311
n.30, 1312 (11th Cir. 2012) (finding that “robbery by sudden snatching”—which
means the taking of property “from the victim’s person, with intent to permanently
or temporarily deprive . . . when, in the course of the taking, the victim was or
became aware of the taking”—qualifies as a violent felony under the ACCA’s
residual clause); United States v. Carmichael, 408 F. App’x 769, 771 (4th Cir.
2011) (holding that common law robbery under North Carolina law, which is
defined as “the felonious, non-consensual taking of money or personal property
from the person or presence of another by means of violence or fear,” qualifies as a
violent felony under § 924(e)(2)(B)(ii)’s residual clause); United States v. Gore,

, 744 (5th Cir. 2011) (Higginbotham, J., concurring) (indicating that,
“as defined by Texas law . . . robbery [is a] violent felon[y] under the [ACCA’s]
residual clause”) (citing United States v. Davis, 

, 287 (5th Cir.2007));

(holding that robbery under Indiana law, which involves
taking property from the person of another by force or putting a person in fear, is a
“crime of violence” under Guideline § 4B1.2’s “use of physical force” clause and
residual clause).

        C. Role of the Modified Categorical Approach

        Mitchell advanced a new argument in his reply brief based on the Supreme
Court’s recent holding in Descamps v. United States, 

(2013), a case
13-5288          United States v. Mitchell                                   Page 11


which clarifies the analytical approach that sentencing courts must use to determine
if a prior conviction is a predicate offense under the ACCA.           Mitchell now
contends that the Tennessee robbery statute is an indivisible statute that defines the
crime of robbery overbroadly, and therefore cannot qualify as an ACCA predicate
under Descamps. According to Mitchell, because “review of plea colloquies or
other approved extra-statutory documents” is appropriate “only where a statute
defines robbery alternatively, not overbroadly,” the “modified [categorical]
approach had no role to play in the case.” He misreads the distinction between
“divisible” and “indivisible” mapped out in Descamps.

       “[W]hen a prior conviction is for violating a so-called ‘divisible statute,’”
which “sets out one or more elements of the offense in the alternative,” we may
resort to a variant of the categorical approach “labeled (not very inventively) the
‘modified categorical approach.’” 

. The modified
categorical approach is a “tool” used in a “narrow range of cases” to identify which
element formed the basis of a defendant’s conviction when a statute bears multiple
alternative elements. 

A court is therefore permitted to “go beyond the
mere fact of conviction[,]” 

, and may “consult a limited
class of documents” to determine which alternative in a divisible statute gave rise
to a conviction, and to compare that conviction to the offense under the ACCA.

. Where the prior conviction resulted from a guilty
plea, we look to documents that identify what facts the defendant “necessarily
admitted” by pleading guilty. United States v. Medina-Almaguer, 

,
423 (6th Cir. 2009). Such documents may include the “charging document, written
plea agreement, transcript of plea colloquy, and any explicit factual finding by the
trial judge to which the defendant assented.” Shepard v. United States, 

, 16 (2005).     On the other hand, if a statute is indivisible, it presents no
alternatives and the inquiry ends. There is simply no reason to turn to the modified
categorical approach. 

.

       A divisible statute is necessary but not sufficient for application of the
modified categorical approach.      Divisibility must be accompanied by another
13-5288         United States v. Mitchell                                   Page 12


condition, namely, ambiguity in the statute of conviction. In the context of the
ACCA, ambiguity results when “an individual can violate the statute [of
conviction] in a way that constitutes a violent felony and in a way that does not.”

. A crime can meet the definition of “violent felony” in
any of three ways: (1) the crime falls under the “use of physical force” clause;
(2) the crime is one of the statute’s enumerated offenses—burglary, arson,
extortion, or a crime involving the use of explosives; or (3) the crime falls under
the residual clause because it poses a serious potential risk of physical injury to
another. § 924(e)(2)(B). In determining the presence or absence of ambiguity, we
compare the statute of conviction to the relevant offense under the ACCA. For
convictions based on one of the enumerated offenses, e.g., burglary, the
enhancement provision is not triggered simply because the crime “happened to be
labeled . . . ‘burglary’ by the laws of the State of conviction.” 

. Rather, in drafting the ACCA, Congress was concerned with “crimes having
certain specified elements.” 

consequence, for convictions based on one of
the enumerated offenses, we compare the elements of the crime of conviction with
the “generic” crime, which has been described as the “generally accepted
contemporary meaning of [the] term,” 

and the “sense in which the term
is now used in the criminal codes of most States.” 

In this way, we
ensure that the comparison to the enumerated ACCA crime is anchored in a
“uniform definition independent of labels.” 

If the relevant statute has
the same elements as the generic ACCA crime or defines the offense more
narrowly, then the prior conviction can serve as an ACCA predicate. 

.

       A crime outside the enumerated offenses qualifies as a predicate offense if it
involves the “use of physical force” as outlined in § 924(e)(2)(B)(i) or poses a
serious potential risk of physical injury as outlined in the residual clause of
§ 924(e)(2)(B)(ii). These two aspects of § 924(e)(2)(B) are not mutually exclusive;
a non-enumerated crime may fit both the “use of physical force” clause and the
residual clause. We compare the elements of conviction for a non-enumerated
crime not to the generic, commonly understood definition of that crime—as doing
13-5288         United States v. Mitchell                                      Page 13


so would circumvent the ACCA entirely—but rather to the definitions set forth in
either the “use of physical force” clause or the residual clause, or both.

       In United States v. Covington, 

, 763 (6th Cir. 2014), we
recently commented, in light of Descamps, that “there are two steps in applying the
categorical approach to determine whether a prior conviction constitutes a crime of
violence under the Guidelines or a violent felony under the ACCA.” The first step
is to “ask whether the statute at issue is divisible.” 

“the court may look to
the Shepard documents—but only to see which alternative version of the offense is
at issue.” 

court must ask whether the offense the statute describes,
as a category, is a crime of violence” under USSG § 4B1.2(a)—or a violent felony
under 18 U.S.C. § 924(e)(2)(B)—which is an inquiry having “as many as three
parts” depending on whether the statute of conviction (1) has as an element “the
use, attempted use, or threatened use of physical force against the person of
another”; (2) matches the elements of one of the enumerated offenses; namely,
burglary, arson, extortion, or a crime involving the use of explosives; or
(3) involves “conduct that presents a serious potential risk of physical injury to
another.” 

       This distillation of Descamps may give the false impression that the
presence of a divisible statute of conviction alone confers on a court the ability to
turn to certain approved, extra-statutory documents. However, this is not so. The
Supreme Court in Descamps, in analyzing a defendant’s conviction for burglary,
repeated the caveat that “[o]ur decisions authorize review of the plea colloquy or
other approved extra-statutory documents only when a statute defines
burglary . . . alternatively, with one statutory phrase corresponding to the generic
crime and another 

(emphasis added).           “In that
circumstance, a court may look to the additional documents . . . .” 

see also 

(recognizing that when a statute “‘refer[s] to several
different crimes,’ not all of which qualify as an ACCA predicate, a court must
determine which crime formed the basis of the defendant’s conviction.” (quoting
Nijhawan v. Holder, 

, 35 (2009)).
13-5288           United States v. Mitchell                                               Page 14


        We make explicit a step in the analysis that the Covington panel alluded to
implicitly: if a statute is divisible—in that it sets out one or more elements of the
offense in the alternative—at least one, but not all of those alternative elements
must depart from: (1) the elements of the generic ACCA crime (if the conviction is
based on an enumerated offense); or (2) the definitions provided in either or both
the “use of physical force” clause and residual clause (if the conviction is based on
a non-enumerated offense).1 See 

(holding that where
“a statute lists multiple, alternative elements, and so effectively creates several
different crimes,” the modified categorical approach acts as a tool to identify the
crime of conviction “[i]f at least one, but not all of those crimes matches the
generic version . . . .”). The Covington panel, after concluding that Michigan’s
prison escape statute is divisible, recognized application of the modified categorical
approach was appropriate because “the statute lists several, alternative” crimes,
“some [of which] involve the element of breaking and some [of which] do not,”
despite not calling out this condition directly. 

.

        Before the Supreme Court in Descamps was a California burglary statute,
which provides that a “person who enters” certain locations “with intent to commit
grand or petit larceny or any felony is guilty of 

(quoting Cal. Penal Code Ann. § 459). California’s definition of burglary does not
require the entry to have been unlawful, in contrast to most burglary laws, which
generally demand breaking and entering or similar conduct.                         

burglary statute is indivisible in that it does “not contain[] alternative
elements,” 

there could be “no uncertainty” as to which element formed
the basis of the defendant’s conviction. 

As such, the categorical
approach “need[ed] no help from its modified partner.” 

California’s burglary statute “sweep[s] so widely” as to “criminalize[] a broader
swath of conduct” than the “generic” definition of burglary, 

the
Supreme Court held that a conviction under the California burglary statute could


        1
           The “either or both” language is meant to depend on whether a party argues the offense is
a violent felony under the “use of physical force” clause, the residual clause, or both.
13-5288         United States v. Mitchell                                      Page 15


not serve as an ACCA predicate. 

In this regard, both the categorical
and modified categorical approaches share the same analytical step:                   a
“compar[ison of] the elements of the crime of conviction . . . with the [definition]
of the [ ] crime [under the ACCA].” 

       The Tennessee robbery statutes at issue here criminalize the intentional
taking of property from the person of another “by violence or putting the person in
fear.” (emphasis added). The “ordinary use” of the conjunction “or” is “almost
always disjunctive, that is, the words it connects are to ‘be given separate
meanings.’” United States v. Woods, 

, 567 (2013) (quoting Reiter v.
Sonotone Corp., 

, 339 (1979)). While “or” can “sometimes introduce
an appositive[,] a word or phrase that is synonymous with what precedes it,” that is
not the case here. 

phrasing here suggests two alternative
elements. In Descamps, the Supreme Court treated as divisible an illustrative
burglary statute featuring “entry into a building or an 

(emphasis in original). So long as certain other conditions were met, an
individual could commit burglary by entering a building or by entering an
automobile. Similarly, Tenn. Code Ann. § 39-2-501(a) specifies robbery as a
“taking . . . by violence or putting . . . in fear.” This wording is nearly identical in
the current statute, which describes robbery as a “theft . . . by violence or putting
. . . in fear.” Tenn. Code Ann. § 39-13-401. Both forms of the robbery statute are
in line with the construction of the example burglary statute highlighted by the
Court as divisible.

       Reinforcing this conclusion is the Supreme Court’s guidance in 

. In that case, the Court considered a Florida state battery
statute, in which battery is defined as occurring “when a person either ‘1. [a]ctually
and intentionally touches or strikes another person against [his] will,’ or
‘2. [i]ntentionally causes bodily harm to another person.’” 

(quoting Fla.
Stat. § 784.03(1)(a)). There, the Court explained that “[b]ecause the elements of
the offense are disjunctive, the prosecution can prove a battery in one of three
ways.” 

Specifically, “[i]t can prove that the defendant
13-5288            United States v. Mitchell                                  Page 16


‘[i]ntentionally caus[ed] bodily harm,’ that he ‘intentionally str[uck]’ the victim, or
that he merely ‘[a]ctually and intentionally touche[d]’ the victim.” 

-37.
Rather than treating the first part of the battery statute as setting forth a single
element, the Court treated the language “touches or strikes” as comprising two
alternative elements.      The Court unequivocally confirmed that its decision in
Johnson “rested on the explicit premise that the law[] contained statutory phrases
that cover several different . . . crimes, not several different methods of committing
one offense.” 

n.2 (internal quotation marks and
brackets omitted). Similarly, the Tennessee robbery statutes at issue here, which
feature the language “by violence . . . or putting in fear,” enunciate two alternative
elements. Therefore, Mitchell’s recasting the statutes as indivisible and overbroad
is incorrect.

        Although the Tennessee robbery statutes are divisible, we need not defer to
the modified categorical approach to determine which alternative formed the basis
of Mitchell’s prior conviction. As already discussed, neither alternative element
departs from the definitions provided in the “use of physical force” clause or the
residual clause. Even if, as Mitchell contends, robbery under Tennessee law is
considered indivisible, the finding would have no bearing on our decision. An
indivisible statute merely limits the court to undertaking the formal categorical
approach; it presents “no uncertainty” as to which element formed the basis of the
prior conviction and, as such, “the categorical approach needs no help from its
modified partner.” 

A straightforward application of the categorical
approach has revealed that the robberies for which Mitchell was convicted are
categorically violent felonies, under both the “use of physical force” clause and the
residual clause.

        Lastly, Mitchell argued that the district court erred in using his 2003
robbery conviction as an ACCA predicate despite having entered an Alford plea,
where he claims to have only stipulated to the facts of a violent felony, but never
admitted guilt as to those facts. However, as Mitchell concedes, “an Alford-type
plea does not prevent the resulting conviction from ever serving as a predicate
13-5288         United States v. Mitchell                                      Page 17


conviction for sentence enhancement.” 

. “Convictions
based on Alford-type pleas can be predicate convictions under the ACCA if the
qualifying crime is inherent in the fact of the prior conviction—i.e., if the statute of
conviction is categorically a ‘violent felony’ . . . .” 

robbery under
Tennessee law is categorically a violent felony, Mitchell’s argument fails.

2.     Mitchell’s Characterization of his 1988 Conviction as an “Act of
       Juvenile Delinquency”
       The government argues that Mitchell never made the claim below that his
1988 conviction for robbery was an “act of juvenile delinquency” under the ACCA
and thus has forfeited his right to raise on appeal a challenge based on that
distinction. Mitchell, without support or explanation, treats his 1988 conviction for
robbery as an “act of juvenile delinquency” and argues, in light of the self-imposed
designation, that the government was required to show he used or carried a firearm,
knife, or explosive device during the crime. We assume, without deciding, that
Mitchell’s claim has not been forfeited and move on to the merits, reviewing de
novo the question of whether his 1988 adult conviction for robbery, committed at
the age of seventeen, qualifies as a predicate offense.

       Directly on point is our previous disposition in United States v. Banks,

(6th Cir. 2012). In that case, the defendant expressly objected to the
PSR’s classification of his adult conviction for robbery as a violent felony, which,
like Mitchell, was committed at the age of seventeen and charged under Tennessee
law. 

In light of § 924(e)(2)(B) employing the disjunctive “or,” we ruled
that “regardless of whether [the defendant’s] robbery qualifies as an ‘act of juvenile
delinquency,’ his adult conviction for that robbery falls under the definition of
‘violent felony’ as a ‘crime punishable by imprisonment for a term exceeding one
year.’” 

original) (citing Tenn. Code Ann. §§ 39-13-401(b), 40-35-
111(b)(3) (setting the minimum penalty for robbery at three years’ imprisonment);
United States v. Taylor, 301 F. App’x 508, 522 (6th Cir. 2008) (citing United States
v. Spears, 

, 1360-61 (11th Cir. 2006); United States v. Lender,

, 156 (4th Cir. 1993)) (holding that, under the disjunctive definition of
13-5288         United States v. Mitchell                                      Page 18


violent felony, the first part of the definition applies to adult convictions, even if
the conviction involves an act committed as a juvenile).

       There is no doubt that Mitchell’s 1988 robbery conviction was an adult
conviction.   In the PSR, Mitchell’s criminal history is divided into “juvenile
adjudications” and “adult criminal convictions”; the 1988 robbery is documented
under the latter category. The conviction was obtained in adult criminal court
(Shelby County Criminal Court). Moreover, under the Tennessee statute in effect
at the time, simple robbery was punishable by imprisonment for not less than five
years nor more than fifteen years. Tenn. Code Ann. § 39-2-501(a). If a defendant
had little or no prior criminal record, he would normally be sentenced within Range
I as a standard offender. See Tenn. Code Ann. § 40-35-105. A Range I sentence
can be within the range of five to ten years. Tenn. Code Ann. § 40-35-109(a)
(Supp. 1986). Therefore, at a minimum, Mitchell would have been subject to five
years’ imprisonment. Ultimately, he received an adult sentence—an eight-year
term of imprisonment in the Tennessee Department of Corrections. His 1988 adult
conviction clearly comports with the ACCA’s definition of “violent felony” and so
qualifies as a predicate offense.

       Mitchell’s reliance on United States v. Wells, 

(6th Cir. 2007),
is misplaced. In Wells, we openly prefaced that “[a]t issue in this appeal is not an
attempt to classify an adult conviction as a predicate offense for armed career
criminal purposes, but rather a juvenile adjudication.” 

(emphasis in
original). Therefore, the case is distinguishable and of no aid to Mitchell.

       AFFIRMED.