PRECEDENTIAL

     UNITED STATES COURT OF APPEALS
          FOR THE THIRD CIRCUIT
              _______________

                   No. 11-2067
                 _______________

B.H., A MINOR, BY AND THROUGH HER MOTHER;
    JENNIFER HAWK; K.M., A MINOR BY AND
           THROUGH HER MOTHER;
          AMY MCDONALD-MARTINEZ

                         v.

      EASTON AREA SCHOOL DISTRICT,
                             Appellant
             _______________

   On Appeal from the United States District Court
       For the Eastern District of Pennsylvania
        (D.C. Civil Action No. 5-10-cv-06283)
   District Judge: Honorable Mary A. McLaughlin
                  _______________

             Argued on April 10, 2012
   Rehearing En Banc Ordered on August 16, 2012
        Argued En Banc February 20, 2013
                _______________
Before: McKEE, Chief Judge, SLOVITER, SCIRICA,
 RENDELL, AMBRO, FUENTES, SMITH, FISHER,
       CHAGARES, JORDAN, HARDIMAN,
GREENAWAY, JR., VANASKIE, and GREENBERG,
                Circuit Judges

            (Opinion filed: August 5, 2013)

Keely J. Collins
John E. Freund, III     [ARGUED]
Jeffrey T. Tucker
King, Spry, Herman, Freund & Faul
One West Broad Street, Suite 700
Bethlehem, PA 18018

      Counsel for Appellant

Seth F. Kreimer
University of Pennsylvania School of Law
3400 Chestnut Street
Philadelphia, PA 19104

Mary Catherine Roper [ARGUED]
American Civil Liberties Union of Pennsylvania
P.O. Box 40008
Philadelphia, PA 19106

Molly M. Tack-Hooper
Berger & Montague
                          2
1622 Locust Street
Philadelphia, PA 19103

Witold J. Walczak
American Civil Liberties Union
313 Atwood Street
Pittsburgh, PA 15213

      Counsel for Appellees

Sean A. Fields
Pennsylvania School Boards Association
400 Bent Creek Boulevard
P.O. Box 2042
Mechanicsburg, PA 17055

      Counsel for Amicus Appellant

Wilson M. Brown, III
Kathryn E. Deal
Drinker, Biddle & Reath
18th & Cherry Streets
One Logan Square, Suite 2000
Philadelphia, PA 19103

Rory Wicks
Gary L. Sirota
1140 South Coast Highway 101
Encinitas, CA 92024
                          3
Amy R. Arroyo
2251 Las Palmas Drive
Carlsbad, CA 92011

Wayne Pollock
Dechert LLP
2929 Arch Street, 18th Floor Cira Centre
Philadelphia, PA 19104

Frank D. LoMonte
Laura Napoli
Student Press Law Center
1101 Wilson Boulevard, Suite 1100
Arlington, VA 22209

Terry L. Fromson
Carol E. Tracey
Women’s Law Project
125 South 9th Street, Suite 300
Philadelphia, PA 19107

David L. Cohen
3320 Market Street, Suite 232
Philadelphia, PA 19104

      Counsel for Amici Appellees


                            4
                    _______________

                       OPINION
                    _______________

SMITH, Circuit Judge, with whom McKEE, Chief
Judge, SLOVITER, SCIRICA, RENDELL, AMBRO,
FUENTES, FISHER, and VANASKIE, Circuit Judges
join.


       Once again, we are asked to find the balance
between a student’s right to free speech and a school’s
need to control its educational environment. In this case,
two middle-school students purchased bracelets bearing
the slogan “I ♥ boobies! (KEEP A BREAST)” as part of
a     nationally recognized        breast-cancer-awareness
campaign. The Easton Area School District banned the
bracelets, relying on its authority under Bethel School
District No. 403 v. Fraser, 

 (1986), to
restrict vulgar, lewd, profane, or plainly offensive speech,
and its authority under Tinker v. Des Moines Independent
Community School District, 

 (1969), to
restrict speech that is reasonably expected to substantially
disrupt the school. The District Court held that the ban
violated the students’ rights to free speech and issued a
preliminary injunction against the ban.

      We agree with the District Court that neither

                             5
Fraser nor Tinker can sustain the bracelet ban. The
scope of a school’s authority to restrict lewd, vulgar,
profane, or plainly offensive speech under Fraser is a
novel question left open by the Supreme Court, and one
which we must now resolve. We hold that Fraser, as
modified by the Supreme Court’s later reasoning in
Morse v. Frederick, 

 (2007), sets up the
following framework: (1) plainly lewd speech, which
offends for the same reasons obscenity offends, may be
categorically restricted regardless of whether it
comments on political or social issues, (2) speech that
does not rise to the level of plainly lewd but that a
reasonable observer could interpret as lewd may be
categorically restricted as long as it cannot plausibly be
interpreted as commenting on political or social issues,
and (3) speech that does not rise to the level of plainly
lewd and that could plausibly be interpreted as
commenting on political or social issues may not be
categorically restricted. Because the bracelets here are
not plainly lewd and because they comment on a social
issue, they may not be categorically banned under
Fraser. The School District has also failed to show that
the bracelets threatened to substantially disrupt the school
under Tinker. We will therefore affirm the District
Court.




                             6
                             I.

A.    Factual background
       As a “leading youth focused global breast cancer
organization,” the Keep A Breast Foundation tries to
educate thirteen- to thirty-year-old women about breast
cancer. Br. of Amicus Curiae KABF at 13. To that end,
it often partners with other merchants to co-brand
products that raise awareness. And because it believes
that young women’s “negative body image[s]” seriously
inhibit their awareness of breast cancer, the Foundation’s
products often “seek[] to reduce the stigma by speaking
to young people in a voice they can relate to.” Id. at 14–
15. If young women see such awareness projects and
products as cool and trendy, the thinking goes, then they
will be more willing to talk about breast cancer openly.

        To “start a conversation about that taboo in a light-
hearted way” and to break down inhibitions keeping
young women from performing self-examinations, the
Foundation began its “I ♥ Boobies!” initiative. Id. at 20–
21. Part of the campaign included selling silicone
bracelets of assorted colors emblazoned with “I ♥
Boobies! (KEEP A BREAST)” and “check y♥urself!
(KEEP A BREAST).” Id. at 21–22. The Foundation’s
website address (www.keep-a-breast.org) and motto
(“art. education. awareness. action.”) appear on the inside
of the bracelet. Id.

                             7
      As intended, the “I ♥ Boobies” initiative was a hit
with young women, quickly becoming one of the
Foundation’s “most successful and high profile
educational campaigns.” Id. at 20–21. Two of the young
women drawn to the bracelets were middle-school
students B.H. and K.M. They purchased the bracelets
with their mothers before the 2010–2011 school year—
B.H. because she saw “a lot of [her] friends wearing” the
bracelets and wanted to learn about them, and K.M.
because of the bracelet’s popularity and awareness
message. App. 72, 92, 106, 442.
       But the bracelets were more than just a new
fashion trend. K.M.’s purchase prompted her to become
educated about breast cancer in young women. The girls
wore their bracelets both to commemorate friends and
relatives who had suffered from breast cancer and to
promote awareness among their friends. Indeed, their
bracelets started conversations about breast cancer and
did so far more effectively than the more-traditional pink
ribbon. App. 73–74. That made sense to B.H., who
observed that “no one really notices” the pink ribbon,
whereas the “bracelets are new and . . . more appealing to
teenagers.” App. 74.

      B.H., K.M., and three other students wore the “I ♥
boobies! (KEEP A BREAST)” bracelets at Easton Area
Middle School during the 2010–2011 school year. A few
teachers, after observing the students wear the bracelets

                            8
every day for several weeks, considered whether they
should take action. The teachers’ responses varied: One
found the bracelets offensive because they trivialized
breast cancer. Others feared that the bracelets might lead
to offensive comments or invite inappropriate touching.
But school administrators also believed that middle-
school boys did not need the bracelets as an excuse to
make sexual statements or to engage in inappropriate
touching. See, e.g., Viglianti Test., App. 196, 198
(testifying that such incidents “happened before the
bracelets” and were “going to happen after the bracelets”
because “sexual curiosity between boys and girls in the
middle school is . . . a natural and continuing thing”).
      In mid- to late September, four or five teachers
asked the eighth-grade assistant principal, Amy
Braxmeier, whether they should require students to
remove the bracelets.       The seventh-grade assistant
principal, Anthony Viglianti, told the teachers that they
should ask students to remove “wristbands that have the
word ‘boobie’ written on them,” App. 343, even though
there were no reports that the bracelets had caused any
in-school disruptions or inappropriate comments.1

1
  In mid-October before the ban was publicly announced,
school administrators received some unrelated reports of
inappropriate touching, but neither the word “boobies”
nor the bracelets were considered a cause of these
incidents.
                            9
       With     Breast    Cancer    Awareness      Month
approaching in October, school administrators
anticipated that the “I ♥ boobies! (KEEP A BREAST)”
bracelets might reappear.2 The school was scheduled to
observe Breast Cancer Awareness Month on October 28,
so the day before, administrators publicly announced, for
the first time, the ban on bracelets containing the word
“boobies.”       Using the word “boobies” in his
announcement, Viglianti notified students of the ban over
the public-address system, and a student did the same on
the school’s television station. The Middle School still
encouraged students to wear the traditional pink, and it
provided teachers who donated to Susan G. Komen for
the Cure with either a pin bearing the slogan
“Passionately Pink for the Cure” or a T-shirt reading
“Real Rovers Wear Pink.”
      Later that day, a school security guard noticed
B.H. wearing an “I ♥ boobies! (KEEP A BREAST)”
bracelet and ordered her to remove it. B.H. refused.
After meeting with Braxmeier, B.H. relented, removed
her bracelet, and returned to lunch. No disruption
occurred at any time that day.

      The following day, B.H. and K.M. each wore their
“I ♥ boobies! (KEEP A BREAST)” bracelets to observe
2
  The Middle School permits students to wear the
Foundation’s “check y♥urself (KEEP A BREAST)”
bracelets.
                           10
the Middle School’s Breast Cancer Awareness Day. The
day was uneventful—until lunchtime. Once in the
cafeteria, both girls were instructed by a school security
guard to remove their bracelets. Both girls refused.
Hearing this encounter, another girl, R.T., stood up and
similarly refused to take off her bracelet. Confronted by
this act of solidarity, the security guard permitted the
girls to finish eating their lunches before escorting them
to Braxmeier’s office. Again, the girls’ actions caused no
disruption in the cafeteria, though R.T. told Braxmeier
that one boy had immaturely commented either that he
also “love[d] boobies” or that he “love[d] her boobies.”

      Braxmeier spoke to all three girls, and R.T. agreed
to remove her bracelet. B.H. and K.M. stood firm,
however, citing their rights to freedom of speech. The
Middle School administrators were having none of it.
They punished B.H. and K.M. by giving each of them
one and a half days of in-school suspension and by
forbidding them from attending the Winter Ball. The
administrators notified the girls’ families, explaining only
that B.H. and K.M. were being disciplined for
“disrespect,” “defiance,” and “disruption.”

       News of the bracelets quickly reached the rest of
the Easton Area School District, which instituted a
district-wide ban on the “I ♥ boobies! (KEEP A
BREAST)” bracelets, effective on November 9, 2010.
The only bracelet-related incident reported by school

                            11
administrators occurred weeks after the district-wide ban:
Two girls were talking about their bracelets at lunch
when a boy who overheard them interrupted and said
something like “I want boobies.” He also made an
inappropriate gesture with two red spherical candies.
The boy admitted his “rude” comment and was
suspended for one day.3

      This was not the first time the Middle School had
banned clothing that it found distasteful. Indeed, the
School District’s dress-code policy prohibits “clothing
imprinted with nudity, vulgarity, obscenity, profanity,
and double entendre pictures or slogans.”4 Under the
policy, seventh-grade students at the Middle School have
been asked to remove clothing promoting Hooters and
Big Pecker’s Bar & Grill, as well as clothing bearing the
phrase “Save the ta-tas” (another breast-cancer-
awareness slogan). Typically, students are disciplined
only if they actually refuse to remove the offending
apparel when asked to do so.

B.    Procedural history

3
  After the district-wide ban was in place, there were
several incidents of middle-school boys inappropriately
touching girls, but they were unrelated to the “I ♥
boobies! (KEEP A BREAST)” bracelets.
4
  B.H. and K.M. do not assert a facial challenge to the
constitutionality of the dress-code policy.
                           12
       Through their mothers, B.H. and K.M. sued the
School District under 42 U.S.C. § 1983.5 Compl., ECF
No. 1 ¶ 3, B.H. v. Easton Area Sch. Dist., No. 5:10-CV-
06283-MAM (E.D. Pa. Nov. 15, 2010). They sought a
temporary restraining order allowing them to attend the
Winter Ball and a preliminary injunction against the
bracelet ban. B.H. v. Easton Area Sch. Dist., 827 F.
Supp. 2d 392, 394 (E.D. Pa. 2011). At the District
Court’s urging, the School District reversed course and
permitted B.H. and K.M. to attend the Winter Ball while
retaining the option to impose a comparable punishment
if the bracelet ban was upheld. Id. The District Court
accordingly denied the motion for a temporary
restraining order. Id.
       The District Court conducted an evidentiary
hearing on the request for a preliminary injunction. It
soon became clear that the School District’s rationale for
disciplining B.H. and K.M. had shifted. Although B.H.’s
and K.M.’s disciplinary letters indicated only that they
were being disciplined for “disrespect,” “defiance,” and
“disruption,” the School District ultimately based the ban


5
    The District Court had both federal-question
jurisdiction under 28 U.S.C. § 1331 and § 1983
jurisdiction under 28 U.S.C. § 1343(a)(3). See Max v.
Republican Comm. of Lancaster Cnty., 

, 199
n.1 (3d Cir. 2009).
                           13
on its dress-code policy6 together with the bracelets’
alleged sexual innuendo. According to the School
District’s witnesses, the Middle School assistant
principals had conferred and concluded that the bracelets
“conveyed a sexual double entendre” that could be
harmful and confusing to students of different physical
and sexual developmental levels. Sch. Dist.’s Br. at 9.
And the principals believed that middle-school students,
who often have immature views of sex, were particularly
likely to interpret the bracelets that way. For its part, the
Foundation explained that no one there “ever suggested
that the phrase ‘I (Heart) Boobies!’ is meant to be sexy.”
App. 150. To that end, the Foundation had denied
requests from truck stops, convenience stores, vending
machine companies, and pornographers to sell the

6
  Even the Middle School administrators seemed unsure
which words would be prohibited by the dress code.
When deposed, Viglianti and principal Angela DiVietro
testified that the word “breast” (as in apparel stating
“keep-a-breast.org” or “breast cancer awareness”) would
be inappropriate because the word “breast” “can be
construed as [having] a sexual connotation.” App. 490,
497. At the District Court’s evidentiary hearing, they
reversed course. Viglianti stated that “keep-a-breast.org”
would be appropriate “[i]n the context of Breast Cancer
Awareness Month,” and DiVeitro no longer believed the
phrase “breast cancer awareness” was vulgar to middle-
school students.
                             14
bracelets.
       After the evidentiary hearing, the District Court
preliminarily enjoined the School District’s bracelet ban.
According to the District Court, B.H. and K.M. were
likely to succeed on the merits because the bracelets did
not contain lewd speech under Fraser and did not
threaten to substantially disrupt the school environment
under Tinker. The District Court could find no other
basis for regulating the student speech at issue. The
School District appealed, and the District Court denied its
request to stay the injunction pending this appeal.

                            II.
       Although the District Court’s preliminary
injunction is not a final order, we have jurisdiction under
28 U.S.C. § 1292(a)(1), which grants appellate
jurisdiction over “[i]nterlocutory orders of the district
courts . . . granting, continuing, modifying, refusing, or
dissolving injunctions.” See Sypniewski v. Warren Hills
Reg’l Bd. of Educ., 

, 252 n.10 (3d Cir.
2002). We review the District Court’s factual findings
for clear error, its legal conclusions de novo, and its
ultimate decision to grant the preliminary injunction for
abuse of discretion. Id. at 252. Four factors determine
whether a preliminary injunction is appropriate:

      (1) whether the movant has a reasonable
      probability of success on the merits; (2)
                            15
      whether the movant will be irreparably
      harmed by denying the injunction; (3)
      whether there will be greater harm to the
      nonmoving party if the injunction is granted;
      and (4) whether granting the injunction is in
      the public interest.

Id. (quoting Highmark, Inc. v. UPMC Health Plan, Inc.,

, 170 (3d Cir. 2001)). The District Court
concluded that all four factors weighed in favor of B.H.
and K.M. In school-speech cases, though, the first
factor—the likelihood of success on the merits—tends to
determine which way the other factors fall. Id. at 258.
Because the same is true here, we focus first on B.H. and
K.M.’s burden to show a likelihood of success on the
merits. Id.

                            III.
       The School District defends the bracelet ban as an
exercise of its authority to restrict lewd, vulgar, profane,
or plainly offensive student speech under Fraser. As to
the novel question of Fraser’s scope, jurists seem to
agree on one thing: “[t]he mode of analysis employed in
Fraser is not entirely clear.” Morse, 551 U.S. at 404.7

7
   The rest of the Supreme Court’s student-speech
jurisprudence might fairly be described as opaque. See
Morse, 551 U.S. at 418 (Thomas, J., concurring) (“I am
afraid that our jurisprudence now says that students have
                            16
On this point, we think the Supreme Court’s student-
speech cases are more consistent than they may first
appear. As we explain, Fraser involved only plainly
lewd speech. We hold that, under Fraser, a school may
also categorically restrict speech that—although not
plainly lewd, vulgar, or profane—could be interpreted by
a reasonable observer as lewd, vulgar, or profane so long
as it could not also plausibly be interpreted as
commenting on a political or social issue. Because the “I
♥ boobies! (KEEP A BREAST)” bracelets are not plainly
lewd and express support for a national breast-cancer-
awareness campaign—unquestionably an important
social issue—they may not be categorically restricted

a right to speak in schools except when they do
not . . . .”); id. at 430 (Breyer, J., concurring in part and
dissenting in part) (“[C]ourts have described the tests
these cases suggest as complex and often difficult to
apply.”); see, e.g., Doninger v. Niehoff, 

,
353 (2d Cir. 2011) (“The law governing restrictions on
student speech can be difficult and confusing, even for
lawyers, law professors, and judges. The relevant
Supreme Court cases can be hard to reconcile, and courts
often struggle with which standard applies in any
particular case.”); Guiles ex rel. Guiles v. Marineau, 

, 326, 331 (2d Cir. 2006) (acknowledging “some
lack of clarity in the Supreme Court’s student-speech
cases” and stating that the “exact contours of what is
plainly offensive [under Fraser] is not so clear”).
                             17
under Fraser.

A.    The Supreme Court’s decision in Fraser
       “[A]s a general matter, the First Amendment
means that government has no power to restrict
expression because of its message, its ideas, its subject
matter, or its content.” Ashcroft v. ACLU, 

,
573 (2002). Of course, there are exceptions. When
acting as sovereign, the government is empowered to
impose time, place, and manner restrictions on speech,
see Ward v. Rock Against Racism, 

, 791
(1989), make reasonable, content-based decisions about
what speech is allowed on government property that is
not fully open to the public, see Ark. Educ. Television
Comm’n v. Forbes, 

, 674–75 (1998), decide
what viewpoints to espouse in its own speech or speech
that might be attributed to it, see Johanns v. Livestock
Mktg. Ass’n, 

, 560 (2005), and categorically
restrict unprotected speech, such as obscenity, see Miller
v. California, 

, 23 (1973).8

8
  Other examples of categorically unprotected speech
include child pornography, see New York v. Ferber, 

, 764–65 (1982), advocacy that imminently
incites lawless action, see Brandenburg v. Ohio, 

, 447–48 (1969) (per curiam), fighting words, see
Chaplinsky v. New Hampshire, 

, 571–72
(1942), true threats, see Watts v. United States, 

      Sometimes, however, the government acts in
capacities that go beyond being sovereign. In those
capacities, it not only retains its sovereign authority over
speech but also gains additional flexibility to regulate
speech. See In re Kendall, 

, 825 (3d Cir.
2013) (collecting examples).          One of those other
capacities is K-12 educator. Although “students do not
‘shed their constitutional rights to freedom of speech or
expression at the schoolhouse gate,’” the First
Amendment has to be “applied in light of the special
characteristics of the school environment” and thus
students’ rights to freedom of speech “are not
automatically coextensive with the rights of adults in
other settings.” Morse, 551 U.S. at 396–97 (internal
quotation marks and citations omitted).

      The Supreme Court first expressed this principle
nearly a half century ago. In 1965, the United States
deployed over 200,000 troops to Vietnam as part of
Operation Rolling Thunder—and thus began the Vietnam
War. That war “divided this country as few other issues
[e]ver have.” Tinker, 393 U.S. at 524 (Black, J.,


705, 708 (1969) (per curiam), commercial speech that is
false, misleading, or proposes illegal transactions, see
Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n
of N.Y., 

, 562, 566–67 (1980), and some
false statements of fact, see United States v. Alvarez, 

, 2546–47 (2012).
                            19
dissenting). Public opposition to the war made its way
into schools, and in one high-profile case, a group of
high-school and middle-school students wore black
armbands to express their opposition. Id. at 504
(majority opinion). School officials adopted a policy
prohibiting the armbands and suspending any student
who refused to remove it when asked. Id. Some students
refused and were suspended. Id. The Supreme Court
upheld their right to wear the armbands. Id. at 514.
Tinker held that school officials may not restrict student
speech without a reasonable forecast that the speech
would substantially disrupt the school environment or
invade the rights of others. Id. at 513. As nothing more
than the “silent, passive expression of opinion,
unaccompanied by any disorder or disturbance on [the
students’] part,” the students’ armbands were protected
by the First Amendment. Id. at 508.

      Under Tinker’s “general rule,” the government
may restrict school speech that threatens a specific and
substantial disruption to the school environment or that
“inva[des] . . . the rights of others.”9 Saxe v. State


9
  We have not yet decided whether Tinker is limited to
on-campus speech. See J.S. v. Blue Mountain Sch. Dist.,

, 926 & n.3 (3d Cir. 2011) (en banc)
(declining to reach this issue); see also id. at 936 (Smith,
J., concurring) (“I write separately to address a question
                            20
College Area Sch. Dist., 

, 211, 214 (3d Cir.
2001) (citing Tinker, 393 U.S. at 504). Since Tinker, the
Supreme Court has identified three “narrow”
circumstances in which the government may restrict
student speech even when there is no risk of substantial
disruption or invasion of others’ rights. Id. at 212. First,
the government may categorically restrict vulgar, lewd,
profane, or plainly offensive speech in schools, even if it
would not be obscene outside of school. Fraser, 478
U.S. at 683, 685. Second, the government may likewise
restrict speech that “a reasonable observer would
interpret as advocating illegal drug use” and that cannot
“plausibly be interpreted as commenting on any political
or social issue.” Morse, 551 U.S. at 422 (Alito, J.,
concurring); see also id. at 403 (majority opinion)
(“[T]his is plainly not a case about political debate over
the criminalization of drug use or possession.”).10 And
third, the government may impose restrictions on school-
sponsored speech that are “reasonably related to
legitimate pedagogical concerns”—a power usually
lumped together with the other school-specific speech
doctrines but that, strictly speaking, simply reflects the
government’s more general power as sovereign over

that the majority opinion expressly leaves open: whether
Tinker applies to off-campus speech in the first place.”).
10
   As we explain in Part III.B(2), the limitations that
Justice Alito’s concurrence places on the majority’s
opinion in Morse are controlling.
                            21
government-sponsored speech.11 Hazelwood Sch. Dist. v.

11
   Compare Pleasant Grove City, Utah v. Summum, 

, 468 (2009) (discussing the government-speech
doctrine and explaining that “[a] government entity may
exercise this same freedom to express its views when it
receives assistance from private sources for the purpose
of delivering a government-controlled message” (citing
Johanns, 544 U.S. at 562)), with Kuhlmeier, 484 U.S. at
271, 273 (reaffirming the government’s same authority to
control speech that might be “reasonably perceive[ed] to
bear the imprimatur of the school” in its role as K-12
educator); see also Eugene Volokh, The First
Amendment and the Government as K-12 Educator, The
Volokh Conspiracy (Oct. 31, 2011, 6:26 PM),
http://www.volokh.com/2011/10/31/the-first-
amendment-and-the-government-as-k-12-educator/
(“[Kuhlmeier] generally reflects broad government-as-
speaker law, and not special rules related to the
government as K-12 educator.”); Michael J. O’Connor,
Comment, School Speech in the Internet Age: Do
Students Shed Their Rights When They Pick Up a
Mouse?, 11 U. Pa. J. Const. L. 459, 469 (2009)
(“Hazelwood . . . simply illustrates the idea that the
school speech arena is not isolated from developments in
wider First Amendment jurisprudence. . . . Hazelwood
recognizes that schools are government actors and
therefore are entitled to control speech that could be
reasonably viewed as originating with them.”); Gia B.
                          22
Kuhlmeier, 

, 273 (1988).
       The first exception is at issue here. We must
determine the scope of the government’s authority to
categorically restrict vulgar, lewd, indecent, or plainly
offensive speech under Fraser. Fraser involved a high-
school assembly during which a student “nominated a
peer for class office through an ‘an elaborate, graphic,
and explicit sexual metaphor.’” Saxe, 240 F.3d at 212
(quoting Fraser, 478 U.S. at 677). Fraser’s speech
“glorif[ied] male sexuality”:

      I know a man who is firm—he’s firm in his
      pants, he’s firm in his shirt, his character is
      firm—but most . . . of all, his belief in you,
      the students of Bethel, is firm. . . . Jeff
      Kuhlman [the candidate] is a man who takes
      his point and pounds it in. If necessary, he’ll
      take an issue and nail it to the wall. He
      doesn’t attack things in spurts, he drives hard,
      pushing and pushing until finally—he
      succeeds. . . . Jeff is a man who will go to the
      very end—even the climax, for each and
      every one of you. . . . So vote for Jeff for
      A.S.B. vice-president—he’ll never come
      between you and the best our high school can

Lee, First Amendment Enforcement in Government
Institutions and Programs, 56 UCLA L. Rev. 1691,
1711–12 (2009) (similar).
                            23
      be.
Fraser, 478 U.S. at 687 (Brennan, J., concurring). In
response, “[s]ome students hooted and yelled; some by
gestures simulated the sexual activities pointedly alluded
to in [Fraser’s] speech.” Id. at 678 (majority opinion).
Still “[o]ther students appeared to be bewildered and
embarrassed by the speech.” Id. The school suspended
Fraser and took him out of the running for graduation
speaker. Id.
       The Supreme Court upheld Fraser’s suspension.
Id. at 683. Rather than requiring a reasonable forecast of
substantial disruption under Tinker, the Court held that
lewd, vulgar, indecent, and plainly offensive student
speech is categorically unprotected in school, even if it
falls short of obscenity and would have been protected
outside school. Saxe, 240 F.3d at 213 (discussing
Fraser); Morse, 551 U.S. at 405 (“Had Fraser delivered
the same speech in a public forum outside the school
context, it would have been protected.”); Fraser, 478
U.S. at 688 (Blackmun, J., concurring) (“If [Fraser] had
given the same speech outside of the school environment,
he could not have been penalized simply because
government officials considered his language to be
inappropriate.”). For this proposition, the Court relied on
precedent holding that the government can restrict
expression that would be obscene from a minor’s
perspective—even though it would not be obscene in an

                            24
adult’s view—where minors are either a captive audience
or the intended recipients of the speech. See Fraser, 478
U.S. at 684–85 (relying on Ginsberg v. New York, 

, 635–37 & nn.4–5 (1968) (upholding criminal
punishment for selling to minors any picture depicting
nudity); Bd. of Educ., Island Trees Union Free Sch. Dist.
No. 26 v. Pico, 

, 870 (1982) (plurality
opinion) (acknowledging that the Free Speech Clause
would allow a local board of education to remove
“pervasively vulgar” books from school libraries); and
FCC v. Pacifica Found., 

, 749–50 (1978)
(rejecting a Free Speech Clause challenge to the FCC’s
broad leeway to regulate indecent-but-not-obscene
material on broadcast television during hours when
children were likely to watch)).

     Fraser did no more than extend these obscenity-to-
minors12 cases to another place where minors are a

12
  See Brown v. Entm’t Merchs. Ass’n, 

,
2735 (2011) (describing Ginsberg as regulating
“obscenity for minors”); Reno v. ACLU, 

,
869 (1997) (reaffirming the government’s power under
Pacifica and Ginsberg to “‘protect[] the physical and
psychological well-being of minors’ which extended to
shield them from indecent messages that are not obscene
by adult standards” (quoting Sable Comm’cns of Cal.,
Inc. v. FCC, 

, 126 (1989))); Pacifica
Found., 438 U.S. at 767 (Brennan, J., dissenting)
                           25
captive audience—schools.         Indeed, as the Court
explained, schools are tasked with more than just
“educating our youth” about “books, the curriculum, and
the civics class.” Id. at 681. Society also expects schools
to “teach[] students the boundaries of socially appropriate
behavior,” including the “fundamental values of ‘habits
and manners of civility’ essential to a democratic
society.”      Id. at 681, 683 (citation omitted).
Consequently, Fraser’s “sexually explicit monologue”
was not protected. Id. at 685.

      It is important to recognize what was not at stake
in Fraser. Fraser addressed only a school’s power over
speech that was plainly lewd—not speech that a
reasonable observer could interpret as either lewd or non-
lewd. See, e.g., Doninger v. Niehoff, 

, 49 (2d
Cir. 2008) (“[Fraser’s] reference to ‘plainly offensive’
speech must be understood in light of the vulgar, lewd,
and sexually explicit language that was at issue in [that]
case.”); Chandler v. McMinnville Sch. Dist., 978 F.2d


(agreeing with the majority that the government could
regulate “variable obscenity” or “obscenity to minors” on
broadcast television, but disagreeing with the majority
that the Carlin monologue met that standard); Erznoznik
v. City of Jacksonville, 

, 213 n.10 (1975)
(describing Ginsberg as involving “obscenity as to
minors”); Ginsberg, 390 U.S. at 635 n.4 (using the label
“variable obscenity”).
                            26
524, 530 (9th Cir. 1992) (interpreting Fraser as limited to
“per se vulgar, lewd, obscene, or plainly offensive”
school speech). After all, the Court believed Fraser’s
speech to be “plainly offensive to both teachers and
students—indeed to any mature person.”13 Fraser, 478
U.S. at 683.

       And because it was plainly lewd, the Court did not
believe that Fraser’s speech could plausibly be
interpreted as political or social commentary.         In
hindsight, it might be tempting to believe that Fraser’s
speech was political because it was made in the context
of a student election. Cf. Citizens United v. FEC, 

, 

, 898 (2010) (describing the
importance of political speech as the “means to hold

13
   Of course, Fraser’s speech might “seem[] distinctly
lacking in shock value” today, especially “from the
perspective enabled by 25 years of erosion of refinement
in the use of language.” Zamecnik v. Indian Prairie Sch.
Dist. No. 204, 

, 877 (7th Cir. 2011); see also
Fraser, 478 U.S. at 691 (Stevens, J., dissenting) (noting
that Clark Gable’s famous use of the word “damn” in
“Frankly, my dear, I don’t give a damn” “shocked the
Nation” when Justice Stevens was a high school student
but had become “less offensive” by the time of Fraser).
Any such change in perspective, however, is irrelevant to
our examination of the Court’s interpretation of Fraser’s
speech and its reasoning.
                            27
officials accountable to the people”). But that kind of
revisionist history is belied by both the logic and
language of Fraser. “Fraser permits a school to prohibit
words that ‘offend for the same reasons that obscenity
offends.’” Saxe, 240 F.3d at 213 (quoting Fraser, 478
U.S. at 685). Obscenity, in turn, offends because it is “no
essential part of any exposition of ideas, and [is] of such
slight social value as a step to truth that any benefit that
may be derived from [it] is clearly outweighed by the
social interest in order and morality.” Fraser, 478 U.S.
at 683 (quoting Pacifica Found., 438 U.S. at 746
(plurality opinion)). In other words, obscenity and
obscenity to minors, like “other historically unprotected
categories of speech,” have little or no political or social
value. United States v. Stevens, 

, 

, 1585 (2010). By concluding that Fraser’s speech
met the obscenity-to-minors standard, the Court
necessarily implied that his speech could not be
interpreted as having “serious” political value. Miller,
413 U.S. at 24.
       In fact, the majority in Fraser made this explicit.
“[T]he Fraser [C]ourt distinguished its holding from
Tinker in part on the absence of any political message in
Fraser’s speech.” Guiles ex rel. Guiles v. Marineau, 

, 326, 328 (2d Cir. 2006). In the Court’s own
words, there was a “marked distinction between the
political ‘message’ of the armbands in Tinker and the
sexual content of [Fraser’s] speech.” Fraser, 478 U.S. at
                            28
680 (emphasis added); see also Defoe ex rel. Defoe v.
Spiva, 

, 332 (6th Cir. 2010) (“Tinker
governs this case because by wearing clothing bearing
images of the Confederate flag, Tom Defoe engaged in
‘pure speech,’ which is protected by the First
Amendment, and thus Fraser would not apply.”).
Several courts of appeals have similarly interpreted
Fraser. Guiles, 461 F.3d at 326, 328; Newsom ex rel.
Newsom v. Albemarle Cnty. Sch. Bd., 

, 256
(4th Cir. 2003) (explaining that Fraser “distinguish[ed]
Tinker on the basis that the lewd, vulgar, and plainly
offensive speech was ‘unrelated to any political
viewpoint’ (quoting Fraser, 478 U.S. at 685)); Chandler,
978 F.2d at 532 n.2 (Goodwin, J., concurring)
(concluding that Fraser does not apply because “this case
clearly involves political speech”). And the Supreme
Court later characterized Fraser’s reasoning the same
way. Morse, 551 U.S. at 404 (noting that Fraser was
“plainly attuned” to the sexual, non-political “content of
Fraser’s speech”). In fact, Morse refused to “stretch[]
Fraser” so far as to “encompass any speech that could fit
under some definition of ‘offensive’” out of a fear that
“much political and religious speech might be perceived
as offensive to some.” Id. at 409. Fraser therefore
involved plainly lewd speech that did not comment on
political or social issues.




                           29
B.    How far does a school’s authority under Fraser
      extend?
       The School District asks us to extend Fraser in at
least two ways: to reach speech that is ambiguously lewd,
vulgar, or profane and to reach speech on political or
social issues.14 The first step is justified, but the second

14
   Fraser differs from this case in a third way: Fraser
involved speech at an official school assembly, whereas
the School District’s bracelet ban extends to the entire
school day, not just school-sponsored functions. But like
other courts of appeals, we do not think that this
difference matters. See, e.g., R.O. ex rel. Ochshorn v.
Ithaca City Sch. Dist., 

, 542 (2d Cir. 2011)
(“[W]e have not interpreted Fraser as limited either to
regulation of school-sponsored speech or to the spoken
word.”); Chandler, 978 F.2d at 529 (concluding that
restriction of vulgar, lewd, and plainly offensive speech
under Fraser is not limited to speech “given at an official
school assembly”); Bystrom by and through Bystrom v.
Fridley High Sch., Indep. Sch. Dist. No. 14, 

, 753 (8th Cir. 1987) (“It is true that [Fraser]
involved a speech given before a student assembly . . . .
[But] [t]his possible difference, in our view, does not
amount to a legal distinction making the Bethel rule
inapplicable here.”). As we explained, Fraser reflected
an extension of the Court’s obscenity-to-minors
jurisprudence, which permits the government to restrict
                            30
lewd speech to children where children are either a
captive audience or the intended recipients of the speech.
Children are just as much of a captive audience in the
hallways, cafeteria, or locker rooms as they are in official
school assemblies and classrooms. Naturally, then, we
have never described a school’s authority under Fraser
as being limited to official school functions and
classrooms. See, e.g., J.S., 650 F.3d at 927 (“The first
exception is set out in Fraser, which we interpreted to
permit school officials to regulate “‘lewd,’ ‘vulgar,’
‘indecent,’ and ‘plainly offensive’ speech in school.”
(emphasis in original) (quoting Saxe, 240 F.3d at 213)).
Although Justice Brennan’s concurrence and Justice
Stevens’s dissent in Fraser suggested that this difference
might matter, nothing in the majority opinion endorsed
their distinction. See Fraser, 478 U.S. at 689 (Brennan,
J., concurring) (opining that Fraser’s “speech may well
have been protected had he given it in school but under
different circumstances, where the school’s legitimate
interests in teaching and maintaining civil public
discourse were less weighty”); id. at 696 (Stevens, J.,
dissenting) (“It seems fairly obvious that [Fraser’s]
speech would be inappropriate in certain classroom and
formal social settings. On the other hand, in a locker
room or perhaps in a school corridor the metaphor in the
speech might be regarded as rather routine comment.”).
Indeed, if Fraser were so limited, then a school’s
authority under Fraser would largely merge with its
                            31
is not.

          1. Under Fraser, schools may restrict
             ambiguously lewd speech only if it cannot
             plausibly be interpreted as commenting on a
             social or political matter.
       Although Fraser involved plainly lewd, vulgar,
profane, or offensive speech that “offends for the same
reasons obscenity offends,” Saxe, 240 F.3d at 213
(quoting Fraser, 478 U.S. at 685), student speech need
not rise to that level to be restricted under Fraser. We
conclude that schools may also categorically restrict
ambiguous speech that a reasonable observer could
interpret as lewd, vulgar, profane, or offensive—unless,
as explained below, the speech could also plausibly be
interpreted as commenting on a political or social issue.
After all, Fraser made clear that “the determination of
what manner of speech in the classroom or in school
assembly is inappropriate properly rests with the school
board.” 478 U.S. at 683. The Supreme Court’s three
other student-speech cases suggest that courts should
defer to a school’s decisions to restrict what a reasonable
observer would interpret as lewd, vulgar, profane, or
offensive. See Morse, 551 U.S. at 403 (explaining that,

power to reasonably regulate school-sponsored speech
under Kuhlmeier, yet we have always viewed Fraser and
Kuhlmeier as separate exceptions to Tinker. See, e.g.,
J.S., 650 F.3d at 927.
                             32
under Tinker, courts determine whether school officials
have “reasonably conclude[d]” that student speech will
substantially disrupt the school); id. at 405 (explaining
that, under Kuhlmeier, courts uphold a school’s
reasonable, pedagogically related restrictions on speech
that an observer could reasonably attribute to the school);
id. at 422 (Alito, J., concurring) (explaining that schools
may restrict student speech that could “reasonably be
regarded as encouraging illegal drug use” and that could
not plausibly be interpreted as commenting on a political
or social issue). This makes sense. School officials
know the age, maturity, and other characteristics of their
students far better than judges do. Our review is
restricted to a cold and distant record. And we must take
into account that these same officials must often act
“suddenly and unexpectedly” based on their experience.
Id. at 409–10 (majority opinion); see, e.g., Walker-
Serrano ex rel. Walker v. Leonard, 

, 416–17
(3d Cir. 2003) (“There can be little doubt that speech
appropriate for eighteen-year-old high school students is
not necessarily acceptable for seven-year-old grammar
school students. Human sexuality provides the most
obvious example of age-sensitive matter . . . .” (citing
Fraser, 478 U.S. at 683–84)); Sypniewski, 306 F.3d at
266 (“What is necessary in one school at one time will
not be necessary elsewhere and at other times.”).

      It remains the job of judges, nonetheless, to
determine whether a reasonable observer could interpret
                            33
student speech as lewd, profane, vulgar, or offensive.
See Morse, 551 U.S. at 402 (taking the same approach
with respect to the message of drug advocacy on
Frederick’s banner); see also Christian Legal Soc’y
Chapter of the Univ. of Cal. v. Martinez, 

,
2988 (2010) (“This Court is the final arbiter of the
question whether a public university has exceeded
constitutional constraints, and we owe no deference to
universities when we consider that question.”). Whether
a reasonable observer could interpret student speech as
lewd, profane, vulgar, or offensive depends on the
plausibility of the school’s interpretation in light of
competing meanings; the context, content, and form of
the speech; and the age and maturity of the students. See,
e.g., Chandler, 978 F.2d at 530 (analyzing the word
“scab” on buttons worn by students during a teacher
strike to determine whether it was a vulgar, offensive
epithet or just “common parlance” and concluding that, at
the motion-to-dismiss stage, Fraser did not apply).
       Although this is a highly contextual inquiry,
several rules apply. A reasonable observer would not
adopt an acontextual interpretation, and the subjective
intent of the speaker is irrelevant. See Morse, 551 U.S. at
401–02 (explaining that Frederick’s desire to appear on
television “was a description of [his] motive for
displaying the banner” and “not an interpretation of what
the banner sa[id]”); see also Saxe, 240 F.3d at 216–17
(noting that students’ intent to offend or disrupt does not
                            34
satisfy Tinker). And Fraser is not a blank check to
categorically restrict any speech that touches on sex or
any speech that has the potential to offend. See Morse,
551 U.S. at 401, 409 (refusing to “stretch[] Fraser” so far
as “to encompass any speech that could fit under some
definition of ‘offensive’ and rejecting the argument that
the “BONG HiTS 4 JESUS” message on Frederick’s
banner could be banned under Fraser, even though it “is
no doubt offensive to some”); accord Eugene Volokh,
May ‘Jesus Is Not a Homophobe’ T-shirt Be Banned
From Public High School As ‘Indecent’ And ‘Sexual’?,
The Volokh Conspiracy (Apr. 4, 2012, 3:36 PM),
http://www.volokh.com/2012/04/04/may-jesus-was-not-
a-homophobe-T-shirt-be-banned-from-public-high-
school-as-indecent-and-sexual/ (“But Fraser . . . hardly
suggested that all speech on political and religious
questions related to sexuality and sexual orientation
could be banned from public high school.”). After all, a
school’s mission to mold students into citizens capable of
engaging in civil discourse includes teaching students of
sufficient age and maturity how to navigate debates
touching on sex.




                            35
      2. Fraser does not permit a school to restrict
         ambiguously lewd speech that can also
         plausibly be interpreted as commenting on a
         social or political issue.
       A school’s leeway to categorically restrict
ambiguously lewd speech, however, ends when that
speech could also plausibly be interpreted as expressing a
view on a political or social issue. Justices Alito and
Kennedy’s concurrence in Morse adopted a similar
protection for political speech that could be interpreted as
illegal drug advocacy.         Their narrower rationale
protecting political speech limits and controls the
majority opinion in Morse, and it applies with even
greater force to ambiguously lewd speech.
       Justice Alito’s concurrence, joined by Justice
Kennedy, provided the crucial fourth and fifth votes in
the five-to-four majority opinion. But the two justices
conditioned their votes on the “understanding that (1)
[the majority opinion] goes no further than to hold that a
public school may restrict speech that a reasonable
observer would interpret as advocating illegal drug use
and (2) it provides no support for any restriction of
speech that can plausibly be interpreted as commenting
on any political or social issue.” Morse, 551 U.S. at 422
(Alito, J., concurring); see id. at 425 (regarding the
categorical regulation of non-political advocacy of
ambiguous illegal drug advocacy “as standing at the far

                            36
reaches of what the First Amendment permits” and
“join[ing] the opinion of the Court with the
understanding that the opinion does not endorse any
further extension”). The purpose of Justice Alito’s
concurrence was to “ensur[e] that political speech will
remain protected within the school setting” (subject, as
always, to Tinker’s substantial-disruption principle).
Ponce v. Socorro Indep. Sch. Dist., 

, 768
(5th Cir. 2007).
       Because the votes of Justices Alito and Kennedy
were necessary to the majority opinion and were
expressly conditioned on their narrower understanding
that speech plausibly interpreted as political or social
commentary was protected from categorical regulation,
that limitation is a binding part of Morse. This
conclusion requires a minor detour. The most familiar
situation in which we follow the narrowest rationale was
expressed t by the Supreme Court in Marks v. United
States: when “no single rationale explaining the result
enjoys the assent of five Justices, the holding of the Court
may be viewed as that position taken by those Members
who concurred in the judgments on the narrowest
grounds.” 

, 193 (1977) (internal quotation
marks and citations omitted). But that situation is not the
only one in which we tally the justices’ views and look
for the narrowest rationale. The Supreme Court and this
Court have both applied the narrowest-grounds approach
in circumstances beyond those posed by Marks,
                            37
including to determine holdings in majority opinions (not
just plurality opinions involving “no single legal rationale
explain[ing] the result”) 15 and to count even dissenting
justices’ votes that, by definition, could not “explain the
result” (not just the votes of those who “concurred in the
judgments”).16 See United States v. Johnson, 

, 65 (1st Cir. 2006) (noting that the Supreme Court has
“moved away” from adhering to the strict circumstances
in Marks).
      And it makes sense that the limitations in Justice
Alito’s concurrence would narrow the majority opinion.
When an individual justice’s vote is not needed to form a
majority, “the meaning of a majority opinion is to be
found within the opinion itself” because “the gloss that
an individual [j]ustice chooses to place upon it is not
authoritative.” McKoy v. North Carolina, 

,

15
   See discussion of Horn and Bishop infra pp. 30–33.
16
   See, e.g., Nichols v. United States, 

, 746
(1994) (combining the views of four dissenters and
Justice Stewart in Baldasar v. Illinois, 

(1980), to form a “holding”); Donovan, 661 F.3d at 182
(“[W]e have looked to the votes of dissenting Justices if
they, combined with votes from plurality or concurring
opinions, establish a majority view on the relevant
issue.”); Student Pub. Interest Research Grp. of N.J., Inc.
v. AT&T Bell Labs., 

, 1451 & n.16 (3d Cir.
1988) (same).
                            38
448 n.3 (1990) (Blackmun, J., concurring). But when an
individual justice joins the majority and is essential to
maintaining the majority, and then writes separately, “the
opinion is not a majority opinion except to the extent that
it accords with his views.” Id. at 462 n.3 (Scalia, J.,
dissenting). Of course, that linchpin justice’s opinion
“cannot add to what the majority opinion holds” by
“binding the other four [j]ustices to what they have not
said” because his views would not be the narrowest
grounds. Id. But that justice’s separate opinion “can
assuredly narrow what the majority opinion holds, by
explaining the more limited interpretation adopted by that
necessary member of the majority.” Id. In that case, the
linchpin justice’s views are “the least common
denominator” necessary to maintain a majority opinion.
Id.; see generally Sonja R. West, Concurring in Part and
Concurring in the Confusion, 

. Rev. 1951
(2006) (advocating the same approach and explaining
that it is consistent with determining precedent from the
traditional Supreme Court’s seriatim opinions).
       Indeed, this is not the first time that we have been
compelled to limit a majority opinion by a linchpin
justice’s narrower concurrence. In Horn v. Thoratec, we
considered whether the federal regulation of medical
devices preempts only state-law “requirement[s]”
specific to medical devices or also preempts general
common-law claims not specific to medical devices (such
as negligence). See 

, 173–74 (3d Cir. 2004).
                            39
That, in turn, required us to analyze the Supreme Court’s
decision in Medtronic v. Lohr, 

 (1996). We
read Part V of the Lohr majority opinion—which Justice
Breyer formally joined as the fifth vote—as saying that
only device-specific state-law requirements, not general
common-law claims, are preempted. See Horn, 376 F.3d
at 174 (noting that the majority in Part V conclud[ed] that
common-law claims “escape[]” preemption because
“their generality leaves them outside” of the preempted
category of device-specific requirements (quoting Lohr,
518 U.S. at 502)); id. at 175 (explaining that “Justice
Breyer joined in some parts of Justice Stevens’ plurality
opinion (thus making it a majority opinion at times),”
including “in Part V”). But we also read Justice Breyer’s
concurrence as reaching the opposite conclusion, despite
his having joined that portion of the majority opinion.
See id. Faced with an apparent conflict between Part V
of the majority opinion and Justice Breyer’s concurrence,
we followed the latter because it was narrower, just as the
Fifth, Sixth, Seventh, Eighth, and Ninth Circuits had
done. Id. at 175–76; see also Martin v. Medtronic, 

, 581–83 (5th Cir. 2001); Kemp v. Medtronic,

, 230 (6th Cir. 2000); Mitchell v. Collagen
Corp., 

, 911–12 (7th Cir. 1997); Papike v.
Tambrands, Inc., 

, 742 (9th Cir. 1997). In
doing so, we rejected our dissenting colleague’s
argument that the narrowest-grounds approach was
“simply inapplicable” because Justice Breyer joined Part
V of the majority opinion and that the “correct course of
                            40
action” in the event of a conflict “would be to follow Part
V as the majority opinion.” Horn, 376 F.3d at 184 &
n.30 (Fuentes, J., dissenting); see id. at 183 (explaining
that the Horn majority and the Seventh and Ninth
Circuits “also perceived a contradiction and chose to
ignore Justice Breyer’s vote for Part V, instead crediting
the apparently contrary reasoning in his concurrence”).

       Likewise, in United States v. Bishop, 

,
576–77 (3d Cir. 1995), we relied on the narrower
concurring views of Justices Kennedy and O’Connor to
limit the majority’s opinion in United States v. Lopez,

 (1995), which they formally joined as the
fourth and fifth votes. We declined to read the majority
opinion so broadly as to upend judicial deference to
Congress’s judgment about whether an activity
substantially implicates interstate commerce, instead
following the concurrence’s view that the majority had
reached a “necessary though limited holding” that still
“counseled great restraint” before finding that Congress
had transgressed its Commerce Clause power. Bishop,
66 F.3d at 590 (quoting Lopez, 514 U.S. at 568
(Kennedy, J., concurring)). As in Horn, we took that
approach notwithstanding our dissenting colleague’s
argument that we should follow the breadth of the
majority opinion and ignore the narrower concurrence
because “Justices O’Connor and Kennedy joined in the
[majority] opinion.” Id. at 591 (Becker, J., concurring in
part and dissenting in part). As even our dissenting
                            41
colleague explained, we followed the narrower views of
Justices O’Connor and Kennedy because they “form[ed]
an intermediate bloc [of the majority] which would view
Lopez as case-specific.” Id. And Horn and Bishop are
not the only examples. See, e.g., United States v.
Monclavo-Cruz, 

, 1288 (9th Cir. 1981)
(relying on the narrowing construction given to the
majority opinion by Justice Powell, who was also a
necessary member of the majority, to limit the majority’s
holding in South Dakota v. Opperman, 

(1976)); United States v. Wilson, 

, 1164
(8th Cir. 1980) (similar).

      To be sure, the Supreme Court once said—in a
case not involving a linchpin concurrence—that federal
courts should not give “much precedential weight” to a
concurring opinion, even if it coheres with the majority
opinion. Alexander v. Sandoval, 

, 285 n.5
(2001); see also Vasquez v. Hillery, 

, 622
n.4 (1986) (describing the Marks rule as “inapplicable” to
an opinion “to which five Justices expressly
subscribed”). Yet we have already decided that this
principle from Alexander is inapplicable to a concurrence
that (1) “cast the so-called ‘swing vote,’ which was
crucial to the outcome of the case and without which
there could be no majority,” and (2) took a narrower
approach than the majority opinion. Horn, 376 F.3d at
174–75 (distinguishing Alexander on this basis).


                           42
       Which brings us back to Justice Alito’s
concurrence in Morse. The linchpin justices in Morse—
Justices Alito and Kennedy—expressly conditioned their
joining the majority opinion on a narrower interpretation
of the opinion—namely, that it did not permit the
restriction of speech that could plausibly be interpreted as
political or social speech. Had they known that lower
courts would ignore their narrower understanding of the
majority opinion—or had the majority opinion expressly
gone farther than their limitations—then, by their own
admission, they would not have joined the majority
opinion. That would have transformed the five-justice
majority opinion into a three-justice plurality opinion,
with their concurring views becoming the controlling
narrowest grounds under an uncontroversial application
of the Marks doctrine. Why, then, should it matter
whether they formally joined the majority opinion or not?

       It should not. Ignoring limitations placed on the
majority opinion by a necessary member of the majority
would mean that four justices could “fabricate a majority
by binding a fifth to their interpretation of what they say,
even though he writes separately to explain his own more
narrow understanding.” McKoy, 494 U.S. at 462 n.3
(Scalia, J., dissenting). That produces inexplicable
anomalies. If a four-justice plurality holds X and Y, and
a fifth justice “concurs in the judgment” to hold only X
and rejects Y, the fifth member’s more limited views
become binding under a straightforward application of
                            43
Marks. The same interpretation is true if the fifth justice
joins the majority opinion and “concurs in part.” Yet if
the same concurring justice joins the majority opinion
while “concurring,” then the majority opinion holding X
and Y becomes binding and the fifth member’s narrower
views evaporate. Such an approach places all of its
weight on the distinction between a justice’s choice to
follow his name with “concurring” instead of “concurring
in part” or “concurring in the judgment.” Cf. West,
Concurring in Part and Concurring in the Confusion,

. Rev. at 1953–54 (explaining why these
“after the comma” phrases cannot bear such weight);
Tristan C. Pelham-Webb, Note, Powelling for Precedent:
“Binding” Concurrences, 64 N.Y.U. Ann. Surv. Am. L.
693, 737 (2009) (same). That elevates formalism over
substance at the expense of ignoring the very conditions
on which a necessary member of the majority expressly
chose to join the majority.

       In short, because Justice Alito’s concurrence
provides “a single legal standard . . . [that] when properly
applied, produce[s] results with which a majority of the
Justices in the case articulating the standard would
agree,” United States v. Donovan, 

, 182 (3d
Cir. 2011) (alterations in original) (internal quotation
marks and citations omitted), his opinion in Morse forms
the “narrowest grounds necessary to secure a majority,”
Planned Parenthood of Se. Pa. v. Casey, 

,
694 n.7 (3d Cir. 1991), aff’d in part and rev’d in part on
                            44
other grounds, 

 (1992). As a result, we
agree with the en banc Fifth Circuit that the limitations
placed on the majority opinion by Justice Alito’s
concurrence are binding on us.17 See Morgan v.
Swanson, 

, 403 (5th Cir. 2011) (en banc)
(majority opinion of Elrod, J.) (describing Justice Alito’s
Morse concurrence as “controlling”); see also Morgan v.
Plano Indep. Sch. Dist., 

, 746 n.25 (5th Cir.
2009) (“We have held Justice Alito’s concurrence to be
the controlling opinion in Morse.” (citing Ponce, 508
F.3d at 768)).

17
   We have had this same intuition previously. See J.S.,
650 F.3d at 927 (“Notably, Justice Alito’s concurrence in
Morse further emphasizes the narrowness of the Court’s
holding.”). And every court of appeals to address this
question (other than the Seventh Circuit) has shared our
intuition. See Morgan, 589 F.3d at 746 n.25; Barr v.
Lafon, 

, 564 (6th Cir. 2008) (treating Justice
Alito’s concurrence as the basis for Morse’s “narrow
holding”); Corder v. Lewis Palmer Sch. Dist. No. 38, 

, 1228 (10th Cir. 2009) (same). The Seventh
Circuit concluded, without citation or support, that the
narrowest-grounds approachdoes not apply where there is
a majority opinion, as in Morse.Nuxoll ex rel. Nuxoll v.
Indian Prairie Sch. Dist. No. 204, 

, 673 (7th
Cir. 2008). But as we explain, we have already rejected
the Seventh Circuit’s formalist approach when it was
urged by dissenting colleagues in Horn and Bishop.
                            45
       Justice Alito would have protected political or
social speech reasonably interpreted to advocate illegal
drug use, and that protection applies even more strongly
to ambiguously lewd speech. In Morse, the Court added
a new categorical exception to Tinker: student speech
that a reasonable observer could interpret as advocating
illegal drug use but that cannot plausibly be interpreted as
addressing political or social issues. Id. at 422. The
exception was justified because illegal drugs pose an
“immediately obvious,” “grave” and “unique threat to the
physical safety of students.” Id. at 425. Despite that
threat, however, the Court held that speech advocating
illegal drug use is not categorically unprotected if it “can
plausibly be interpreted as commenting on any political
or social issue, including speech on issues such as the
wisdom of the war on drugs or of legalizing marijuana
for medicinal use.” Id. at 422 (internal quotation marks
omitted). Even with that limitation, the Court made clear
that this new exception to Tinker “stand[s] at the far
reaches of what the First Amendment permits.” Id. at
425.
      If speech posing such a “grave” and “unique threat
to the physical safety of students” can be categorically
regulated only when it cannot “plausibly be interpreted as
commenting on any political or social issue”—and that
regulation nonetheless “stand[s] at the far reaches of
what the First Amendment permits”—then there is no
reason why ambiguously lewd speech should receive any
                            46
less protection when it also “can plausibly be interpreted
as commenting on any political or social issue.” Id. at
422, 425. One need not be a philosopher of Mill or
Feinberg’s stature18 to recognize that harmful speech
posing an “immediately obvious” threat to the “physical
safety of students,” id. at 425, presents a far graver threat
to the educational mission of schools—thereby
warranting less protection—than ambiguously lewd
speech that might undercut teaching “the appropriate
form of civil discourse” to students, Fraser, 478 U.S. at
683. It would make no sense to afford a T-shirt
exclaiming “I ♥ pot! (LEGALIZE IT)” protection under
Morse while declaring that a bracelet saying “I ♥
boobies! (KEEP A BREAST)” is unprotected under
Fraser.

      Those limits are persuasive on their own terms,
even if we disregard the controlling limitations of Justice
Alito’s Morse concurrence.          Fraser reflects the
longstanding notions that “not all speech is of equal First

18
   John Stuart Mill and Joel Feinberg are both known for,
among other things, their groundbreaking work on the
relationship between harm and offense and how conduct
of each type might be subject to criminalization. See
generally Joel Feinberg, Harm to Others: The Moral
Limits of the Criminal Law (1984); Joel Feinberg,
Offense to Others: The Moral Limits of the Criminal Law
(1985); John Stuart Mill, On Liberty (1859).
                             47
Amendment importance” and that “speech on matters of
public concern . . . is at the heart of the First
Amendment’s protection.” Snyder v. Phelps, 

, 1215 (2011) (quotation marks and citations
omitted); see also Connick v. Myers, 

, 145
(1983) (“[S]peech on public issues occupies the highest
rung of the hierarchy of First Amendment values, and is
entitled to special protection.” (internal quotation marks
and citations omitted)). And it is only a limited
exception to the otherwise “bedrock principle” of the
First Amendment that “the government may not prohibit
the expression of an idea simply because society finds the
idea itself offensive or disagreeable.” Texas v. Johnson,

, 414 (1989); see also Sable Commc’ns of
Cal., Inc. v. FCC, 

, 126 (1989) (“Sexual
expression which is indecent but not obscene is protected
by the First Amendment.”). The Supreme Court has
never held that schools may bore willy-nilly through that
bedrock principle. But it has made clear that “minors are
entitled to a significant measure of First Amendment
protection” and the government does not “have a free-
floating power to restrict the ideas to which children may
be exposed.” Brown v. Entm’t Merchs. Ass’n, 

, 2736 (2011). To be sure, Fraser rejected the idea
that “simply because an offensive form of expression
may not be prohibited to adults making what the speaker
considers a political point, the same latitude must be
permitted to children in a public school.” Fraser, 478
U.S. at 682. As we have explained, though, Fraser was
                           48
limited to plainly lewd speech, and that refusal to protect
a student’s plainly lewd speech where the same speech
by an adult would be protected does not extend to
political speech that is not plainly lewd. On that score,
our conclusion puts us in good company with five
justices in Morse19 who were expressly unwilling to
permit a categorical exception to Tinker that would
intrude on political or social speech and two justices20

19
    In addition to Justices Alito and Kennedy, three
dissenting justices (Justices Stevens, Souter, and
Ginsburg) would not have extended the Morse exception
to political or social speech. These five justices instead
split over whether Morse’s speech could reasonably be
interpreted as advocating illegal drug use. Morse, 551
U.S. at 444, 448 (Stevens, J., dissenting) (concluding that
Morse’s banner is constitutionally protected because it
could not reasonably be interpreted as advocating illegal
drug use and was at most a “minority[] viewpoint” in
“the national debate about a serious issue” deserving
First Amendment protection).
20
   In the majority opinion, Chief Justice Roberts and
Justice Scalia refused to “stretch[] Fraser” so far as to
“encompass any speech that could fit under some
definition of ‘offensive’” specifically to protect “political
and religious speech [that] might be perceived as
offensive to some.” Morse, 551 U.S. at 409; see also id.
at 403 (majority opinion) (“But not even Frederick
argues that the banner conveys any sort of political or
                             49
who all but said as much.
       What’s more, this limitation is consistent with our
previous intuitions as well as those of the Sixth and
Second Circuits. See Saxe, 240 F.3d at 213 (Alito, J.)
(noting that the “dichotomy” between Fraser and Tinker
is “neatly illustrated by the comparison between Cohen’s
[“Fuck the Draft”] jacket and Tinker’s armband”); Defoe,
625 F.3d at 335 n.6 (rejecting the Eleventh Circuit’s
extension of Fraser to displays of the Confederate flag
and instead holding that such displays “by students [are]
protected political speech that school officials may only
regulate by satisfying the Tinker standard” (citing Barr v.
Lefon, 538 F.3d. 554, 569 n.7 (6th Cir. 2008))); Guiles,


religious message. Contrary to the dissent’s suggestion,
this is plainly not a case about political debate over the
criminalization of drug use or possession.”); id. at 406
n.2 (“[T]here is no serious argument that Frederick’s
banner is political speech . . . .”). Although Justice
Thomas joined that portion of the majority opinion, he
would have concluded that “the First Amendment, as
originally understood, does not protect student speech in
public schools” and overruled Tinker. Id. at 410–11
(Thomas, J., concurring). Justice Breyer would have
avoided the “difficult First Amendment issue” and
concluded that “qualified immunity bars [Morse’s] claim
for monetary damages.”         Id. at 425 (Breyer, J.,
concurring in the judgment in part and dissenting in part).
                            50
461 F.3d at 325 (holding Fraser inapplicable because the
T-shirt was not “as plainly offensive as the sexually
charged speech considered in Fraser . . . [,] especially
when considering that [it was] part of an anti-drug
political message”).
       Consequently, we hold that the Fraser exception
does not permit ambiguously lewd speech to be
categorically restricted if it can plausibly be interpreted
as political or social speech.

      3. Under Fraser, schools may restrict plainly
         lewd speech regardless of whether it could
         plausibly be interpreted as social or political
         commentary.
      As the Supreme Court made clear in Fraser,
though, schools may restrict plainly lewd speech
regardless of whether it could plausibly be interpreted to
comment on a political or social issue. Fraser, 478 U.S.
at 682 (“[T]he First Amendment gives a high school
student the classroom right to wear Tinker’s armband,
but not Cohen’s [“Fuck the Draft”] jacket.”). That is true
by definition. Plainly lewd speech “offends for the same
reasons obscenity offends” because the speech in that
category is “no essential part of any exposition of ideas”
and thus carries very “slight social value.” Id. at 683
(quoting Pacifica Found., 438 U.S. at 746 (plurality
opinion)). As with obscenity in general, obscenity to
minors, and all other historically unprotected categories
                            51
of speech, “the evil to be restricted so overwhelmingly
outweighs the expressive interests, if any, at stake, that
no process of case-by-case adjudication is required”
because “the balance of competing interests is clearly
struck.” Stevens, 130 S. Ct. at 1585–86 (quoting New
York v. Ferber, 

, 763–64 (1982)). In other
words, we do not engage in a case-by-case determination
of whether obscenity to minors—and by extension,
plainly lewd speech under Fraser—carries social value.
As a result, schools may continue to regulate plainly
lewd, vulgar, profane, or offensive speech under Fraser
even if a particular instance of such speech can
“plausibly be interpreted as commenting on any political
or social issue.” Morse, 551 U.S. at 422 (Alito, J.,
concurring).

       In response, the School District recites a mantra
that has Fraser providing schools the ultimate discretion
to define what is lewd and vulgar. It relies on the
Supreme Court’s sentiment that schools may define their
“basic educational mission” and prohibit student speech
that is inconsistent with that mission. Kuhlmeier, 484
U.S. at 266–67.21 Indeed, before Morse, some courts of

21
  See also Fraser, 478 U.S. at 683 (“[T]he determination
of what manner of speech in the classroom or in school
assembly is inappropriate properly rests with the school
board.”); Pico, 457 U.S. at 864 (“[F]ederal courts should
not ordinarily ‘intervene in the resolution of conflicts
                           52
appeals adopted that broad interpretation of the Supreme
Court’s student-speech cases. See, e.g., LaVine v. Blaine
Sch. Dist., 

, 988 (9th Cir. 2001) (“[A]
school need not tolerate student speech that is
inconsistent with its basic educational mission.”); Boroff
v. Van Wert City Bd. of Educ., 

, 470 (6th
Cir. 2000) (“[W]here Boroff’s T-shirts contain symbols
and words that promote values that are so patently
contrary to the school’s educational mission, the School
has the authority, under the circumstances of this case, to
prohibit those T-shirts [under Fraser].”).
      Whatever the face value of those sentiments, such
sweeping and total deference to school officials is
incompatible with the Supreme Court’s teachings. In
Tinker, Hazelwood, and Morse, the Supreme Court
independently evaluated the meaning of the student’s
speech and the reasonableness of the school’s


which arise in the daily operation of school systems.’”
(quoting Epperson v. Arkansas, 

, 104
(1968))); Wood v. Strickland, 

, 326 (1975)
(“It is not the role of the federal courts to set aside
decisions of school administrators which the court may
view as lacking a basis in wisdom or compassion.”); see
also Kuhlmeier, 484 U.S. at 273 (“[T]he education of the
Nation’s youth is primarily the responsibility of parents,
teachers, and state and local school officials, and not of
federal judges.”).
                            53
interpretation and actions. There is no reason the
school’s authority under Fraser should receive special
treatment. More importantly, such an approach would
swallow the other student-speech cases, including Tinker,
effectively eliminating judicial review of student-speech
restrictions. See Guiles, 461 F.3d at 327 (making this
point). That is precisely why the Supreme Court in
Morse explicitly rejected total deference to school
officials:
      The opinion of the Court does not endorse the
      broad argument advanced by petitioners and
      the United States that the First Amendment
      permits public school officials to censor any
      student speech that interferes with a school's
      “educational mission.” . . . The “educational
      mission” argument would give public school
      authorities a license to suppress speech on
      political and social issues based on
      disagreement with the viewpoint expressed.
      The argument, therefore, strikes at the very
      heart of the First Amendment.
Morse, 551 U.S. at 423 (Alito, J., concurring).

       Instead, Morse settled on a narrower view of
deference, deferring to a school administrator’s
“reasonable judgment that Frederick’s sign qualified as
drug advocacy” only if the speech could not plausibly be
interpreted as commenting on a political or social issue.
                            54
Morse, 551 U.S. at 441 (Stevens, J., dissenting); see also
id. at 408 (majority opinion) (“[S]chools [may] restrict
student expression that they reasonably regard as
promoting illegal drug use.”); id. at 422 (Alito, J.,
concurring) (“[A] public school may restrict speech that a
reasonable observer would interpret as advocating illegal
drug use . . . .”). Our approach to lewd speech provides
the same degree of deference to schools as the Court did
in Morse. We defer to a school’s reasonable judgment
that an observer could interpret ambiguous speech as
lewd, vulgar, profane, or offensive only if the speech
could not plausibly be interpreted as commenting on a
political or social issue.
       The School District invokes a parade of horribles
that, in its view, would follow from our framework:
protecting ambiguously lewd speech that comments on
political or social issues—like the bracelets in this case—
will encourage students to engage in more egregiously
sexualized advocacy campaigns, which the schools will
be obliged to allow. See Pa. Sch. Bd. Ass’n Amicus Br.
in Supp. of Appellant at 19 (listing examples, including
“I ♥ Balls!” apparel for testicular cancer, and “I ♥ Va Jay
Jays” apparel for the Human Papillomaviruses); App.
275–76 (raising the possibility of apparel bearing the
slogans “I ♥ Balls!” or “I ♥ Titties!”). Like all slippery-
slope arguments, the School District’s point can be
inverted with equal logical force. If schools can
categorically regulate terms like “boobies” even when the
                            55
message comments on a social or political issue, schools
could eliminate all student speech touching on sex or
merely having the potential to offend. See Frederick
Schauer, Slippery Slopes, 99 Harv. L. Rev. 361, 381
(1985) (“[I]n virtually every case in which a slippery
slope argument is made, the opposing party could with
equal formal and linguistic logic also make a slippery
slope claim.”). The ease of turning a slippery-slope
argument on its head explains why the persuasiveness of
such a contention does not depend on its logical validity.
Id. Instead, the correctness of a slippery-slope argument
depends on an empirical prediction that a proposed rule
will increase the likelihood of some other undesired
outcome occurring. Id. (“To some people, one argument
will seem more persuasive than the other because the
underlying empirical reality . . . makes one equally
logical possibility seem substantially more likely to occur
than the other.”); see also Eugene Volokh, The
Mechanism of the Slippery Slope, 116 Harv. L. Rev.
1026, 1066–71 (2003) (making a similar point in the
context of extending precedent). Because courts usually
lack the data necessary for such a prediction, “fear of . . .
what’s at the bottom of a long, slippery slope is not a
good reason for today’s decision.” Marozsan v. United
States, 

, 1499 (7th Cir. 1988) (en banc)
(Easterbrook, J., dissenting). “The terror of extreme
hypotheticals produces much bad law,” and so our
answer to the School District’s “extreme hypothetical[s]”
is that we will “cross that bridge when we come to it.”
                             56
Id.
       To make matters worse, the School District has
greased the supposedly slippery slope by omitting any
empirical evidence. We have no reason to think either
that the parents of middle-school students will be willing
to allow their children to wear apparel advocating
political or social messages in egregious terms or that a
student will overcome the typical middle-schooler’s
embarrassment, immaturity, and social pressures by
wearing such apparel. And many of the School District’s
hypotheticals pose no worries under our framework. A
school could categorically restrict an “I ♥ tits! (KEEP A
BREAST)” bracelet because, as the Supreme Court
explained in Pacifica, the word “tits” (and also
presumably the diminutive “titties”) is a patently
offensive reference to sexual organs and thus obscene to
minors. See Pacifica Found., 438 U.S. at 745–46
(plurality opinion) (explaining that the comedian George
Carlin’s seven “dirty” words, which includes “tits,”
“offend for the same reasons that obscenity offends”);
see also LaVine, 257 F.3d at 989 (concluding that a poem
“filled with imagery of violent death and suicide” was
not “vulgar, lewd, obscene, or plainly offensive because
it was “not ‘an elaborate, graphic, and explicit sexual
metaphor’ as was the student’s speech in Fraser, nor
[did] it contain the infamous seven words that cannot be
said on the public airwaves”); cf. FCC v. Fox Television
Stations, Inc., 

, 517–18 (2009) (concluding
                           57
it was not arbitrary or capricious for the FCC to regulate
even “isolated uses of sexual and excretory words,”
including Carlin’s seven “dirty” words, because “[e]ven
isolated utterances can be made in pander[ing], . . .
vulgar and shocking manners” and can thus “constitute
harmful first blow[s] to children” (alterations in
original)). The same is true of a student’s drawings of
stick figures in sexual positions, even if used to promote
contraceptive use. Cf. R.O. ex rel. Ochshorn City Sch.
Dist., 

, 543 (2d Cir. 2011). And even if
students engage in more questionable speech, the school
retains the government’s normal sovereign authority to
regulate speech as well as its additional powers as
educator to restrict speech under Tinker, Kuhlmeier, and
Morse. See, e.g., Hardwick v. Heyward, 

,
440 (4th Cir. 2013) (holding that a school’s prohibition
on wearing T-shirts depicting the Confederate battle flag
was permissible under Tinker because of a history of
racial tension and disruptions related to the Confederate
flag).
       By contrast, there is empirical support for the
opposite worry. Some schools, if empowered to do so,
might eliminate all student speech touching on sex or
merely having the potential to offend. Indeed, the
Middle School’s administrators seemed inclined to do
just that. They initially testified that they could ban the
word “breast,” even if used in the context of a breast-
cancer-awareness campaign, because the word, by itself,
                            58
“can be construed as [having] a sexual connotation.”
App. 490, 497. If anything, the fear of a slippery slope
cuts against the School District.

       In a similar vein, we need not speculate on
context-dependent hypotheticals to give guidance to
schools and district courts. The fault lines of our
framework are adequately mapped out in the rest of First
Amendment jurisprudence.          The Supreme Court’s
obscenity-to-minors case law marks the contours of
plainly lewd speech. See, e.g., Brown v. Entm’t Merchs.
Ass’n, 

, 2735 (refusing to extend the
categorical nonprotection for obscenity to minors to
speech that is violent from a minor’s perspective);
Ginsberg, 390 U.S. at 638 (approving a state prohibition
on selling minors sexual material that would be obscene
from the minor’s perspective).             Those contours
necessarily admit of some flexibility and can be
“adjust[ed] . . . ‘to social realities by permitting the
[sexual] appeal of this type of material to be assessed”
from the minors’ perspective.          Id.; see also Fox
Television Stations, Inc., 556 U.S. at 520 (explaining that
based on the obscenity-to-minors case law, the FCC
properly “dr[aws] distinctions between the offensiveness
of particular words based upon the context in which they
appeared” on case-by-case basis without having to rely
on empirical evidence as to the degree of offensiveness).
And the government is not a stranger to determining
whether speech plausibly comments on a political or
                            59
social issue. For that, we look to case law on whether
speech involves a matter of public concern. See, e.g.,
Garcetti v. Ceballos, 

, 418 (2006)
(“Pickering and the cases decided in its wake identify
two inquiries to guide interpretation of the constitutional
protections accorded to public employee speech. The
first requires determining whether the employee spoke as
a citizen on a matter of public concern. . . . If the answer
is yes, then the possibility of a First Amendment claim
arises.”). Of course, these rules lack “perfect clarity”—
just as every legal rule contains fuzzy borders. Brown,
131 S. Ct. at 2764 (Breyer, J., dissenting); cf. United
States v. Williams, 

, 304 (2008) (“[P]erfect
clarity and precise guidance have never been required
even of regulations that restrict expressive activity.”).
Even so, just because a “precise standard” for political
speech or plain lewdness (obscenity to minors) “proves
elusive,” it is still “easy enough to identify instances that
fall within a legitimate regulation.” Brown, 131 S. Ct. at
2764 (Breyer, J., dissenting). Over time, the fault lines
demarcating plainly lewd speech and political or social
speech will settle and become more rule-like as precedent
accumulates.

       To recap: Under the government’s sovereign
authority, a school may categorically ban obscenity,
fighting words, and the like in schools; the student-
speech cases do not supplant the government’s sovereign
powers to regulate speech. See, e.g., Doe v. Pulaski
                             60
Cnty. Special Sch. Dist., 

, 626, 626–27 (8th
Cir. 2002) (en banc) (holding that the government, as K-
12 educator, could punish a student for making a true
threat); Cuff ex rel. B.C. v. Valley Cent. Sch. Dist., 

, 118 (2d Cir. 2012) (Pooler, J., dissenting)
(“Indeed, despite the expansion of school-specific
exceptions to the First Amendment’s general prohibition
against government restrictions on speech, certain well-
settled rules apply to adults and adolescents alike.”).
Under Fraser, a school may categorically restrict plainly
lewd, vulgar, or profane speech that “offends for the
same reasons obscenity offends” regardless of whether it
can plausibly be interpreted as commenting on social or
political issues. Saxe, 240 F.3d at 213 (quoting Fraser,
478 U.S. at 685). As we have explained, see supra at
20–21, plainly lewd speech cannot, by definition, be
plausibly interpreted as political or social commentary
because the speech offends for the same reason obscenity
offends and thus has slight social value. Fraser also
permits a school to categorically restrict ambiguous
speech that a reasonable observer could interpret as
having a lewd, vulgar, or profane meaning so long as it
could not also plausibly be interpreted as commenting on
a social or political issue. But Fraser does not permit a
school to categorically restrict ambiguous speech that a
reasonable observer could interpret as having a lewd,
vulgar, or profane meaning and could plausibly interpret
as commenting on a social or political issue. And of
course, if a reasonable observer could not interpret the
                           61
speech as lewd, vulgar, or profane, then Fraser simply
does not apply. As always, a school’s other powers over
student speech under Tinker, Kuhlmeier, and Morse
remain as a backstop.

C.    The Middle School’s ban on “I ♥ boobies!
      (KEEP A BREAST)” bracelets
       Under this framework, the School District’s
bracelet ban is an open-and-shut case. The “I ♥ boobies!
(KEEP A BREAST)” bracelets are not plainly lewd. The
slogan bears no resemblance to Fraser’s “pervasive
sexual innuendo” that was “plainly offensive to both
teachers and students.” Fraser, 478 U.S. at 683.
Teachers had to request guidance about how to deal with
the bracelets, and school administrators did not conclude
that the bracelets were vulgar until B.H. and K.M. had
worn them every day for nearly two months. In addition,
the Middle School used the term “boobies” in
announcing the bracelet ban over the public address
system and the school television station. What’s more,
the bracelets do not contain language remotely akin to the
seven words that are considered obscene to minors on
broadcast television. Pacifica Found., 438 U.S. at 745–
46 (plurality opinion); LaVine, 257 F.3d at 989
(concluding that speech was not vulgar, lewd, obscene, or
plainly offensive because it was “not ‘an elaborate,
graphic, and explicit sexual metaphor’ as was the
student’s speech in Fraser, nor [did] it contain the

                           62
infamous seven words that cannot be said on the public
airwaves” under Pacifica). Indeed, the term “boobie” is
no more than a sophomoric synonym for “breast.” And
as the School District also concedes, a reasonable
observer would plausibly interpret the bracelets as part of
a national breast-cancer-awareness campaign, an
undeniably important social issue. Oral Arg. Tr. at
10:11–16; see also K.J. ex rel. Braun v. Sauk Prairie Sch.
Dist., No. 11-CV-622, slip op. at 14 (W.D. Wis. Feb. 6,
2012) (“When one reads the entire phrase, it is clearly a
message designed to promote breast cancer awareness.”).
Accordingly, the bracelets cannot be categorically
banned under Fraser.22

                           IV.
       Fraser, of course, is only one of four school-
specific avenues for regulating student speech.23 The
22
   Because we conclude that the slogan is not plainly
lewd and is plausibly interpreted as commenting on a
social issue, the bracelets are protected under Fraser. As
a result, we need not determine whether a reasonable
observer could interpret the bracelets’ slogan as lewd.
23
   As the Supreme Court has recently reaffirmed, there
might be other exceptions to Tinker that have not yet
been identified by the courts. See Morse, 551 U.S. at
408–09 (identifying a new exception to the Tinker
framework for speech that is reasonably interpreted as
advocating illegal drug use and that is not plausibly
                            63
parties rightly agree that Kuhlmeier and Morse do not
apply: no one could reasonably believe that the Middle
School was somehow involved in the morning fashion
decisions of a few students, and no one could reasonably
interpret the bracelets as advocating illegal drug use.
       That leaves only Tinker as possible support for the
School District’s ban. Under Tinker’s “general rule,” the
government may restrict school speech “that threatens a
specific and substantial disruption to the school
environment” or “inva[des] . . . the rights of others.”
Saxe, 240 F.3d at 211 (citing Tinker, 393 U.S. at 504).
“[I]f a school can point to a well-founded expectation of
disruption—especially one based on past incidents
arising out of similar speech—the restriction may pass


interpreted as commenting on any political or social
issue). Compare id. at 405 (“Fraser established that the
mode of analysis set forth in Tinker is not absolute.”),
and id. at 406 (“And, like Fraser, [Kuhlmeier] confirms
that the rule of Tinker is not the only basis for restricting
student speech.”), with id. at 423 (Alito, J., concurring)
(“I join the opinion of the Court on the understanding
that the opinion does not hold that the special
characteristics of the public schools necessarily justify
any other speech restrictions.” (emphasis added)). Here,
however, the School District relies solely on the existing
school-speech framework and does not propose any new
bases for restricting student speech.
                             64
constitutional muster.” Id. at 212; J.S. v. Blue Mountain
Sch. Dist., 

, 928 (3d Cir. 2011) (en banc)
(“[T]he School District need not prove with absolute
certainty that substantial disruption will occur.”). The
School District has the burden of showing that the
bracelet ban is constitutional under Tinker. See J.S., 650
F.3d at 928. That it cannot do.

       Tinker meant what it said: “a specific and
significant fear of disruption, not just some remote
apprehension of disturbance.” Id. Tinker’s black
armbands did not meet this standard, even though the
armbands “caused comments, warnings by other
students, the poking of fun at them, . . . a warning by an
older football player that other, nonprotesting students
had better let them alone,” and the “wreck[ing]” of a
math teacher’s lesson period. Tinker, 393 U.S. at 517
(Black, J., dissenting).

      Here, the record of disruption is even skimpier.
When the School District announced the bracelet ban, it
had no more than an “undifferentiated fear or remote
apprehension of disturbance.” Sypniewski, 307 F.3d at
257. The bracelets had been on campus for at least two
weeks without incident. B.H., 

 at 408;
see also App. 13 (“[N]one of the three principals had
heard any reports of disruption or student misbehavior
linked to the bracelets. Nor had any of the principals
heard reports of inappropriate comments about

                           65
‘boobies.’”). That track record “speaks strongly against a
finding of likelihood of disruption.” Sypniewski, 307
F.3d at 254.

       The School District instead relies on two incidents
that occurred after the ban. In one, a female student told
a teacher that she believed some boys had remarked to
girls about their “boobies” in relation to the bracelets—
an incident that was never confirmed. B.H., 

 at 408. In the other, two female students were
discussing the bracelets during lunch, and a boy
interrupted them to say “I want boobies” while “making
inappropriate gestures with two spherical candies.” Id.
The boy was suspended for a day. Id.
       Even assuming that disruption arising after a
school’s speech restriction could satisfy Tinker—a
question we need not decide today—these two isolated
incidents hardly bespeak a substantial disruption caused
by the bracelets. “[S]tudent expression may not be
suppressed simply because it gives rise to some slight,
easily overlooked disruption, including but not limited to
‘a showing of mild curiosity’ by other students,
‘discussion and comment’ among students, or even some
‘hostile remarks’ or ‘discussion outside of the
classrooms’ by other students.” Holloman ex rel.
Holloman v. Harland, 

, 1271–72 (11th Cir.
2004) (internal quotation marks and citations omitted).
Given that Tinker’s black armband—worn to protest a

                           66
controversial war and divisive enough to prompt
reactions from other students—was not a substantial
disruption, neither is the “silent, passive expression” of
breast-cancer awareness.24 Tinker, 393 U.S. at 508. If

24
   According to B.H. and K.M., Tinker’s substantial-
disruption standard does not permit a school to restrict
speech because of the heckler’s veto of other students’
disruptive reactions. See Appellees’ Br. at 35 (emphasis
added). Because no forecast of substantial disruption
would be reasonable on this record under any meaning of
that term, we need not determine the precise interplay
between the anti-heckler’s veto principle present
elsewhere in free-speech doctrine and Tinker’s
substantial-disruption standard in public schools.
Compare Zamecnik, 636 F.3d at 879 (noting that Tinker
endorsed both the heckler’s veto doctrine and the
substantial-disruption test and concluding that other
students’ harassment of “Zamecnik because of their
disapproval of her [“Be Happy, Not Gay” T-shirt] is not a
permissible ground for banning it”), and Holloman, 370
F.3d at 1275–76 (interpreting Tinker as endorsing an
anti-heckler’s veto principle, concluding that “[w]hile the
same constitutional standards do not always apply in
public schools as on public streets, we cannot afford
students less constitutional protection simply because
their peers might illegally express disagreement through
violence instead of reason”), with Taylor v. Roswell
Indep. Sch. Dist., 

, 38 (10th Cir. Apr. 8,
                            67
anything, the fact that these incidents did not occur until
after the School District banned the bracelets suggests
that the ban “exacerbated rather than contained the
disruption in the school.” J.S., 650 F.3d at 931 (drawing
this same conclusion on a similar record).
       Undeterred, the School District invokes the other
half of Tinker’s general rule, arguing that the bracelets
invade other students’ Title IX rights to be free from
sexual harassment. See Tinker, 393 U.S. at 513. Under
Title IX, students may sue federally-funded schools that
“act[] with deliberate indifference” to “harassment that is
so severe, pervasive, and objectively offensive . . . that
the victim students are effectively denied equal access to
an institution’s resources and opportunities.” Saxe, 240
F.3d at 205–06 (quoting Davis ex rel. LaShonda D. v.
Monroe Cnty. Bd. of Educ., 

, 651 (1999)).
According to the School District, the “I ♥ boobies!
(KEEP A BREAST)” bracelet was “deemed
inappropriate for school due to the likelihood of a
resultant increase in student-on-student sexual
harassment.” Sch. Dist.’s Br. at 54.


2013) (“Plaintiffs note that most disruptions occurred
only because of wrongful behavior of third parties and
that no Plaintiffs participated in these activities. . . . This
argument might be effective outside the school context,
but it ignores the ‘special characteristics of the school
environment.’” (quoting Tinker, 393 U.S. at 506)).
                              68
       That argument suffers from several flaws, not the
least of which is the School District’s failure to raise it in
the District Court and that Court’s consequent failure to
address it. Freeman v. Pittsburgh Glass Works, LLC,

, 249 (3d Cir. 2013) (“We generally refuse
to consider issues that the parties have not raised below.”
(citing Singleton v. Wulff, 

, 120 (1976))).
But there is an even more basic reason why the School
District’s invocation of Title IX is not the shield it claims
to be. Even assuming that protecting students from
harassment under Title IX would satisfy Tinker’s rights-
of-others prong,25 the School District does not explain

25
   As we have repeatedly noted, “the precise scope of
Tinker’s ‘interference with the rights of others’ language
is unclear.” Saxe, 240 F.3d at 217 (quoting Tinker, 393
U.S. at 504); DeJohn v. Temple Univ., 

, 319
(3d Cir. 2008). And the Supreme Court has “never
squarely addressed whether harassment, when it takes the
form of pure speech, is exempt from First Amendment
protection.” Saxe, 240 F.3d at 207. We need not address
either of these points today. Even if Tinker permits
school regulation of pure speech that would constitute
“harassment” under Title IX, the School District has not
offered any explanation or evidence of how passively
wearing the “I ♥ boobies! (KEEP A BREAST)” bracelets
would create such a severe and pervasive environment in
the Middle School. Cf. Saxe, 240 F.3d at 204 (Alito, J.)
(“There is no categorical ‘harassment exception’ to the
                             69
why the bracelets would breed an environment of
pervasive and severe harassment. See, e.g., DeJohn v.
Temple Univ., 

, 320 (3d Cir. 2008)
(“[U]nless harassment is qualified with a standard akin to
a severe or pervasive requirement, [an anti-]harassment
policy may suppress core protected speech.”); Saxe, 240
F.3d at 217 (rejecting a school district’s similar argument
that it could ban speech creating a “hostile environment”
without showing that the particular speech covered by the
policy would create a severe or pervasive environment);
see also Nuxoll ex rel. Nuxoll v. Indian Prairie Sch. Dist.
No. 204, 

, 676 (7th Cir. 2008) (“[I]t is
highly speculative that allowing the plaintiff to wear a T-
shirt that says “Be Happy, Not Gay” would have even a
slight tendency to provoke such incidents [of student-on-
student harassment], or for that matter to poison the
educational atmosphere.”).

      The bracelet ban cannot be upheld on the authority
of Tinker.
                            V.
       Because the School District’s ban cannot pass
scrutiny under Fraser or Tinker, B.H. and K.M. are likely
to succeed on the merits. In light of that conclusion, the

First Amendment’s free speech clause.”); Rodriguez v.
Maricopa Cnty. Cmty. College Dist., 

, 708
(9th Cir. 2010) (agreeing with Saxe’s statement).
                            70
remaining preliminary-injunction factors also favor them.
The ban prevents B.H. and K.M. from exercising their
right to freedom of speech, which “unquestionably
constitutes irreparable injury.” K.A. ex rel. Ayers v.
Pocono Mountain Sch. Dist., 

, 113 (3d Cir.
2013) (quoting Elrod v. Burns, 

, 373 (1976)
(plurality opinion)). An after-the-fact money judgment
would hardly make up for their lost opportunity to wear
the bracelets in school. See Elrod, 427 U.S. at 374 n.29
(“The timeliness of political speech is particularly
important.”).
       And the preliminary injunction does not “result in
even greater harm to” the School District, the non-
moving party. Allegheny Energy, Inc. v. DQE, Inc., 

, 158 (3d Cir. 1999). The School District
complains that unless the bracelet ban stands, it “has no
clear guidance” on how to enforce its dress code.
Appellant’s Br. at 60. But the injunction addresses only
the School District’s ban of the “I ♥ boobies! (KEEP A
BREAST)” bracelets. It does not enjoin the School
District’s regulation of other types of apparel, such as the
“Save the ta-tas” T-shirt or testicular-cancer-awareness
apparel bearing the phrase “feelmyballs.org.” Whether
the injunction stays or goes, the School District will have
to continue making individualized assessments of
whether it may restrict student speech consistent with the
First Amendment, just as school administrators have
always had to do. See, e.g., Castorina ex rel. Rewt v.
                            71
Madison Cnty. Sch. Bd., 

, 543 (6th Cir.
2001) (“The foregoing discussion of the three Supreme
Court . . . cases demonstrates the importance of the
factual circumstances in school speech cases . . . .”). The
District Court’s injunction against the bracelet ban does
not change that.

       Lastly, granting the preliminary injunction furthers
the public interest. The School District argues that the
injunction eliminates its “authority to manage its student
population” and thus harms the public. Appellant’s Br. at
61. Again, that hyperbolic protest ignores the narrow
breadth of the injunction, which addresses only the
constitutionality of the bracelet ban under the facts of this
case.       More importantly, allowing a school’s
unconstitutional speech restriction to continue
“vindicates no public interest.” K.A., 

,
at *11 (citation omitted). For these reasons, the District
Court did not abuse its discretion by enjoining the School
District’s bracelet ban.
                *      *     *      *      *
      School administrators “have a difficult job,” and
we are well-aware that the job is not getting any easier.
Morse, 551 U.S. at 409. Besides the teaching function,
school administrators must deal with students distracted
by cell phones in class and poverty at home, parental
under- and over-involvement, bullying and sexting,
preparing students for standardized testing, and ever-
                             72
diminishing funding. When they are not focused on
those issues, school administrators must inculcate
students with “the shared values of a civilized social
order.” Fraser, 478 U.S. at 683; see also McCauley v.
Univ. of the V.I., 

, 243 (3d Cir. 2010)
(quoting Brown v. Bd. of Educ., 

, 493
(1954)) (“Public elementary and high school education is
as much about learning how to be a good citizen as it is
about multiplication tables and United States history.”).
       We do not envy those challenges, which require
school administrators “to make numerous difficult
decisions about when to place restrictions on speech in
our public schools.” Morgan v. Swanson, 

,
420 (5th Cir. 2011) (en banc) (majority opinion of Elrod,
J.). And the School District in this case was not
unreasonably concerned that permitting “I ♥ boobies!
(KEEP A BREAST)” bracelets in this case might require
it to permit other messages that were sexually oriented in
nature. But schools cannot avoid teaching our citizens-
in-training how to appropriately navigate the
“marketplace of ideas.” Just because letting in one idea
might invite even more difficult judgment calls about
other ideas cannot justify suppressing speech of genuine
social value. Tinker, 393 U.S. at 511 (“The classroom is
peculiarly the ‘marketplace of ideas.’ The Nation’s
future depends upon leaders trained through wide
exposure to that robust exchange of ideas which
discovers truth ‘out of a multitude of tongues,’ (rather)
                           73
than through any kind of authoritative selection.’”
(quoting Keyishian v. Bd. of Regents of Univ. of State of
N.Y., 

, 603 (1967))); see id. at 511
(“[S]chool officials cannot suppress ‘expressions of
feelings with which they do not wish to contend.’”
(citation omitted)).

      We will affirm the District Court’s order granting a
preliminary injunction.




                           74
HARDIMAN, Circuit Judge, dissenting with whom
CHAGARES, JORDAN, GREENAWAY, JR., and
GREENBERG, join.

       Today the Court holds that twelve-year-olds have a
constitutional right to wear in school a bracelet that says ―I ♥
boobies! (KEEP A BREAST).‖ Because this decision is
inconsistent with the Supreme Court‘s First Amendment
jurisprudence, I respectfully dissent.

                               I

       My colleagues conclude that the Supreme Court‘s
decision in Bethel School District No. 403 v. Fraser, 

 (1986), cannot justify the Easton Area School District‘s
bracelet ban ―because [the bracelets] comment on a social
issue.‖ Maj. Typescript at 6. This limitation on the ability of
schools to regulate student speech that could reasonably be
deemed lewd, vulgar, plainly offensive, or constituting sexual
innuendo finds no support in Fraser or its progeny. The
Majority‘s ―high value speech‖ modification of Fraser is
based on the following two premises it derives from the
Supreme Court‘s decision in Morse v. Frederick, 

 (2007): first, that Justice Alito‘s concurrence in Morse is
the ―controlling‖ opinion in that case, Maj. Typescript at 21
n.10, 43, 45, 47; and second, that Morse ―modified‖ the
Supreme Court‘s decision in Fraser, Maj. Typescript at 6,
46–51. Both premises are wrong.

                               A

       I begin with the Majority‘s first premise, namely, that
Justice Alito‘s concurrence in Morse is the ―controlling‖
opinion in that case, despite the fact that Chief Justice




                               1
Roberts‘s majority opinion was joined in full by four other
Justices. Maj. Typescript at 36–46. This distinctly minority
view is contrary both to the understanding of Morse
expressed by eight of our sister Courts of Appeals and to what
we ourselves have repeatedly articulated to be the Court‘s
holding in Morse. By endorsing the Fifth Circuit‘s mistaken
understanding of Morse, the Majority applies an incorrect
legal standard that leads to the unfortunate result the Court
reaches today.

        The notion that Justice Alito‘s concurrence in Morse is
the controlling opinion flows from a misunderstanding of the
Supreme Court‘s ―narrowest grounds‖ doctrine as established
in Marks v. United States, 

 (1977). In Marks,
the petitioners had been convicted of distributing obscene
materials pursuant to jury instructions that were modeled on
the definition of obscenity articulated in Miller v. California,

 (1973). Marks, 430 U.S. at 190. Because the
petitioners‘ conduct occurred before the Court had decided
Miller, they argued that due process entitled them ―to jury
instructions not under Miller, but under the more favorable
[obscenity] formulation of Memoirs v. Massachusetts.‖ Id.
That formulation was unclear, however, because the Memoirs
Court had issued a fractured decision; no more than three of
the six Justices who voted for the judgment endorsed any one
of three separate opinions, each of which articulated a
different standard for obscenity.           See Memoirs v.
Massachusetts, 

, 414, 418 (1966) (plurality
opinion) (Justice Brennan, joined by Chief Justice Warren
and Justice Fortas, stating that obscenity may be proscribed if
it is ―utterly without redeeming social value‖); id. at 421, 424
(Black and Douglas, JJ., concurring in judgment) (concurring
separately on the grounds that obscenity cannot be




                               2
proscribed); id. at 421 (Stewart, J., concurring in judgment)
(concurring on the grounds that only hard-core pornography
is proscribable as obscene). The lack of a majority opinion in
Memoirs led the Sixth Circuit in Marks to reject the
petitioners‘ argument that the plurality‘s ―utterly without
redeeming social value‖ standard was the governing rule. It
reasoned that because ―the Memoirs standards never
commanded the assent of more than three Justices at any one
time . . . Memoirs never became the law.‖ Marks, 430 U.S. at
192 (describing the lower court‘s holding).

       On appeal, the Supreme Court rejected the Sixth
Circuit‘s reasoning and articulated the following standard:
―When a fragmented Court decides a case and no single
rationale explaining the result enjoys the assent of five
Justices, ‗the holding of the Court may be viewed as that
position taken by those members who concurred in the
judgments on the narrowest grounds . . . .‘‖ Id. at 193
(quoting Gregg v. Georgia, 

, 169 n.15 (1976)
(plurality opinion)). Based on this reasoning, the Court
concluded that because three Justices joined the plurality
opinion and Justices Black and Douglas ―concurred on
broader grounds,‖ ―[t]he view of the Memoirs plurality . . .
constituted the holding of the Court and provided the
governing standards.‖ Marks, 430 U.S. at 193–94.

       As Marks demonstrates, the narrowest grounds rule is
a necessary tool for deciphering the holding of the Court
when there is no majority opinion. See, e.g., Grutter v.
Bollinger, 

, 325 (2003) (attempting to apply the
Marks rule to derive a holding in the ―fractured decision‖
Regents of the University of California v. Bakke, 

 (1978)). Contrary to the Majority‘s holding today,
neither Marks nor other Supreme Court decisions support the




                              3
―unprecedented argument that a statement of legal opinion
joined by five Justices of th[e] Court does not carry the force
of law,‖ Vasquez v. Hillery, 

, 261 n.4 (1986).
Rather, the narrowest grounds rule applies only to ―discern a
single holding of the Court in cases in which no opinion on
the issue in question has garnered the support of a majority.‖
Id.; cf. Black‘s Law Dictionary 1201 (9th ed. 2009) (defining
a ―majority opinion‖ as ―[a]n opinion joined in by more than
half the judges considering a given case‖).

       Unable to find persuasive Supreme Court authority to
buttress its novel reading of Marks, the Majority argues that
our Court has ―applied the narrowest-grounds approach in
circumstances beyond those posed by Marks, including to
determine holdings in majority opinions.‖ Maj. Typescript at
37–38 (footnotes, citation, and internal quotation marks
omitted). For support, the Majority cites our decisions in
Horn v. Thoratec Corp., 

 (3d Cir. 2004), and
United States v. Bishop, 

 (3d Cir. 1995). Maj.
Typescript at 39–42. Neither case counsels the Majority‘s
application of the narrowest-grounds doctrine to interpret
Morse.

       In Horn, we looked to Justice Breyer‘s concurrence in
Medtronic v. Lohr, 

 (1996), for guidance on how
to address an issue central to our case, but that the Lohr Court
discussed only in dicta. See Horn, 376 F.3d at 175–76
(comparing Justice Breyer‘s ―more narrow‖ view on
preemption with ―Justice Stevens‘ sweeping pronouncement
[in his plurality opinion] that [the statute at issue] almost
never preempts a state common law claim‖). Likewise, in
Bishop, we cited Justice Kennedy‘s concurrence in United
States v. Lopez, 

 (1995), in order to reinforce the
already established principle that courts must exercise ―‗great




                               4
restraint‘ before a court finds Congress to have overstepped
its commerce power‖ despite Lopez‘s revolutionary holding.
Bishop, 66 F.3d at 590 (quoting Lopez, 514 U.S. at 568
(Kennedy, J., concurring)). Critically, in neither of these
cases did we indicate a belief that a concurring Justice can
create a new rule of law simply by both asking and answering
a question left unaddressed by the majority opinion. In fact,
we noted that Justice Breyer‘s concurrence in Horn was
particularly persuasive because ―Justice Breyer did not
discuss issues in his concurring opinion that Justice Stevens,
writing on behalf of the four-judge plurality, did not reach.‖
Horn, 376 F.3d at 175. That is not the case here.

        The Majority concedes that a concurring ―justice‘s
opinion ‗cannot add to what the majority opinion holds‘ by
‗binding the other four [j]ustices to what they have not said.‘‖
Maj. Typescript at 39 (quoting McKoy v. North Carolina, 

, 462 n.3 (1990) (Scalia, J., dissenting)). Yet by
holding that Justice Alito‘s concurrence ―controls the
majority opinion in Morse,‖ Maj. Typescript at 36, the
Majority violates this very principle. The majority in Morse
noted that ―this is plainly not a case about political debate,‖
Morse, 551 U.S. at 403, and refused to address what the result
of the case would have been had Frederick‘s banner been
―political.‖   The Majority implies that Justice Alito‘s
concurrence provides a definitive, ―controlling‖ answer to fill
the void left by the Morse majority opinion, but the Supreme
Court has disavowed this approach: ―The Court would be in
an odd predicament if a concurring minority of the Justices
could force the majority to address a point they found it
unnecessary (and did not wish) to address, under compulsion
of [the dissent‘s] new principle that silence implies
agreement.‖ Alexander v. Sandoval, 

, 285 n.5




                               5
(2001). Put another way, a majority ―holding is not made
coextensive with the concurrence because [the majority]
opinion does not expressly preclude (is ‗consistent with[]‘ . .
.) the concurrence‘s approach.‖ Id.

       Notwithstanding the Majority‘s statement to the
contrary, we have never applied the Marks rule to hold that a
concurrence may co-opt an opinion joined by at least five
Justices. Rather, consistent with Marks, ―we have looked to
the votes of dissenting Justices if they, combined with votes
from plurality or concurring opinions, establish a majority
view on the relevant issue.‖ United States v. Donovan, 

, 182 (3d Cir. 2011) (emphasis added); see also
Student Pub. Interest Research Grp. of N.J., Inc. v. AT&T
Bell Labs., 

, 1451 & n.16 (3d Cir. 1988). In
Donovan, we used Marks to analyze the Supreme Court‘s
―fractured‖ decision in Rapanos v. United States, 

 (2006), a case in which only three other Justices joined
Justice Scalia‘s plurality opinion and four others dissented.
Donovan, 661 F.3d at 179, 182. Nowhere did we suggest that
Marks would have been applicable had Rapanos featured a
single majority opinion. Likewise, in Planned Parenthood of
Southeastern Pennsylvania v. Casey, 

 (3d Cir.
1991), rev’d on other grounds, 

 (1992), we held
that Marks stands for the proposition that ―the controlling
opinion in a splintered decision is that of the Justice or
Justices who concur on the ‗narrowest grounds.‘‖ Casey, 947
F.2d at 693 (emphasis added). We then applied this principle
while interpreting the Supreme Court‘s plurality decisions in
Webster v. Reproductive Health Services, 

(1989), and Hodgson v. Minnesota, 

 (1990). See
Casey, 947 F.3d at 695–96 (noting that in Webster ―[t]he five
Justices in the majority issued three opinions,‖ none of which




                              6
garnered five votes on the legal issue in dispute, and that
―Hodgson was decided in a similar manner‖). Once again,
we gave no indication that Marks would have applied had
five Justices or more joined the same opinion.

        I also find it significant that, in the six years since
Morse was decided, nine of ten appellate courts have cited as
its holding the following standard articulated by Chief Justice
Roberts in his opinion for the Court: ―[A] principal may,
consistent with the First Amendment, restrict student speech
at a school event, when that speech is reasonably viewed as
promoting illegal drug use,‖ Morse, 551 U.S. at 403.1 Not

       1
         See Doninger v. Niehoff, 

, 345 (2d Cir.
2011) (―[T]he Supreme Court has determined that public
schools may ‗take steps to safeguard those entrusted to their
care from speech that can reasonably be regarded as
encouraging illegal drug use‘ because of the special nature of
the school environment and the dangers posed by student
drug use.‖ (citations omitted)); Hardwick ex rel. Hardwick v.
Heyward, 

, 435 (4th Cir. 2013) (―[S]chool
officials can regulate student speech that can plausibly be
interpreted as promoting illegal drugs because of ‗the dangers
of illegal drug use.‘‖ (citation omitted)); Defoe ex rel. Defoe
v. Spiva, 

, 332–33 (6th Cir. 2010) (―As this
Court has already recognized, however, the Morse holding
was a narrow one, determining no more than that a public
school may prohibit student expression at school or at school-
sponsored events during school hours that can be ‗reasonably
viewed as promoting drug use.‘‖ (citation omitted));
Zamecnik v. Indian Prairie Sch. Dist. No. 204, 

,
877 (7th Cir. 2011) (noting that promoting ―the use of illegal
drugs, [is] a form of advocacy in the school setting that can be




                               7
one of these courts indicated that Justice Alito‘s concurrence
controls, or that his dicta regarding ―political or social
speech‖ altered or circumscribed the Court‘s holding in
Morse. We too have articulated the import of Morse
consistent with these eight appellate courts: ―[I]n Morse, the
Court held that ‗schools may take steps to safeguard those
entrusted to their care from speech that can reasonably be
regarded as encouraging illegal drug use.‘‖ K.A. ex rel. Ayers
v. Pocono Mountain Sch. Dist., 

, 107 (3d Cir.


prohibited without evidence of disruption‖ (citation omitted));
D.J.M. ex rel. D.M. v. Hannibal Pub. Sch. Dist. No. 60, 

, 761 (8th Cir. 2011) (―Chief Justice Roberts
reviewed the Court‘s approach in these prior decisions before
holding ‗that schools may take steps to safeguard those
entrusted to their care from speech that can reasonably be
regarded as encouraging illegal drug use.‘‖ (citation
omitted)); Redding v. Safford Unified Sch. Dist. No. 1, 

, 1094 (9th Cir. 2008), rev’d on other grounds, 

 (2009) (―[S]chools can ‗restrict student expression
that they reasonably regard as promoting illegal drug use.‘‖
(citation omitted)); Corder v. Lewis Palmer Sch. Dist. No. 38,

, 1228 (10th Cir. 2009) (―[A] public school
may prohibit student speech at school or at a school-
sponsored event during school hours that the school
‗reasonably view[s] as promoting illegal drug use.‘‖ (citation
omitted)); Boim v. Fulton Cnty. Sch. Dist., 

, 984
(11th Cir. 2007) (―[T]he special characteristics of the school
environment and the governmental interest in stopping
student drug abuse . . . allow schools to restrict student
expression that they reasonably regard as promoting illegal
drug use.‖ (citation omitted)).




                              8
2013) (citation omitted).2 This widespread consensus is
further proof that Chief Justice Roberts‘s majority opinion,
not Justice Alito‘s concurrence, is the controlling opinion in
Morse.

       Before today, only the Fifth Circuit had held
otherwise. See Morgan v. Plano Indep. Sch. Dist., 

, 746 n.25 (5th Cir. 2009) (―We have held Justice Alito‘s
concurrence to be the controlling opinion in Morse.‖ (citing
Ponce v. Socorro Indep. Sch. Dist., 

, 768 (5th
Cir. 2007)); see also Morgan, 589 F.3d at 745 n.15
(interpreting the holding in Morse to be ―that schools may
regulate speech that a reasonable observer would interpret as
advocating illegal drug use and that could not be interpreted
as commenting on any political or social issue‖ (internal
quotation marks omitted)).3 However, the Fifth Circuit did

       2
         The Majority cites our opinion in J.S. ex rel. Snyder
v. Blue Mountain Sch. Dist., 

 (3d Cir. 2011), as
evidence that we ―previously‖ had the ―intuition‖ that Justice
Alito‘s concurrence controls the Supreme Court‘s opinion in
Morse. Maj. Typescript at 45 n.17. But in J.S., as in K.A., we
explicitly noted that the Supreme Court ―held that ‗the special
characteristics of the school environment and the
governmental interest in stopping drug abuse allow schools to
restrict student expression that they reasonably regard as
promoting illegal drug use.‘‖ 650 F.3d at 927 (emphasis
added) (quoting Morse, 551 U.S. at 408) (alterations, citation,
and internal quotation marks omitted).
       3
        The Majority claims that both the Sixth Circuit and
Tenth Circuit agree with the Fifth Circuit that Justice Alito‘s
concurrence is controlling. See Maj. Typescript at 45 n.17




                              9
not cite Marks or any other ―narrowest grounds‖ case and
provided no justification to support its conclusion that Justice
Alito‘s concurrence is the controlling opinion in Morse. As
the Seventh Circuit has aptly noted:

       The plaintiff calls Justice Alito‘s concurrence
       the ―controlling‖ opinion in Morse because
       Justices Alito and Kennedy were part of a five-
       Justice majority, so that their votes were crucial
       to the decision. But they joined the majority
       opinion, not just the decision, and by doing so
       they made it a majority opinion and not merely,
       as the plaintiff believes (as does the Fifth
       Circuit, Ponce v. Socorro Independent School


(citing Barr v. Lafon, 

, 564 (6th Cir. 2008), and
Corder, 566 F.3d at 1228). I disagree. In Barr, the Sixth
Circuit recognized Chief Justice Roberts‘s articulation that ―a
public school may prohibit student speech at school or at a
school-sponsored event during school hours that the school
‗reasonably view[s] as promoting illegal drug use‘‖ as the
Court‘s ―narrow holding.‖ 538 F.3d at 564 (citation omitted).
Although the opinion went on to discuss Justice Alito‘s
concurrence, the Sixth Circuit never opined that the
concurrence controls or otherwise modifies what the court
had previously described as Morse‘s ―narrow holding.‖ See
id.; see also Defoe, 625 F.3d at 332–33 & n.5 (describing the
same ―narrow‖ holding in Morse before discussing Justice
Alito‘s concurrence in a footnote). The same can be said for
the Tenth Circuit‘s decision in Corder, which essentially
parrots Barr‘s description of Morse‘s majority opinion and
Justice Alito‘s concurrence. See Corder, 566 F.3d at 1228
(quoting Barr, 538 F.3d at 564).




                              10
       District, 

, 768 (5th Cir. 2007)), a
       plurality opinion. The concurring Justices
       wanted to emphasize that in allowing a school
       to forbid student speech that encourages the use
       of illegal drugs the Court was not giving
       schools carte blanche to regulate student
       speech. And they were expressing their own
       view of the permissible scope of such
       regulation.

Nuxoll ex rel. Nuxoll v. Indian Prarie Sch. Dist. # 204, 

, 673 (7th Cir. 2008) (emphasis added) (citation
omitted). This interpretation of the relationship between
Justice Alito‘s concurrence and the majority opinion in Morse
is the correct one because it is faithful to Marks and its
progeny.

        For the reasons stated, I would not read Justice Alito‘s
concurrence as altering or circumscribing a majority opinion
for the Court that he joined in toto. Thus, the Court‘s holding
in Morse remains the familiar articulation that has been
consistently stated, time and again, by this Court and eight
other Courts of Appeals: ―[A] principal may, consistent with
the First Amendment, restrict student speech at a school
event, when that speech is reasonably viewed as promoting
illegal drug use.‖ Morse, 551 U.S. at 403.

                               B

       If Justice Alito‘s concurrence is not the ―controlling‖
opinion in Morse, the Majority has committed legal error by
engrafting his dicta regarding ―social or political‖
commentary as a limitation upon the ability of schools to
regulate speech that runs afoul of Fraser. But even assuming,




                              11
arguendo, that Justice Alito‘s concurrence alters or
circumscribes the Court‘s opinion in Morse, it is far from
clear that it had anything to say about the realm Fraser carved
out of Tinker v. Des Moines Independent Community School
District, 

 (1969).

        Tinker established the general rule that ―student
expression may not be suppressed unless school officials
reasonably conclude that it will ‗materially and substantially
disrupt the work and discipline of the school.‘‖ Morse, 551
U.S. at 403 (quoting Tinker, 393 U.S. at 513); see also, e.g.,
Saxe v. State Coll. Area Sch. Dist., 

, 211 (3d Cir.
2001). Tinker‘s ―substantial disruption‖ test does not apply in
every case, however. As then-Judge Alito wrote when he was
a member of this Court, ―the Supreme Court has carved out a
number of narrow categories of speech that a school may
restrict even without the threat of substantial disruption.‖ Id.
at 212; see also J.S., 650 F.3d at 927 (emphasizing that the
exceptions to Tinker are ―narrow‖). First came Fraser, in
which the Supreme Court held that schools may restrict the
manner in which a student conveys his message by forbidding
and punishing the use of lewd, vulgar, indecent, or plainly
offensive speech. See Fraser, 478 U.S. at 680–86. Then, in
Hazelwood School District v. Kuhlmeier, 

(1988), the Court held that administrators may regulate
speech that is school-sponsored or could reasonably be
viewed as the school‘s own speech. Id. at 272–73. Most
recently, in Morse the Court held that ―schools may take steps
to safeguard those entrusted to their care from speech that can
reasonably be regarded as encouraging illegal drug use.‖
Morse, 551 U.S. at 397.

      As these cases indicate, ―[s]ince Tinker, every
Supreme Court decision looking at student speech has




                              12
expanded the kinds of speech schools can regulate.‖ Palmer
ex rel. Palmer v. Waxahachie Indep. Sch. Dist., 

,
507 (5th Cir. 2009); cf. Morse, 551 U.S. at 417 (Thomas, J.,
concurring) (observing that ―the Court has since scaled back
Tinker‘s standard, or rather set the standard aside on an ad
hoc basis‖). In derogation of this consistent trend, the
Majority makes us the first United States Court of Appeals to
suggest that Morse has circumscribed Fraser, thereby
limiting the ability of teachers and administrators to regulate
student speech.

        In addition to overriding the careful steps taken to
allow schools to regulate student speech since Tinker, the
Majority errs by placing Morse at the center of a case that has
nothing whatsoever to do with illegal drug use. That Morse is
not central to this case is borne out by the way the case was
litigated and adjudicated. The District Court concluded that
only the standards of Tinker and Fraser are implicated, and
neither party ever argued otherwise. See B.H. v. Easton Area
Sch. Dist., 

 392, 394 (E.D. Pa. 2011) (―The
two Supreme Court cases examining student speech that are
most relevant to this case are Fraser and Tinker.‖). The
School District primarily contends that the ―I ♥ boobies!‖
bracelets are proscribable because they express sexual
innuendo that can reasonably be classified in the middle
school context as lewd, vulgar, and indecent speech.
Plaintiffs rejoin that the word ―boobies‖ is neither inherently
sexual nor vulgar, especially when conspicuously tied to
breast cancer awareness. Until the case reached the en banc
Court, no party or judge had suggested that Morse provided
the governing standard for this dispute. And rightly so,
because this is a Fraser case, not a Morse case, and there are
critical differences between the two.




                              13
       Courts have recognized, time and again, that the three
exceptions to Tinker‘s general rule are independent ―carve-
outs.‖ See, e.g., Saxe, 240 F.3d at 212–14. The Supreme
Court has given no indication—either in Morse or any of its
subsequent decisions—that it has modified the standard, first
articulated in Fraser more than 25 years ago, that governs
how schools are to regulate speech they may reasonably deem
lewd, vulgar, indecent, or plainly offensive. Moreover,
although the appellate courts have had dozens of
opportunities to do so, no court has suggested that Morse
qualified Fraser in any way. Since Morse, we have had
occasion to consider Fraser and have consistently
―interpreted [it] to permit school officials to regulate ‗lewd,
vulgar, indecent, and plainly offensive speech in school.‘‖
J.S., 650 F.3d at 927 (quoting Saxe, 240 F.3d at 213)
(emphasis and internal quotation marks omitted); see also
K.A., 710 F.3d at 107 (―In [Fraser], the Court held that
schools may restrict the manner in which a student conveys
his message by forbidding and punishing the use of lewd,
vulgar, indecent, and plainly offensive speech.‖ (citation
omitted)); Layshock ex rel. Layshock v. Hermitage Sch. Dist.,

, 212–13 (3d Cir. 2011) (same).

       In fact, the appellate opinions addressing Morse,
Fraser, and Kuhlmeier treat them as independent analytical
constructs that permit schools to regulate certain types of
speech that would otherwise be protected under Tinker. See,
e.g., Hardwick, 711 F.3d at 435 n.11 (―[W]e must continue to
adhere to the Tinker test in cases that do not fall within any
exceptions that the Supreme Court has created until the Court
directs otherwise.‖); Doninger, 642 F.3d at 353–54
(―[B]ecause the t-shirts were not vulgar, could not reasonably
be perceived to bear the School‘s imprimatur, and did not




                              14
encourage drug use, they could be subject to regulation
different from that permissible for adults in non-school
settings only if they threatened substantial disruption to the
work and discipline of the School.‖ (citations omitted)). It is
especially notable that even the Fifth Circuit, which
mistakenly held that Justice Alito‘s concurrence in Morse is
―controlling,‖ continues to treat the Tinker carve-outs as
independent exceptions rather than overlapping categories of
proscribable speech. See Morgan, 589 F.3d at 745 n.15 (5th
Cir. 2009) (characterizing Fraser as ―holding schools may
prohibit lewd, vulgar, obscene or plainly offensive student
speech‖ and, in the same string citation, separately
characterizing Morse as ―holding that schools may regulate
speech ‗that a reasonable observer would interpret as
advocating illegal drug use‘ and that could not be ‗interpreted
as commenting on any political or social issue‘‖ (citations
omitted)). The Majority‘s own analysis demonstrates that
threshold questions in a school speech case are whether the
speech at issue is governed by one of the three Tinker carve-
outs and, if not, whether the school acted properly under
Tinker. See Maj. Typescript at 63–64.

       In addition, we have emphasized that the carve-outs
touch on ―several narrow categories of speech that a school
may restrict even without the threat of substantial disruption.‖
K.A., 710 F.3d at 107 (emphasis added) (internal quotation
marks omitted). This does not mean, as the Majority
suggests, that the carve-outs narrow one another. See Maj.
Typescript at 45 n.17 (citing J.S., 650 F.3d at 927). Rather, it
is simply a recognition that they are narrow within their
separate spheres. Indeed, courts have been especially careful
to underscore the narrowness of the Court‘s holding in Morse.
See, e.g., Defoe, 625 F.3d at 332–33 (―[T]he Morse holding




                              15
was a narrow one, determining no more than that a public
school may prohibit student expression at school or at school-
sponsored events during school hours that can be ‗reasonably
viewed as promoting drug use.‘‖ (emphasis added) (citation
omitted)); Barr, 538 F.3d at 564 (same); B.W.A. v.
Farmington R-7 Sch. Dist., 

, 741 (8th Cir. 2009)
(same).

        In J.S., we too recognized the ―narrowness of the
Court‘s holding‖ in Morse. J.S., 650 F.3d at 927.4 There, we
declared that Morse did not apply to a school‘s punishment of
a student for creating a MySpace profile using graphic
language and imagery to disparage her teacher, see J.S., 650
F.3d at 932 n.10 (―Indisputably, neither Kuhlmeier nor Morse
governs this case.‖). Instead, we indicated that ―the only way
for the punishment to pass constitutional muster is if . . . J.S.‘s
speech can be prohibited under the Fraser exception to
Tinker.‖ Id. at 931–32. If the proper standard under Fraser is
the Majority‘s formulation of whether a student‘s lewd
speech may ―plausibly be interpreted as commenting on a
social or political issue,‖ surely we would have considered

       4
         The Majority believes that this clause serves as an
indicator that Justice Alito‘s concurrence narrowed the
holding in Morse and, in turn, narrowed the speech that
schools can proscribe under Fraser. See Maj. Typescript at
45 n.17. Contrary to the Majority‘s implication, in J.S. we
neither addressed Justice Alito‘s discussion of student speech
that touches on matters plausibly related to a social or
political issue nor indicated a belief that his concurrence
somehow modified the Morse Court‘s majority opinion,
which we quoted verbatim as the Court‘s holding. See J.S.,
650 F.3d at 927.




                                16
whether J.S.‘s online profile touched on any such issue.
Instead of doing so, we applied the Fraser test while
disavowing the relevance of Morse.

       The fact that courts have maintained analytical
separation among the different Tinker carve-outs makes sense
because the Supreme Court created each one for a unique
purpose. In K.A. we addressed these ―vital interests that
enable school officials to exercise control over student speech
even in the absence of a substantial disruption.‖ K.A., 710
F.3d at 107. The vital interest at issue in Morse that ―allow[s]
schools to restrict student expression that they reasonably
regard as promoting illegal drug use‖ is ―the special
characteristics of the school environment, and the
governmental interest in stopping student drug abuse.‖ Id.
(quoting Morse, 551 U.S. at 408). Fraser allowed schools to
punish ―lewd, indecent, or offensive speech,‖ 478 U.S. at 683,
to further ―society‘s . . . interest in teaching students the
boundaries of socially appropriate behavior,‖ K.A., 710 F.3d
at 107 (quoting Fraser, 478 U.S. at 681). And in Kuhlmeier,
the interest that ―entitle[s] [educators] to exercise greater
control over [school-sponsored publications]‖ is ―to assure
that participants learn whatever lessons the activity is
designed to teach, that readers or listeners are not exposed to
material that may be inappropriate for their level of maturity,
and that the views of the individual speaker are not
erroneously attributed to the school.‖ K.A., 710 F.3d at 107
(quoting Kuhlmeier, 484 U.S. at 271).              The Court‘s
willingness to curtail the First Amendment rights of students
to enable schools to achieve these important goals vindicates
the principle that ―the rights of students ‗must be applied in
light of the special characteristics of the school
environment.‘‖ Morse, 551 U.S. at 397 (quoting Kuhlmeier,




                              17
484 U.S. at 266). Because each case was intended to address
a separate concern, I disagree with the Majority that language
qualifying one type of carve-out applies equally to the others.

        In sum, Morse‘s ―narrow‖ holding does not apply
unless a school has regulated student speech that it viewed as
advocating illegal drug use. Notwithstanding its critical
reliance on Morse, at one point the Majority seems to agree
that Morse does not apply to this case when it states that ―no
one could reasonably interpret the bracelets as advocating
illegal drug use.‖ Maj. Typescript at 64. The Majority can‘t
have it both ways. The decision to engraft Justice Alito‘s
Morse concurrence onto Fraser erodes the analytical
distinction between the two lines of cases and turns this
appeal into some sort of Fraser/Morse hybrid. ―The law
governing restrictions on student speech can be difficult and
confusing, even for lawyers, law professors, and judges. The
relevant Supreme Court cases can be hard to reconcile, and
courts often struggle to determine which standard applies in
any particular case.‖ Doninger, 642 F.3d at 353. By using
Morse to modify the distinct carve-out established in Fraser,
the Majority has muddied the waters and further encumbered
the ability of educators to run their schools.

        The Majority attempts to make more palatable its
decision to engraft Morse‘s supposed prohibition of ―any
restriction of speech that can plausibly be interpreted as
commenting on any political or social issue‖ onto Fraser.
For instance, it claims that ―the [Supreme] Court did not
believe that Fraser‘s speech could plausibly be interpreted as
political or social commentary.‖ Maj. Typescript at 27. By
claiming that such an interpretation of Matthew Fraser‘s
―speech nominating a fellow student for student elective
office,‖ Fraser, 478 U.S. at 677, is wholly ―implausible,‖ the




                              18
Majority demonstrates the difficulties that arise when it
blends together the disparate Tinker carve-outs.

       As the Majority rightly notes, the Fraser Court opined
that there was a ―marked distinction between the political
‗message‘ of the armbands in Tinker and the sexual content of
Fraser‘s speech.‖ Maj. Typescript at 28–29 (quoting Fraser,
478 U.S. at 680). That does not mean, however, that it was
implausible to conclude that Fraser‘s speech was political. If
it were truly implausible to ―interpret[] [Fraser‘s speech] as
commenting on any political or social issue,‖ one must
wonder why the United States Court of Appeals for the Ninth
Circuit characterized Fraser‘s speech as ―student political
speech-making‖ and a ―campaign speech[].‖ Fraser v. Bethel
Sch. Dist. No. 403, 

, 1363 (9th Cir. 1985),
rev’d, 

 (1986); id. at 1368 (Wright, J.,
dissenting). The three appellate judges who heard Fraser‘s
case were deemed by the Supreme Court to have erred when
they likened his speech to Tinker‘s armband, but that does not
mean that it was ―implausible‖ for those three judges to view
Fraser‘s speech as political. It was, after all, a campaign
speech.

        A brief hypothetical further demonstrates the problems
posed by the Majority‘s plausibility-based articulation of the
Fraser carve-out. Suppose a student makes a speech at a
school assembly. Like Matthew Fraser‘s speech, the content
is about supporting a candidate for office, but the sexual
references are muted enough such that the Majority would
deem them ―ambiguously lewd‖ instead of ―plainly lewd.‖ If
the student‘s speech is about a classmate running for school
office, the Majority would say that the school may punish the
speaker. But if an identical speech is given and the
classmate‘s name is replaced with the name of a candidate for




                             19
president, mayor, or even school board, the Majority would
conclude that the First Amendment insulates the student‘s
speech. In my view, the two speeches are indistinguishable
under Fraser.

       In sum, the Majority‘s approach vindicates any speech
cloaked in a political or social message even if a reasonable
observer could deem it lewd, vulgar, indecent, or plainly
offensive. In both cases, the inappropriate language is
identical, but the speech is constitutionally protected as long
as it meets the Majority‘s cramped definition of ―politics‖ or
its as-yet-undefined notion of what constitutes ―social
commentary.‖ Fraser repudiated this very idea. ―The First
Amendment guarantees wide freedom in matters of adult
public discourse . . . . It does not follow, however, that
simply because the use of an offensive form of expression
may not be prohibited to adults making what the speaker
considers a political point, the same latitude must be
permitted to children in a public school.‖ Fraser, 478 U.S. at
682 (emphasis added).

                               II

       As noted, the Majority holds that ―Fraser . . . permits a
school to categorically restrict ambiguous speech that a
reasonable observer could interpret as having a lewd, vulgar,
or profane meaning,‖ but only ―so long as it could not also
plausibly be interpreted as commenting on a social or political
issue.‖ Maj. Typescript at 61. It is important to emphasize
here that, despite my disagreement with the second part of the
Majority‘s formulation, I agree fully with its understanding of
the objective-reasonableness inquiry compelled under Fraser.
See Maj. Typescript 32–35 (discussing why ―courts should
defer to a school‘s decisions to restrict what a reasonable




                              20
observer would interpret as lewd, vulgar, profane, or
offensive‖).5


      5
         Though I believe an objective-reasonableness test is
the correct interpretation of Fraser, its level of generality
leaves something to be desired, particularly when one
considers that the lower courts will look to our decision for
guidance. The Majority states that ―[i]t remains the job of
judges . . . to determine whether a reasonable observer could
interpret student speech as lewd, profane, vulgar, or
offensive.‖ Maj. Typescript at 33–34. But who is this
―reasonable observer‖? The Majority gives us clues: he
―would not adopt an acontextual interpretation‖ and would
consider ―the plausibility of the school‘s interpretation in
light of competing meanings; the context, content, and form
of the speech; and the age and maturity of the students.‖ Maj.
Typescript at 34. I would add several more considerations.
Most importantly, evolving societal norms counsel that what
is ―objectively‖ considered ―lewd, profane, vulgar, or
offensive‖ one day may not be so the next. See, e.g., Fraser,
478 U.S. at 691 (Stevens, J., dissenting) (―‗Frankly, my dear,
I don‘t give a damn.‘ When I was a high school student, the
use of those words in a public forum shocked the Nation.
Today Clark Gable‘s four-letter expletive is less offensive
than it was then.‖). Furthermore, given the diversity of
opinions and perspectives across our country, the type of
speech that may reasonably fall into one of the proscribable
categories would vary widely from one community to the
next. These considerations highlight the importance of
ensuring that ―the determination of what manner of speech in
the classroom or in school assembly is inappropriate properly
rests with the school board.‖ Fraser, 478 U.S. at 683.




                             21
        The Majority did not find that the school‘s
interpretation of the bracelets‘ message as lewd was
objectively unreasonable. See Maj. Typescript at 63 n.22
(―[W]e need not determine whether a reasonable observer
could interpret the bracelets‘ slogan as lewd.‖). Thus, had the
Majority not engrafted Justice Alito‘s concurrence in Morse
onto the Fraser standard, my colleagues might agree that the
school did not violate the First Amendment when it
proscribed the bracelet. Because the Majority chose not to
analyze whether the school was reasonable in determining
that the bracelet could be proscribed under Fraser, however, I
will briefly discuss why that is so.

       In this close case, the ―I ♥ boobies! (KEEP A
BREAST)‖ bracelets would seem to fall into a gray area
between speech that is plainly lewd and merely indecorous.
Because I think it objectively reasonable to interpret the
bracelets, in the middle school context, as inappropriate
sexual innuendo and double entendre, I would reverse the
judgment of the District Court and vacate the preliminary
injunction.

       The District Court correctly ascertained the standard of
review to apply in a case that arises under Fraser, but
proceeded to misapply that standard. First, by emphasizing
whether Plaintiffs intended a vulgar or sexual meaning in
their ―I ♥ boobies!‖ bracelets and determining that a non-
sexual, breast-cancer-awareness interpretation of the bracelets
was reasonable, the Court inverted the proper question.
Instead of asking whether it was reasonable to view the
bracelets as an innocuous expression of breast cancer
awareness, the District Court should have asked whether the
school officials‘ interpretation of the bracelets—i.e., as
expressing sexual attraction to breasts—was reasonable. So




                              22
long as the School District‘s interpretation was objectively
reasonable, the ban did not contravene the First Amendment
or our school-speech jurisprudence.

       Second, in its substantive conclusion that ―I ♥
boobies!‖ cannot reasonably be regarded as lewd or vulgar,
the District Court highlighted the bracelets‘ social value while
disregarding their likely meaning to immature middle-
schoolers.6 As the School District argues, the fact that



       6
       In fact, we have questioned the applicability of the
Supreme Court‘s student speech jurisprudence in the
elementary and middle school settings:

       [A]t a certain point, a school child is so young
       that it might reasonably be presumed the First
       Amendment does not protect the kind of speech
       at issue here. Where that point falls is subject
       to reasonable debate.

       In any event, if third graders enjoy rights under
       Tinker, those rights will necessarily be very
       limited.    Elementary school officials will
       undoubtedly be able to regulate much—perhaps
       most—of the speech that is protected in higher
       grades.    When officials have a legitimate
       educational reason—whether grounded on the
       need to preserve order, to facilitate learning or
       social development, or to protect the interests of
       other students—they may ordinarily regulate
       public elementary school children‘s speech.




                              23
Plaintiffs‘ laudable awareness message could be discerned
from the bracelets does not render the School District‘s ban
unconstitutional. ―I ♥ boobies!‖ not only expresses support
for those afflicted with breast cancer, but also conveys a
sexual attraction to the female breast.

       It is true that certain facts indicate that a sexual
interpretation of the ―I ♥ boobies!‖ bracelets may be at the
outer edge of how a reasonable observer would interpret
speech. Most obviously, the bracelets always modify the ―I ♥
boobies!‖ phrase with ―(KEEP A BREAST)‖ or other breast-
cancer-awareness messages. ―When one reads the entire


Walker-Serrano ex rel. Walker v. Leonard, 

,
417–18 (3d Cir. 2003); see also Walz ex rel. Walz v. Egg
Harbor Twp. Bd. of Educ., 

, 276 (3d Cir. 2003)
(noting that ―the age of the students bears an important
inverse relationship to the degree and kind of control a school
may exercise: as a general matter, the younger the students,
the more control a school may exercise‖). Other appellate
courts share our misgivings, noting that ―the younger the
children, the more latitude the school authorities have in
limiting expression.‖ Zamecnik, 636 F.3d at 876 (citing
Muller ex rel. Muller v. Jefferson Lighthouse Sch., 

, 1538–39 (7th Cir. 1996)); see also Nuxoll, 523 F.3d at
673 (when a school regulates the speech of children that are
―very young . . . the school has a pretty free hand‖); Morgan,
659 F.3d at 386 (―[I]n public schools, the speech appropriate
for eighteen-year-old high school students is not necessarily
acceptable for seven-year-old grammar school students.
Indeed, common sense dictates that a 7-year-old is not a 13-
year-old, and neither is an adult.‖ (alterations, citations, and
internal quotation marks omitted)).




                              24
phrase, it is clearly a message designed to promote breast
cancer awareness.‖ K.J. v. Sauk Prairie Sch. Dist., No. 11-
cv-622, slip op. at 14 (W.D. Wis. Feb. 6, 2012).
Additionally, school administrators did not immediately
recognize the bracelets as vulgar or lewd; students had been
wearing the bracelets for two months before they were
banned, and teachers had to request guidance on whether and
how to deal with the bracelets. Moreover, the school itself
was compelled to use the word ―boobies‖ over the public
address system and school television station in order to
describe the proscribed bracelets, which suggests that the
word alone is not patently offensive.

       Notwithstanding the facts supporting Plaintiffs‘ case, I
conclude that ―I ♥ boobies!‖ can reasonably be interpreted as
inappropriate sexual double entendre. In the middle school
context, the phrase can mean both ―I support breast-cancer-
awareness measures‖ and ―I am attracted to female breasts.‖
Many twelve- and thirteen-year-old children are susceptible
to juvenile sexualization of messages that would be
innocuous to a reasonable adult. Indeed, at least one bracelet-
wearer acknowledged that ―immature‖ boys might read a
lewd meaning into the bracelets and conceded that she
understood why the school might want to ban the bracelets,
B.H., 

 at 399, and other students parroted the
phrase on the bracelets while conveying sexual attraction to
breasts. Another school administrator has concluded that the
bracelets at issue here ―elicit attention by sexualizing the
cause of breast cancer awareness.‖ Sauk Prairie, No. 11-cv-
622, at 4. And as Judge Crabb, the only other federal judge to
consider these bracelets, put it in Sauk Prairie, ―hints of
vulgarity and sexuality‖ in the bracelets ―attract attention and
provoke conversation, a ploy that is effective for [KABF‘s]




                              25
target audience of immature middle [school] students.‖ Id. at
15. Finally, as the Gender Equality amicus brief points out,
breasts are ubiquitously sexualized in American culture.

       The Easton Area Middle School principals‘
willingness to say ―boobies‖ to the entire school audience
does not imply that the word does not have a sexual meaning;
it merely suggests that ―boobies‖ is not plainly lewd.
Moreover, although KABF‘s decision not to market its
products through porn stars and at truck stops is laudable, the
interest such organizations have shown in the bracelets is
further evidence that the bracelets are read by many to
contain a sexual meaning. And the ―I ♥ boobies!‖ bracelets‘
breast cancer message is not so obvious or overwhelming as
to eliminate the double entendre. For one thing, the bracelets
come in many colors other than the shade of pink widely
associated with the fight against breast cancer.

        Additionally, although Plaintiffs and their amici argue
that the casual language of the ―I ♥ boobies!‖ bracelets is
intended to make breast cancer issues more accessible and
less stigmatized for girls and young women, that purpose
does not undermine the plausibility of a sexual interpretation
of the bracelets. Nor does the fact that these Plaintiffs‘
mothers were happy not only to purchase the bracelets for
their teenage daughters but also to wear them render the
bracelets immune from school regulation. The mothers‘
intent that the bracelets convey a breast-cancer-awareness
message, like Plaintiffs‘ own subjective motive, is irrelevant
to interpreting the meaning of the speech.

       Likewise, the School District administrators‘
subjective beliefs, expressed at the time of the ban and later
during this litigation, do not affect my determination of




                              26
whether it is objectively reasonable to infer a sexualized
meaning from the bracelets. Their failure to use the words
―lewd,‖ ―vulgar,‖ ―indecent,‖ or ―plainly offensive‖ is not
fatal to their claim of regulatory authority. Similarly, some
principals‘ inconsistent testimony regarding what other
breast-cancer-related phrases they might censor does not
make the phrase at issue here more or less vulgar. Therefore,
it is not probative that administrators intermittently indicated
that they thought the word ―breast‖ by itself has an
impermissible sexual connotation.

       Plaintiffs rely on the initial statements by teachers at
the middle school that the word ―breast‖ alone in any context
and the phrases ―breast cancer awareness‖ and ―keep-a-
breast.org‖ could also be banned to argue that the School
District has left them no other means to convey their breast-
cancer-awareness message. But those words were not
banned—indeed, students are permitted to wear KABF‘s
―check y♥urself!! (KEEP A BREAST)‖ bracelets—and the
administrators changed their position prior to the evidentiary
hearing, opining that such phrases would not be inappropriate
at school. Also significant is the fact that the Easton Area
Middle School has not stifled the message of breast cancer
awareness; in the course of a robust breast cancer awareness
campaign it merely imposed a permissible restriction on the
way in which that message may be expressed. See Saxe, 240
F.3d at 213 (―Fraser speaks to the form and manner of
student speech, not its substance. It addresses the mode of
expression, not its content or viewpoint.‖ (citation omitted)).

       Nor is Plaintiffs‘ position saved by the fact that the ―I
♥ boobies!‖ phrase was ―chosen to enhance the effectiveness
of the communication to the target audience.‖ B.H., 827 F.
Supp. 2d at 406. The District Court‘s focus on the strategic




                              27
purpose of the words and format used in the bracelets was
misguided. If indecency were permitted in schools merely
because it was intended to advance some laudable goal,
Matthew Fraser‘s speech would have been constitutionally
protected insofar as he intended to win the attention of his
classmates while advocating the election of his friend.

        Finally, if we were to hold that the breast cancer
message here makes any sexual reading of the bracelets
unreasonable, schools would be obliged to permit more
egregiously sexual advocacy messages. As Ms. DiVietro
acknowledged, ―other bodily parts in the human anatomy . . .
can get cancer and . . . other types of slang terms‖ would have
to be condoned. App. 275. DiVietro raised the specter of an
―I ♥ Balls‖ slogan to support testicular cancer awareness. Id.
at 275–76. These examples are not speculative. The
Testicular Cancer Awareness Project sells ―feelmyballs‖
bracelets to encourage male self-examinations and general
awareness.      See Testicular Cancer Awareness Project,
http://www.feelmyballs.org/shop/front.php (last visited June
3, 2013). If middle school students have a constitutional right
to wear ―I ♥ boobies!‖ bracelets, it would be difficult to
articulate a limiting principle that would disallow these other
catchy phrases, so long as they were aimed at some socially
beneficial objective.

       Simply stated, the District Court correctly articulated
the proper standard of review to be applied in cases that
implicate Fraser (such as this one), but it strayed from that
standard when evaluating the reasonableness of Plaintiffs‘
intended meaning. For that reason, and because the School
District‘s reading of ―I ♥ boobies!‖ as inappropriate sexual
double entendre was a reasonable interpretation in the middle
school context, I would hold that Plaintiffs cannot




                              28
demonstrate a likelihood of success on the merits of their
claim. Accordingly, the District Court abused its discretion in
granting a preliminary injunction.

                *             *             *

       As this case demonstrates, running a school is more
complicated now than ever before. Administrators and
teachers are not only obliged to teach core subjects, but also
find themselves mired in a variety of socio-political causes
during school time. And they do so in an era when they no
longer possess plenary control of their charges as they did
when they acted in loco parentis. See, e.g., Morse, 551 U.S.
at 413–16 (Thomas, J., concurring). The decisions school
administrators must make regarding the deportment of their
students—what they say, what they wear, or what they do—
require common sense and good judgment. Many of those
decisions will involve matters about which reasonable people
can disagree. In the close cases, such as this one, there is
virtue in deferring to the reasonable judgments of those
responsible for educating our nation‘s youth. With respect, I
dissent.




                              29
GREENAWAY, JR., Circuit Judge, dissenting, with whom
CHAGARES, JORDAN, HARDIMAN and GREENBERG,
join.

        My colleagues have determined today that “I ♥
boobies” is an ambiguous phrase that may connote an
attraction to female breasts, but which falls under the
protection of the First Amendment in the middle school
context because it may plausibly be interpreted as
commenting on a political or social issue. Reasonable minds
may come to varying conclusions on this test, but one thing is
not open to debate: a school district faced with the same
dilemma in the coming weeks, months, or years is given no
greater guidance regarding its ability to determine whether a
particular message may be proscribed than before the
Majority opinion issued.

        The Majority lauds the intent of the two middle
schoolers responsible for introducing “I ♥ boobies! (KEEP A
BREAST)” bracelets into their school, which encouraged
serious discussion regarding a medical issue of increasing
social import. Appellees‟ actions may or may not reflect an
admirable maturity, but the intent of Appellees is not at issue.
In many cases, when the First Amendment is implicated, the
intent of the speakers will be admirable or at worst benign.
The Majority concludes that, as long as the ambiguous speech
may be interpreted by a reasonable person as plausibly related
to a political or social issue, it is protected. Despite its
express disavowal of intent as a consideration, the Majority
inadvertently re-injects the students‟ intent into the fray by
mandating an analysis of whether a political or social issue is
addressed by the speech. This is improper but it is not my
sole criticism.




                               1
       The Majority‟s test leaves school districts essentially
powerless to exercise any discretion and extends the First
Amendment‟s protection to a breadth that knows no bounds.
As such, how will similarly-situated school districts apply this
amorphous test going forward? The Majority‟s test has two
obvious flaws. First, what words or phrases fall outside of the
ambiguous designation other than the “seven dirty words”?
Second, how does a school district ever assess the weight or
validity of political or social commentary? The absence of
guidance on both of these questions leaves school districts to
scratch their heads.

        Practical problems with the Majority‟s test abound.
Where and how do school districts line-draw regarding the
nouns used to describe the subject matter of the particular
awareness campaign? The Majority has established that at
opposite ends of the spectrum are “boobies,” on the one hand,
and “tits,” one of the “seven dirty words,” on the other hand.
What lies between those two extremes and how a school
district is to make a principled judgment going forward
remain open questions. No doubt, there are some words and
phrases that all would agree should be afforded no protection
in the middle school context, despite their use in promoting
an important social issue. My recalcitrance to extend First
Amendment protection to the slogan at hand is simple — why
is this word, “boobies,” different? Why does it deserve
protection? Is “boobies” a term that is inherently innocuous
or sophomoric, as the Majority asserts? As noted in the
Majority, “ta tas” is used as the descriptive term in some
breast cancer awareness campaigns. The ambiguity of “ta
tas” in this context is beyond question. What also seems
beyond question is that the school district, according to the
Majority, must lay dormant to a student‟s use of “ta tas” or




                               2
any synonym of “breast” (other than “tits”) as long as the
student is commenting on a political or social issue, here,
breast cancer awareness. The lack of certitude or a workable
parameter unnecessarily handcuffs school districts.

        What of the circumstance when an anatomically
correct term is used in an awareness campaign? Applying the
Majority‟s test, “I ♥ penises,” “I ♥ vaginas,” “I ♥ testicles,”
or “I ♥ breasts” would apparently be phrases or slogans that
school districts would be powerless to address. Would the
invocation of any of these slogans in a cancer awareness
effort fail to garner protection under the Majority‟s test? It
would appear not. What of the other slogans that the
Majority mentions in its opinion that are sufficiently
ambiguous? The Majority blithely states that “it does not
enjoin the School District‟s regulation of other types of
apparel, such as the „Save the ta-tas‟ T-shirt or testicular-
cancer-awareness         apparel    bearing      the       phrase
„feelmyballs.org.‟” (Maj. Op. 71.) This is exactly my
concern. What may a school district do? These phrases are
both ambiguous and speak to political and social issues. How
is a school district now better able to discern when it may
exercise its discretion to impede the use of a particular slogan,
as it relates to an awareness program, than before the issuance
of this opinion?

       The other practical problem which arises from
application of the Majority‟s test is judging the validity of
political and social comment. In the context of these social
awareness campaigns, when would the students‟ involvement
not invoke political or social comment? The constriction of
“plausibly be interpreted as” adds little to our discourse. For
instance, when would a student using a term that is admittedly
ambiguous not be able to assert that the use of the offending




                               3
word, term, or phrase is speech that is commenting on a
political or social issue? What is the balancing that a school
district can/should/may engage in to determine the merit or
value of the proposed political or social comment? The
unabashed invocation of a lewd, vulgar, indecent or plainly
offensive term is not what is at issue here; what is at issue is
the notion that we have established a test which effectively
has no parameters. The political or social issue prong entirely
eviscerates the school district‟s authority to effectively
evaluate whether the student‟s speech is indeed protected.
This shortcoming in the application of the test exemplifies its
inherent weakness — a failure to resolve the conundrum
school districts face every day.

       In light of the Majority‟s approach, school districts
seeking guidance from our First Amendment jurisprudence in
this context will find only confusion. I cannot adhere to this
approach. I respectfully dissent.




                               4