The Freedom of Non-Speech

Free Speech Beyond Words. Mark V. Tushnet,[1] Alan K. Chen,[2] and Joseph Blocher.[3] NYU Press. 2017. Pp. vii + 260. $28.00 (Cloth).

Enrique Armijo[4]

Introduction

How does one achieve eternal bliss? By saying dada. How does one become famous? By saying dada . . . . How can one get rid of everything that smacks of journalism, worms, everything nice and right, blinkered, moralistic, europeanised, enervated? By saying dada . . . .

I don’t want words that other people have invented. All the words are other people’s inventions. I want my own stuff, my own rhythm, and vowels and consonants too, matching the rhythm and all my own.[5]

On the evening of August 29, 1952, at Maverick Concert Hall, an open-air theater deep in the Catskill Mountains’ forest preserve on the outskirts of Woodstock, New York, the virtuoso pianist David Tudor strode out onstage before an unsure audience, sat at a piano, set a stopwatch, closed the keyboard lid, and for 30 seconds, did nothing. He then opened the lid, reclosed it, reset the watch, and sat silently again for an additional 2 minutes and 23 seconds. He opened and closed the lid, again sat silently at the piano keys for another timed minute and 40 seconds, stood up, and walked offstage.[6]

Though Tudor was careful to make no audible sounds during the performance, the sheet music at his piano was not blank. The composer’s notation read “Tacet.[7] For any instrument or instruments.” The score noted three movements and their lengths, demarcated in the performance by Tudor’s opening and closing of the piano’s keyboard lid, and, in a subsequent recreation of the score, a series of vertical lines intended to indicate the passage of time.[8]

Composer John Cage, the author of that score, would say that the piece, then tentatively called Four Movements, later named 4’33” after its total length, was inspired by artist Robert Rauschenberg’s White Paintings, which Cage had encountered at a Rauschenberg solo show the year before Tudor’s performance.[9] The White Paintings were five paneled works “painted on canvas in a smooth, unmodulated white.”[10] Cage interpreted the images not as a means to project the artist’s own expression, but rather as “backdrops against which the flux of the world might stand out.”[11] Cage would describe the White Paintings as “mirrors of the air,” and “airports for the lights, shadows, and particles” of the rooms in which the paintings were displayed.[12] Their lack of form focused the viewer’s attention on naturally occurring images, such as the changes in light and shadow in the surrounding space. The message Rauschenberg sought to communicate could only be expressed by stripping away the paintings’ content. When he began work on a new set of White Paintings, Rauschenberg instructed his assistant to “[p]aint them so they look like they haven’t been painted.”[13]

With 4’33”, Cage sought to supply the same aesthetic experience through musical composition. The work’s use of silence, when combined with the socially self-imposed silence of the concert hall,[14] was intended to frame the other sounds that occurred in that space during that time, both natural and man-made. By composing a soundless score, Cage sought to teach the audience to listen for those sounds that are all around them that modern society had trained them to forget were there. Writing about that first 1952 performance, Cage said of the audience that

[w]hat they thought was silence, because they didn’t know how to listen, was full of accidental sounds. You could hear the wind stirring outside during the first movement. During the second, raindrops began pattering on the roof, and during the third the people themselves made all kinds of interesting sounds as they talked or walked out.[15]

Despite—indeed because of—its lack of compositional sound, 4’33” thus serves, as critic and composer James Pritchett writes, as “a tribute to the experience of silence, a reminder of its existence and its importance for all of us.”[16] It is only through silence, Cage believed, that we can notice the “permanent presence of the sounds all around us,” and can come to learn that those sounds are “worthy of attention.”[17] In today’s wireless, cacophonous, and ever-connected world, this is a reminder that we need more than ever.

* * *

The story of 4’33”’s inspiration, composition, first performance, and evolving meaning is not merely a story about a work of silence. Rather, it is a story about a work whose author decided that the work be silent, and a story of the message that the work intended to communicate—a message that could only be expressed by not communicating, in particular the use of silence to demonstrate that, as Cage said, “there’s no such thing as silence.”[18] It is a story of expressive choice, in particular Cage’s decision to reject locution so as to convey his intended meaning.[19] Which means that for First Amendment scholars, it presents a test case.

In their illuminating and timely new book, Free Speech Beyond Words, Mark Tushnet, Alan Chen, and Joseph Blocher ask whether the First Amendment applies to expression that, like 4’33”, but also like Lewis Carroll’s nonsense poem Jabberwocky or Jackson Pollock’s splatter painting, does not use words or conventionally representational imagery. It is a question that, as the authors show, the United States Supreme Court has long assumed has an affirmative answer. Even though such works do not use words to express ideas or otherwise convey meaning, and even though the Court is asking whether such works are within “the freedom of speech,” the Court has unanimously declared that they are “unquestionably shielded” by the Speech Clause (p. 2).[20] The question that Tushnet, Chen, and Blocher seek to answer is why that is so.

Why ask why such works are “shielded” by the First Amendment, then, especially since the Court has long treated the answer as a given? According to the authors, the Court runs doctrinal risks when it glides past what Frederick Schauer calls the “coverage” First Amendment question in its rush to reach the “protection” question.[21] Tushnet et al. begin with the unassailable premise that communication involves much more than the use of words to convey meaning, and thus “speech” for First Amendment purposes must entail more than that as well (p. 1). But that premise cannot alone drive the Court’s conclusion that wordless communication is covered by the Amendment.

Setting the boundaries of the First Amendment is a project of determining coverage, not protection, and we need independent justifications for finding coverage exists in order for free speech law to cohere. Additionally, ignoring coverage is doubly dangerous, the authors argue, precisely because the tools the Court uses in deciding the protection question have become so settled in favor of protecting speech, while the tools for determining coverage questions are, in many cases and to borrow John Cage’s term, tacet.[22] First Amendment protection rules are so outcome-determinative that the answer to the protection question in cases like U.S. v. Stevens,[23] Sorrell v. IMS Health,[24] and Brown v. Entertainment Merchants’ Association[25] is doing boundary-setting work which should properly be addressed at the preliminary point of deciding whether First Amendment coverage exists or not (p. 3).

This approach, in the view of the authors, is exactly backwards. The question of what the First Amendment covers is a more consequential question than whether the First Amendment protects a particular speaker or group of speakers in a particular case.[26] Yet the Court continually bogs down in the application of First Amendment rules where the outcome with respect to those rules is in little doubt, and with almost no rigorous predicate analysis of whether the First Amendment applies in the first instance.

A look at the Court’s recent work bears this reading out. Pages are a rough proxy at best for the depth of legal analysis within those pages.[27] But to demonstrate the authors’ premise, below is a chart showing the three aforementioned First Amendment cases, the length of the slip opinions in those cases, and the approximate amount of discussion, as measured by pages in those opinions, that the Court’s members spend analyzing the coverage question with respect to the claim at issue as opposed to the protection question:

 

Case	Total pages	Coverage analysis	Protection analysis
Stevens	52	5 pp. (maj. op.) 9 pp. (diss.)	11 pp. (maj. op.) 9 pp. (diss.)
Sorrell	53	6 pp. (maj. op.) 0 p. (diss.)	16 pp. (maj. op.) 24 pp. (diss.)
Brown	92	½ p. (maj. op.)[28] 0 p. (concur. op.)[29] 20 pp. (Thomas diss.)[30] 0 p. (Breyer diss.)	15 pp. (maj. op.) 0 p. (concur. op.) 0 p. (Thomas diss.) 12 pp. (Breyer diss.)

 

Putting aside Justice Thomas’ dissent in Brown—his commitment to original public meaning-based interpretation causes Thomas to sometimes be an outlier on the current Court with respect to treatment of First Amendment coverage questions[31]—the contrast within this admittedly small sample size is stark. The Court spent more than triple the amount of time in its opinions’ pages discussing the extent of the First Amendment’s protection as it did discussing whether the First Amendment applied to the relevant communicative acts at all. This is a bizarre way to decide cases.[32]

First of all, it carves out the First Amendment from the hardiest canon of construction in all of public law: the doctrine of constitutional avoidance.[33] Of course, the more conventional version of constitutional avoidance instructs that courts should strive to interpret statutes so as to avoid raising constitutional questions. But an analogous form of constitutional avoidance occurs when, as A. E. Dick Howard notes, the Court “pass[es] up a more difficult constitutional question in favor of another which, albeit constitutional, was hardly controversial.”[34] As the authors write, “[t]he whole point of treating the First Amendment as having boundaries is to avoid in-depth analysis of cases involving uncovered conduct” (p. 4, emphasis in original). If the Court were to actually engage the question of First Amendment coverage, it could dispense with protection questions when it had before it an allegedly communicative act that was not sufficiently expressive to fall within the Amendment. But because protection doctrine is well developed (some might instead say larded up), with its many multipart tests and tiers of scrutiny, and coverage doctrine is largely an act of assumption, legal analysis is pulled toward questions which have existing tools available to answer them. It is temptingly easy to apply longstanding doctrines and balancing tests that weigh speaker and state interests, and similarly tempting to assume coverage exists because there are fewer tools to answer that question. But by placing its focus on protection over coverage, the Court is choosing the question that is superficially more difficult, but is in fact more straightforward, and certainly less controversial.

Without clear guideposts for lower courts concerning what the First Amendment covers and what it does not, those courts are forced to follow the Supreme Court’s lead and assume coverage—that is, assume the Constitution’s applicability to the allegedly expressive act in question—and then apply protection doctrine. This is the exact opposite order of decision elsewhere in constitutional law, where the analysis begins, and often ends, with whether the Constitution applies to the individual’s act. For example, in the substantive due process area, every case begins with a careful description of the right being asserted, followed by an analysis of whether that right is sufficiently fundamental to fall within the liberty component of the Due Process Clause.[35] Consider too the Second Amendment, as manifested in the District of Columbia v. Heller case—there the majority and dissenting opinions spent dozens of pages analyzing whether the Amendment covered the use and possession of guns in self-defense.[36] In other areas of constitutional law, if the answer to the coverage question is in the negative, the analysis in effect ends there, other than some pro forma analytical t-crossing and i-dotting with respect to whether the law in question is rational. Coverage, not protection, is the battleground upon which most of the constitutional questions of the day are fought—except battles over the freedom of speech.

So, there is little doubt that Free Speech Beyond Words identifies an analytic gap in the Court and academy’s First Amendment work. The next question up for answer is how well the authors fill it. As I discuss in detail below, the book’s three authors each make novel and valuable contributions with respect to how we should think about constitutional coverage for “speech” that does not employ language or convey articulable ideas. But the book as a whole can be read another way: as a broadside, and extremely effective, critique of listener-based First Amendment theories and doctrines that hang the presence or absence of First Amendment coverage on, as the Court stated in Spence v. Washington, whether “the likelihood was great that a [speaker’s particularized message] would be understood by those who viewed it.”[37] As the authors prove, the question whether an audience understands a speaker’s expressive act to convey a particular message fails to resolve, and even frustrates, meaningful inquiry into the issue of First Amendment coverage for works that do not use words or convey a particular idea. Indeed, perhaps a more accurate (though much less marketable) title for the book might have been Free Speech Beyond Spence.

This Review proceeds as follows. Part I sets out and comments on Chen, Tushnet, and Blocher’s arguments. In Part II, I offer my own alternative, arguably more orthodox, framework for analyzing claims to First Amendment coverage for what I call “non-speech”—works of art and other expressive acts that, like those discussed by the authors, do not use language to communicate specifically articulable ideas. In brief, and drawing from other areas of constitutional law, I argue that the First Amendment covers not simply the act of expression, but also what comes before: the autonomy-based right to decide what to say. As discussed above, John Cage’s use of silence in 4’33” was unquestionably an expressive act, even though it used no words at all. Locating the Speech Clause’s protections earlier in the expressive act permits First Amendment doctrine to rest comfortably upon the idea that the Amendment covers expression that does not use words but that nevertheless is intended to communicate.

4’33” may be expressive because of its deliberate use of silence, but it cannot follow that every use of silence necessarily constitutes an expressive act. Part III sets out potential problems associated with extending First Amendment coverage over expressive non-speech acts, pursuant either to the authors’ approaches or my own, in particular the need for limiting principles given a First Amendment theory and doctrine that seems, at least in recent cases and law journal articles, to only expand. If the First Amendment protects both speech and non-speech, and everyone seems to agree that it does, then the final need is for tools to identify when conduct is neither speech nor non-speech—to draw the boundaries of First Amendment coverage in a way that is clear and predictable to speakers, lawmakers, and reviewing courts.

1. Testing Coverage’s Boundaries
2. The Expressive C#

The first category of wordless expression that Free Speech Beyond Words takes on is instrumental music. The problem is straightforward, as expressed by Alan Chen, the author of this chapter: Why does the First Amendment independently cover the “musical, as opposed to [the] lyrical, component of such expression” (p. 15)? The First Amendment protects the lyrics of Jimi Hendrix’s Purple Haze, which came to Hendrix in a dream after having read a science fiction book called Night of Light by Philip José Farmer.[38] Chen states that by all accounts, the Amendment protects the song’s opening guitar riff, independent of the protection for the lyrics.[39] But why? If Hendrix had never put words to the song at all, why is it so clear that the First Amendment’s coverage would still apply? What is the theoretical basis for the conclusion that, for First Amendment purposes, instrumental music, which is wordless, is “constitutionally equivalent” to speech (p. 15)?

Despite that call of the question, however, the issue is not hypothetical. As Chen shows, governments that have censored music have usually done so due to the music’s lyrical content, though this has not always been the case (p. 27). Nazi Germany enforced its visions of cultural purity by banning the performance and broadcast of what it called “degenerate music,” with the race of the music’s composer as the usual stand-in for degeneracy. Strict Islamists often ban instrumental music because of its associations with secular lifestyles and interests. And public schools here in the United States have been sued for barring students from playing religiously themed instrumental music during school-sanctioned programs (p. 29).

Chen argues that constitutional coverage of instrumental music does not rest comfortably with any of the governing theories underlying the First Amendment. First, it is difficult to argue that music without lyrics serves either self-governance in the Meiklejohnian sense or public discourse in the Postian sense (pp. 32-33).[40] Democracy-based rationales for the First Amendment are, as Chen notes, “simply unhelpful” (p. 35) in justifying constitutional protections for artistic expression. Though Chen does not put as fine a point on it as this, these rationales for the most part fail because they depend in large measure on an exchange of information between speaker and listener; however, a composition that no one other than the composer heard at all, instrumental or not, is certainly covered by the First Amendment, so coverage cannot depend on the transmissionary aspects of expression. A prudish Minnesota legislator who sought to prevent any of the hundreds of hours of music in Prince’s vault from being released would have about as much success in defending against a court challenge as he would in trying to ban Darling Nikki from the radio and record stores.[41] Similarly, truth-seeking rationales such as the marketplace of ideas don’t do the job of justifying coverage of instrumental music, since, as Chen notes, music without lyrics does not usually communicate a specific idea; indeed, the primary expressive point of instrumental music is often to “say” what words cannot (p. 37)—though he does note that marketplace-based theories fit well with coverage for instrumental music if one broadens the definition of “truth” to consider aesthetic or cultural truth. To the extent that current First Amendment theory supports coverage of instrumental music, autonomy theory is the best fit. As Chen argues, a composer may intend to communicate an idea through the use of music, or instead “have no intent at all except to create something beautiful . . . .” (p. 43).

Having discussed the existing theoretical approaches to the coverage question, Chen then turns to specific rationales for First Amendment coverage of instrumental music that may or may not harmonize with those approaches. Chen notes correctly that performing and listening to lyric-less music have played essential roles in a host of areas of primary constitutional concern, from individual identity formation (think: a Deadhead spinning to the Grateful Dead’s Slipknot!) to cultural identity preservation (think: Tuvan throat singing by nomadic sheepherders in the Mongolian mountains) to nationalism (think: a march by John Philip Sousa). But Chen’s project is to justify an independent basis for First Amendment protection for instrumental music, and these examples do not provide especially compelling reasons supporting that distinction, since the analytical work they do is actually to collapse any distinction between instrumental and other kinds of music, or for that matter other forms of art. Hearing the Beatles on The Ed Sullivan Show played a critical role on the identity formation of millions of American youngsters. Whitney Houston’s Star-Spangled Banner at the 1991 Super Bowl moved as many hearts toward America as have Sousa’s marches.

Furthermore, some of Chen’s examples of government suppression of instrumental music are not a perfect fit for his argument that it needs independent protection in the first place. In most of the cases of government suppression that he discusses, the music was not banned for lacking lyrical content; Nazis did not ban instrumental music because it was instrumental, but rather because they were anti-Semites and racists who hated its composers for being Jewish or Black. The presence or absence of lyrical content in the music was irrelevant to the government’s efforts to suppress. So too the cases involving school boards’ banning of student performances of religious music; there, the schools’ concern was not banning instrumental music, but avoiding liability under the Establishment Clause. In the end, Chen’s chapter does not so much as supply an independent basis for instrumental music’s protection as argue that despite its lack of message-conveying lyrical content, instrumental music is actually not so different from other forms of expression after all (p. 66). That is an important First Amendment-related conclusion, and it certainly supplies an adequate rationale for the Amendment’s coverage. But if the goal here is to demonstrate what is different about instrumental music, then the argument is only partially successful. If instrumental music is censored for reasons indistinguishable from music more generally, then it is hard to see why we need a distinct basis for its First Amendment coverage.

Chen also works through an independent theoretical rationale for First Amendment coverage that focuses primarily on the music’s cognitive effects on listeners for instrumental music (pp. 46-48), and finds these justifications to be mostly incomplete. He is right that these theories do not do the trick, but it is for a more simple reason than the ones he gives. Listener-based theories for protecting nonrepresentational art like instrumental music inevitably fail because of their focus on listener effects over speaker intent. The best theoretical grounding for protecting instrumental music on the same footing as speech, one that I will present in more detail in Part II below, rests not with its effects, but rather with its creation. Embedded in instrumental music, and indeed in all conventionally nonrepresentational art, are artistic choices—place the coda here rather than there, run the riff through the fuzz pedal or not, or as Chen quotes Judge Frank Easterbrook, decide “the right place for the major third” (p. 23)[42]—that are themselves covered, on the ground they are no different, or no less expressive, than a lyricist’s choice to use the sun as a metaphor instead of the moon.

As Chen says, “instrumental music is a unique way of expressing and experiencing culture” (p. 68). But the upshot of his essay is actually that with respect to the theory and justifications underlying First Amendment coverage, instrumental music is not really that unique at all.

1. Nonrepresentational Art as “A Gesture of Liberation, from Value—Political, Aesthetic, Moral”[43]

While Chen’s chapter addresses questions about instrumental music, Tushnet’s chapter in Free Speech Beyond Words explores issues around First Amendment coverage of nonrepresentational art. Of the book’s three primary chapters, Tushnet asks most directly how to justify the Supreme Court’s conclusion in Hurley v. Irish-American Gay, Lesbian, & Bisexual Group of Boston that Jackson Pollock’s paintings are “unquestionably” protected by the First Amendment. And those justifications, Tushnet argues, expose several of the Amendment’s “unexpected facets” as well (p. 70).

Those “unexpected facets” include several interesting issues such as the tensions between First Amendment coverage questions and other doctrines such as copyright and trademark law (pp. 109-112) and commercial speech (p. 112). Tushnet’s primary contribution, however, is to demonstrate the many ways in which governing First Amendment theories fail to justify the Amendment’s application to nonrepresentational art. The assumed conclusion that all art is covered because it is expressive, Tushnet argues, necessarily falters because there are a range of other human activities that can be intended to communicate as well—many of which no court would assume are covered by the Speech Clause (p. 85). Ticket scalping; panhandling; selling bootleg T-Shirts at Fenway Park that say “Yankees Suck”[44]; scolding one’s children in public; even political crimes like graffiti or assassination—all of these activities can be spun by the people engaged in them as both expressive and autonomy-affirming, and thus for First Amendment coverage purposes, no different than art.[45] Additionally, as I discuss in more detail below, nonrepresentational art, both on the canvas and via the mission statements of many of its practitioners, consistently declares itself as directly opposed to the communication of ideas (p. 75). A speaker’s “intent to express something” alone thus cannot provide the justification for First Amendment coverage of nonrepresentational art.

To address these questions, Tushnet excavates the presumption that all of art is covered by the First Amendment, and details how that presumption can carry over to art that is not intended to convey a particular message. The presumption, Tushnet notes, arose for the first time while the Court was wrestling with free speech questions concerning, of all things, obscenity. In deciding what was obscene in cases such as Miller v. California,[46] Tushnet notes, the Court “assumed that material that can be described as sufficiently artistic cannot be obscene” (p. 102)—in other words, the Court defined obscenity by what it was not, namely, art. Art was covered by the First Amendment, so if obscenity was outside the Amendment’s coverage, it could not be artistic.[47] Accordingly, if the presence or absence of artistic value determines whether a particular expression is covered by the First Amendment, then the Court is primarily not talking in those cases about the freedom of speech, but rather the freedom to make art.

But who decides what art is? Nonrepresentational art presents a difficult puzzle for the project of defining artistic value, because it is not intended to represent any one reality, and thus is aimed at subjective meaning-making more so than other more conventional forms of art. Abstractionists like Pollock abandoned representational art’s fidelity to the real appearance of things in order to deliberately depart from shared meaning; as the quote opening this Section notes, the primary motivating principle for much of nonrepresentational art is to free artmaking from the constraints of aesthetic value.[48] Similarly, critics often describe experimental art as not simply removed from a common set of aesthetics, but anti-aesthetic—representing “a position of openness, of inquiry, of uncertainty, of discovery.”[49] The experimental artist’s intent is usually not to communicate an idea in the conventional sense, but to facilitate an experience in each viewer or listener that is differentiated rather than shared. In nonrepresentational art, this mission is usually accomplished through the “deliberate abandonment of established forms.”[50] The communicative dynamic intended by the artist/speaker, in other words, is one-to-one rather than one-to-many. In contrast to most other kinds of expression covered by the First Amendment, the whole point of much nonrepresentational art is not to impress a particular meaning upon viewers.

Despite this conflict, however, Tushnet offers a way forward. By working through Justice Souter’s opinion in Hurley, Tushnet notes that for the First Amendment to cover, the “meaning” that observers take from a given expressive act need not be “univocal” (p. 104). “The reasonable observer,” he writes, “must understand that the object on view is expressive, though not all observers will agree on what it expresses” (p. 104). Some viewers must understand the expressive act to be intending to say something—even though there is little shared understanding as to what that message precisely is.[51] This interpretation, which Chen also articulates with respect to instrumental music, takes listener autonomy seriously, but it is listener autonomy in its most radical form: each individual listener or viewer is entitled to her own interpretation of a nonrepresentational work.

So, in the context of art, not all viewers have to believe a particular expression has artistic value in order for the First Amendment to cover it. Coverage theory is free from Spence’s uniform listener meaning requirement. Some can see Duchamp’s Dadaist ready-made Fountain and notice how the curved porcelain repeats and refers to the shapes used in representational depictions of the female form displayed elsewhere in the museum.[52] Others will see what one art critic called a severing of “the traditional link between the artist’s labor and the merit of the work”[53]—and perhaps notice, like John Cage hoped his audience at Maverick Concert Hall would in 1952, the aesthetic beauty displayed throughout everyday industrial life that modernization has trained them to ignore. Others will see a urinal turned on its back and wonder if a plumber forgot something at his job site.

Accordingly, the fact that a work of art’s meaning is inscrutable or contested is not a barrier to constitutional coverage. Tushnet’s argument that First Amendment coverage for nonrepresentational art need not rely on a shared aesthetic meaning thus shares an elegant symmetry with nonrepresentational art itself. But it still carries a thread of the very claim that the radical autonomy theorists that birthed nonrepresentational art seemed to reject: that art must be understood by enough people, or the reasonable (read “right”) person, as saying something in order to be covered. We can surely conclude now that Duchamp intended to say something with Fountain in 1917. But what if no other person at the time saw anything other than a urinal?

The First Amendment definitely covers Fountain now that its meaning, or many meanings, to be as precise as Tushnet advocates we should be, can be discerned from the work. But was that necessarily the case in 1917, when we can’t be so sure that anyone saw anything being communicated at all? If nonrepresentational art is usually ahead of its time, then is it usually ahead of the First Amendment as well?[54] It would be odd indeed if the theory justifying First Amendment coverage of nonrepresentational art provided a weaker justification for coverage at the time the art was made, since this would be the point at which the art was most in need of protection from government censorship.

Speaking of censorship, of the three forms of expression discussed in Free Speech Beyond Words, Tushnet’s case probably presents the most to fear. On the one hand, he is correct to note that “[d]irect regulation of artworks as such is rare, and what there is almost always takes the form of content-neutral regulations that readily pass the relevant doctrinal tests” (p. 109). On the other hand, governments have shown time and again a willingness to impose their standards of what constitutes “good” art, and nonrepresentational art usually finds itself across from “good.”

All of us of a certain age will recall when in 1999, then-New York City Mayor Rudy Giuliani threatened to terminate the city’s funding for and lease with the Brooklyn Museum of Art for the museum’s hosting of the British exhibition Sensation, a show which included, among other works, artist Chris Ofili’s painting of a black Virgin Mary that incorporated elephant dung.[55] To be sure, Giuliani’s complaint was primarily due to what he viewed as the offensive representational aspect of Ofili’s work, but many of the mayor’s comments at the news conference discussing his opposition to Sensation leveled charges that could just as easily be applied to nonrepresentational works—comments like “[a]nything that I can do isn’t art,” and “[i]f I can do it, it’s not art, because I’m not much of an artist.”[56] Similarly, in the 1930s, the New Deal’s Public Works of Art Project subsidized more than 15,000 paintings, murals, prints, crafts and sculptures, all of which depicted what PWAP’s requirements called the “American scene,” and nearly none of which were abstract works. Artists on the federal payroll were instructed that their works should not include anything “experimental [or] unconventional”—evidence of an “administrative bias against abstractionist practices in American art.”[57] During the Second Red Scare, Michigan Congressman George Dondero condemned the foreign roots of nonrepresentational art, and deemed abstraction as “inherently threatening to American values and ideals.”[58] Dondero attacked nonrepresentational art from the floor of Congress, claiming that “[c]ubism aims to destroy by designed disorder . . . Dadaism aims to destroy by ridicule . . . [a]bstraction aims to destroy by the creation of brainstorms . . . [and] [s]urrealism aims to destroy by the denial of reason.”[59] And though Tushnet focuses more on the conduct of domestic governments than Chen, there are international examples here as well; Fascists and Stalinists criticized nonrepresentational art as dangerously opposed to mass culture.[60] Unlike Chen’s examples, most of which involve government censorship of music for reasons largely incidental to the music’s lack of lyrics, the nonrepresentational nature of the art being attacked here is the very reason for governmental scrutiny and scorn. So, the need for a freestanding theory justifying First Amendment coverage for nonrepresentational art is genuine, and Tushnet does essential work in providing one.

1. “All mimsy were the borogoves”[61]: Sense in the Nonsense

Finally, in a chapter by Joseph Blocher, Free Speech Beyond Words considers the First Amendment coverage question through the lens of nonsense speech. Of the three authors, Blocher has given himself by far the hardest swath to mow. Scholars and the Court have often tied First Amendment coverage to the presence or absence of meaning (p. 114); nonsense, based on this framing, is meaning’s opposite. Blocher uses nonsense as a tool to cast considerable doubt on the reflexive notion that “meaning is a prerequisite for constitutional coverage” (p. 115). In particular, he shows the damage it would do to the First Amendment if that notion were correct. Examining nonsense from a theoretical perspective demonstrates that meaning does not, and indeed cannot, carry the constitutional weight that many commentators and judges have assumed it does.

Blocher begins by distinguishing among types of nonsense, since the rationales for First Amendment coverage might vary depending on the particular kind of nonsense at issue. First, there is what Blocher calls “overt nonsense,” namely, the “intentional and apparent” use of words that, individually or in combination, “lack[] meaning” (pp. 120-121), or the Jabberwockys or William Burroughs-inspired cut-up poems of the world.[62] This labeling is consistent with critical commentaries on nonsense literature, which understand “nonsense as a deliberate strategy”[63]—or, to use the terminology of this Review, the use of nonsense as a deliberate communicative choice to favor rhyme, tonality, alliteration, rhythm, or other kinds of acoustics-based wordplay over linguistic-based semantic meaning. There is also “covert nonsense” (p. 121)—those expressive acts where a speaker intends meaning, but the meaning is not received, or if a meaning is received, it is not the one the speaker intended. And though Blocher does not focus on this problem, failing to lock down First Amendment coverage for nonsense is problematic because it is a very short step from Jabberwocky’s overt nonsense to uses of words that sound or read more like speech, but nevertheless make no sense. One need not leaf through a compilation of obscure online experimental flarf poetry to appreciate this point.[64] Here is a sample of the lyrics from I Am Trying to Break Your Heart, the opening song from the band Wilco’s Yankee Hotel Foxtrot, one of the most critically acclaimed rock albums of the 2000s. You tell me what they mean.

I want to hold you in the Bible-black predawn

You’re quite a quiet domino, bury me now

Take off your Band-Aid because I don’t believe in touchdowns

What was I thinking when I said hello?[65]

In a doctrinal world where First Amendment rules and theory are meaning-dependent, coverage of nonsense—even speech that is intended to have meaning but winds up being nonsense for the reasons Blocher describes—is at risk. However, as he notes, there is little debate that coverage does indeed reach nonsense of all kinds, or more specifically that stream-of-consciousness composition, automatic writing, and surrealist literature are “speech” for First Amendment purposes. To resolve this inconsistency, Blocher turns away from law and towards analytic philosophy—particularly its rejection, primarily via the work of Ludwig Wittgenstein in the late 1940s and early 1950s, of the concept of “representational meaning” (pp. 116-117).

By representational meaning, philosophers meant that language only had meaning when it connected to “extralinguistic concepts” (p. 133)—namely, in the words of Oliver Wendell Holmes, “the facts for which they stand” (p. 134).[66] In other words, as Wittgenstein argued, “the limits of my language mean the limits of my world” (p. 136)[67]: because the existence of reality depended on the use of language to describe it, language that did not describe reality had no meaning. But in much the same way as nonrepresentational artists began seeing artmaking as its own end rather than as simply a means to make references to things that previously existed in the visual world, the analytic thinkers began to see the defects and limitations in representational meaning, and adopted a different approach: language’s meaning depended not on its “representation of the world,” but rather on its use— “whether or not” the speakers have “followed the rules” of what Wittgenstein called “the relevant language game” in which the speakers are communicating (p. 140). If the speakers’ use is following the applicable socially understood rules of the applicable language game, then their use of language will be processed and understood by their co-party or co-parties in the communicative act. On the other hand, if the speakers’ use does not follow those rules, the use results in nonsense. My uses of the terms “NAFTA,” “Obamacare,” or “LBGTQIA” follow the rules of the game of “faculty lounge language,” and thus have meaning, but under the rules of the game of “father-young daughter language,” my uses would produce nonsense.[68]

This shift is instructive for law, Blocher claims, because current First Amendment doctrine relies upon representational meaning-based analysis in several foundational respects. Whenever a speaker, scholar, or court argues that acts are covered by the First Amendment if they “convey[] ‘ideas’ or ‘information’” (p. 137),[69] they are falling into the same trap that the analytic theorists worked through when they abandoned representational meaning in favor of a focus on language’s use. Spence v. Washington, with its dual requirements that a speaker “inten[ded] to convey a particular[] message” and there existed a great “likelihood . . . that the message would be understood by those who view[] it” in order for the First Amendment to apply, is the apex of the representational meaning approach (p. 138).[70] But by tying coverage to representation-based meaning, nonsense, some forms of symbolic speech, and a range of other expressive activity are by implication left outside the First Amendment altogether.

In defending his premise that the First Amendment must cover nonsense, Blocher is not as quick to question listener-based First Amendment values as are Chen and Tushnet. Protecting nonsense is consistent with marketplace theory, Blocher argues, because “if the marketplace model requires judges to be agnostic as to truthfulness, it seems they should also be agnostic as to meaningfulness” (p. 125). On first glance, this reads like autonomy theory wrapped in a marketplace-of-ideas blanket. If the marketplace of ideas model protects nonsense because the function of the market is not to facilitate the search for objective fact but rather solely what John Stuart Mill called “the likings and dislikings”[71] of individual reason and judgment, then there isn’t much if any daylight between the collective search for truth and the individual search for self-fulfillment through expressive liberty. But Blocher is absolutely correct to argue that the argument over whether nonsense in fact has value, expressive or otherwise, is a form of searching for truth (p. 126), in much the same way as Tushnet shows that arguing over what Duchamp meant by the Fountain is evidence that the thing being argued over itself has First Amendment value. This argument, however, has its own limitations.

Nonsense, or at least what Blocher calls overt nonsense, is usually an explicit rejection of the premise that words have a communally agreed upon meaning. As Hugo Ball wrote in the Dadaist Manifesto excerpted above for the epigram to this Review, “words are other people’s inventions . . . I want my own stuff.”[72] It is certainly so that many have used the Manifesto, or Carroll’s Jabberwocky, or Wittgenstein’s Tractatus Logico-Philosophicus, a work that Blocher describes as proudly “proclaim[ing] itself to be nonsensical” (p. 126), in their own explorations of what truth is, or what it can be. But to return to an argument made above in response to some of Tushnet’s claims, what about the overt nonsense that no one argues about at all? Claims that First Amendment coverage attaches to nonsense because “meaningless speech . . . can have value as a means to truth” (p. 127), standing alone, is a version of the “art is important if it has aesthetic value” that nonrepresentational artists and nonsense writers expressly rejected.

This value-based argument is what Chen in the context of instrumental music calls an “associative claim.” It attaches First Amendment coverage not to the expressive act itself, but rather to the “association” of that act with “particular events or historical contexts” (p. 34)—here, the debates over those acts, the contexts in which they were made, and the degree to which they contribute to later debates. Blocher has to be arguing that the First Amendment covers useless as well as useful nonsense, on the ground that usefulness might one day blossom where once there was only uselessness. Because he can’t be arguing that the case for First Amendment coverage exists only when nonsense is useful. Most nonsense is useless, and proudly so.

Additionally, Tushnet’s “mutivocal” coverage solution described above seems to be at least partially responsive to Blocher’s covert nonsense problem. Under a multivocality theory, First Amendment coverage will attach if some listeners ascribe some meaning, or more broadly some intent to make meaning, to the speaker’s communicative act. Accordingly, looking for meaning per se in determining the existence or absence of First Amendment coverage does not rely on any common understanding on what that meaning might be. And Tushnet does not need analytic philosophy to reach this conclusion. Indeed Blocher, like Tushnet, points to Hurley as a “rejection of the representational approach and an endorsement of the idea that meaning lies in form and use” (p. 142).

Blocher argues that, in effect, the Speech Clause covers an individual’s right as to whether to make sense or not. The First Amendment’s primary concern is not with what we eventually say, but rather in the decisions that precede the communicative act. Which brings us back to John Cage.

1. Non-Speech as the Choice Not To Speak

It is no coincidence that Tushnet and Chen, and to a lesser extent Blocher, all independently come to the conclusion that the collective value-based theories underlying the First Amendment are mostly inadequate for explaining why the Amendment covers instrumental music, nonrepresentational art, and nonsense writing. The making of art is an individualistic act that manifests the author’s expressive choices, and thus any theory whose justifications are essentially collective in nature, or depend on others’ identifying or interpreting those choices, will not suffice. Collective theories concern themselves primarily with the exchange of ideas between the speaker and the audience; however, First Amendment coverage cannot depend on an audience’s interpretation of the artist’s communication, since as Tushnet, Chen and Blocher show, the meaning-making of the audience is often independent of the meaning (or lack of meaning) that the artist intended. We are all deconstructivist free speech libertarians now.

In Part II.A below, I argue that the First Amendment covers an individual’s decision as to whether or not to speak, whether through non-use of words or via inaction. If this is true, then the Amendment necessarily also covers one’s right to express oneself without words. Part II.B discusses another area of constitutional law, namely the Fifth Amendment, where the connection between autonomy and silence is better theorized. Later, in Part III, I attempt to draw some limiting principles around these arguments.

1. Fundamental Rights-Based Decisional Autonomy and Free Speech

At its core, underlying the First Amendment’s coverage of expressive acts such as nonrepresentational art and instrumental music is a theory that resonates more with current substantive due process doctrine than that of free speech. Governing much of the Court’s Fourteenth Amendment jurisprudence is an understanding of the liberty component of the Amendment’s Due Process Clause that protects the individual’s decision-making authority on issues critical to that individual’s self-conception and actualization—decisions on issues that are so fundamentally personal in nature that the Court has long said the government’s lawmaking authority cannot reach them.

Furthermore, and critically for present purposes, the Court’s fundamental rights jurisprudence under the Fourteenth Amendment recognizes that the sub-rights around that decisional core, such as the right to decide to conceive a child or to marry, for example, also include the inverse sides of those same rights: the decision not to conceive, or not to marry.[73] This view of fundamental-rights-protection-as-decisional-rights-protection recognizes that the Constitution necessarily protects inaction to the same degree as action, as both are expressions of autonomy and free will. A law that obliges married couples to use contraception, for example, is just as constitutionally defective as one that bars them from doing so, and for the same reasons.

The First Amendment’s coverage is similarly multi-sided. Even though the First Amendment expressly applies to the freedom of speech, its coverage and protection must necessarily apply to “non-speech”—acts that because of a deliberate choice by the speaker do not use words, but nevertheless are intended to communicate a message. It must cover both the decision to use a particular set of words and the decision to not use words, including the decision to be silent. It follows that the use of words or not as a proxy for the presence or absence of expressiveness is a red herring, because the decision to not use words is itself an expressive act covered by the First Amendment.

To be sure, cases like West Virginia State Board of Education v. Barnette affirm that the Speech Clause protects one’s right not to speak a government message, there the right to not recite the Pledge of Allegiance, if one’s conscience so compels.[74] Twenty-five years later, in Wooley v. Maynard, the Court declared that

the right of freedom of thought protected by the First Amendment against state action includes both the right to speak freely and the right to refrain from speaking at all. . . . The right to speak and the right to refrain from speaking are complementary components of the broader concept of “individual freedom of mind.”[75]

Likewise, Hurley itself, to which both Tushnet and Blocher give extended attention, is fundamentally a right not to speak case. There, parade organizers did not want to include the gay and lesbian Irish-American group that sought to participate in the parade, because the organizers believed that forcing them to do so would force them to express approval of that group’s message.[76]

The right to non-speech relates to the right to be free from articulating the government’s message, as recognized in Barnette and Wooley, but it is not coextensive with that right. Where the choice to be silent has a distinct expressive value, the interest being protected is not simply the conscience-related right to speak one’s own mind or not in lieu of another’s specific message. An example of the First Amendment covering one’s right not to speak would be presented by Louis Michael Seidman’s proposal of a hypothetical law mandating that Americans cast votes on Election Day.[77] The right implicated here is slightly different from that described in Barnette, Wooley, and Hurley, where the rights at issue were to be free from government “restrict[ing] or modify[ing] the message” a speaker “wishes to express,” “bear an offensive statement personally,” or “affirm a [specific] moral or political commitment.”[78] Even though the government’s compulsory voting law is not telling the individual for whom to vote, there is no doubt that the First Amendment would cover an individual’s decision to not comply with the statute’s mandate. Through its power of coercion, the state is in effect trying to force the individual to speak where she prefers to remain silent. And this silence has expressive content, particularly with respect to the social context of nonvoting as a commonly understood example of protest through inaction. A similar infringement of a right to not speak would be presented by a hypothetical public financing system that directed a small portion of taxpayers’ payments to go toward presidential candidates’ campaigns. There, the speaker is being forced to communicate a message, by way of a political contribution, that the First Amendment protects the speaker’s right to decide not to make.[79]

Even though “withdrawing in disgust is not the same as apathy,” both sentiments, though opposites, are often expressed in the same way—namely by not speaking at all.[80] And the First Amendment has long been recognized to cover anonymity, or the right “not to reveal one’s identity when exercising one’s affirmative right to express oneself”; as Martin Redish writes, anonymity is a “traditionally recognized subcategory of the constitutional guarantee of silence.”[81] Like silence, anonymity, “as part of the suite of editorial decisions about what to include or exclude . . . is itself a speech act.”[82]

Assuming this is all correct, it is a short step to go from finding that the First Amendment covers “non-speech” to finding it also covers nonverbal or nonrepresentational expressive acts like those discussed in Free Speech Beyond Words. Inaction, or non-speech, is one outcome of autonomy’s exercise, but so too are other expressive choices. The Amendment covers David Levine’s decision to draw a caricature of Henry Kissinger having sex with a woman with a head shaped like the Earth’s globe and call it “Screwing the World,”[83] but it also protects Jackson Pollock’s decision not to paint a similar portrait, and to splatter paint instead. If the First Amendment protects Woody Guthrie’s decision to join lyrics to music in the service of social justice, it must also protect John Cage’s decision to express himself through 4’33”’s deliberate absence of sound. A right that covers how one speaks must also cover whether one speaks at all.

However—and consistent with many, but not all of the arguments made in Free Speech Beyond Words—the presence or absence of First Amendment coverage for a particular speech act cannot turn on whether the decision to speak or not has communicative content as assessed by someone other than the speaker. As discussed above, it is easy to imagine acts intended to be expressive that no reasonable observer would understand to be even communicative, let alone to express a “particularized message,” per Spence. Including listener understanding as part of the calculus for First Amendment coverage purposes cuts against the commitment to expressive autonomy, as illustrated in cases where the expression in question does not use words to convey meaning.

Other constitutional contexts have better explored the theoretical connection between autonomy and silence. Next, I will briefly touch on one such context and see what we can learn from it.

 

1. The Autonomy-Based Right to Remain Silent

Another area of constitutional law in which the connection between silence and autonomy is more fully theorized is in the Fifth Amendment context, in particular the post-Miranda v. Arizona development of the right against self-incrimination. Miranda itself relied on the proposition that an individual possesses a right to remain silent “unless he chooses to speak in the unfettered exercise of his own will,” and as Lisa Kern Griffin notes, Miranda “references the concept of free choice nine times.”[84] Griffin argues that “misperceptions about what silence communicates” prejudice criminal defendants, and that silence does not indicate that the speaker has nothing to say, but rather “indicates the need for a space within which to make choices,” protects the “freedom to choose what to say to whom and when to say it,” and “leaves room for individuals to form their own plans.”[85] Similarly, the aforementioned work of Louis Michael Seidman notes that because silence is the predicate for speech, silence must be protected “in order to give meaning to speech . . . . [F]or speech to be truly free, there must also be silence.”[86]

Silence and choice, in other words, are inextricable. In the view of these and other criminal justice scholars, in the Fifth Amendment context, focusing on the observer’s interpretation of the expressive content of silence or lack thereof is not part of the analysis; it is part of the problem. The expressive value of silence, as well as other forms of “non-speech,” cannot be left to the ear or the eye of the beholder. Indeed, doing so can be, and often is, prejudicial to the speaker who has chosen to express herself through silence rather than words.

There is much that First Amendment theorists can learn from this approach. At first glance, the stakes with respect to the First and Fifth Amendments are very different. The risks around self-incrimination relate to “what the state does to criminals . . . . [W]hen a person incriminates herself, the state puts her in prison, or perhaps even takes her life.”[87] However, determining whether the First Amendment covers a particular act can certainly decide whether one’s liberty is at risk. Gregory Lee Johnson, the petitioner in Texas v. Johnson,[88] was initially sentenced to a year’s incarceration for burning an American flag outside of Dallas’ City Hall. When the Court was assessing whether or not Johnson’s conduct was covered by the First Amendment, they were literally deciding whether or not he would go to prison. Despite its citing of the Spence test, the Court’s focus was accordingly on Johnson’s intent to communicate and the “context in which [his conduct] occurred.”[89] To be sure, the Court also noted that the “overtly political nature” of Johnson’s act was “overwhelmingly apparent.”[90] But First Amendment coverage could not rise or fall on Johnson’s audience’s understanding of his message—or whether they understood it to be a message at all.

III. A Search for Limiting Principles

It is simple to state, as the Supreme Court has, that free speech “means more than simply the right to talk and to write.”[91] But finding the border between expressive and nonexpressive conduct has been a much more difficult task. That difficulty is compounded by, among other things, the standard approach to current First Amendment problems, which is discussed in more detail in this Part.

1. The Specter of First Amendment Expansionism

The arguments such as those made in Free Speech Beyond Words and in this Review defending the Speech Clause’s coverage of nominally non-expressive acts must be viewed within a broader First Amendment context. That context is one where interpretations of the Amendment’s applicability and protections nearly always expand rather than contract. As Tushnet notes in an earlier article, First Amendment scholars, at least historically, have long been collectively biased in favor of First Amendment expansionism. He writes:

Scholars of the First Amendment seem to “like” the Amendment [in a way that scholars of other constitutional topics do not “like” the Amendments that are the topic of their study]. . . . What I mean by “liking” the First Amendment is something like this: A First Amendment scholar hears that a court has held that some local ordinance or state statute violates the First Amendment, and his or her initial thought is that the decision is presumptively correct.[92]

This bias, Tushnet claims, works as a one-way ratchet with respect to First Amendment scholarship’s, and, by extension, lawyers’ and judges’, views as to the Amendment’s reach. Individual arguments concerning the legality, or even the normative merits, of particular regulations that might be argued to infringe upon speech run up against the unified presumption that most if not all expressive acts are covered by the First Amendment, and that those acts enjoy maximum protection. This is a systemic problem, Tushnet claims. It leads to a faux balancing where the government’s regulatory interest always loses, or as Justice Scalia might have said, the dice are always loaded in favor of free speech.[93]

Most interestingly, Tushnet also argues that even though many (most?) of those same scholars believe that the Roberts Court’s expansion of the First Amendment has taken a wrong turn in areas such as campaign finance and commercial speech, the reasons the scholars give are based not on the competing considerations and counter-majoritarian difficulties wrestled with across other areas of constitutional law, but rather on the Amendment itself. Citizens United v. FEC and McCutcheon v. FEC are wrong not because the Court showed insufficient Thayerian deference to elected officials in a core area of legislative expertise, but rather because the Court ignored the free speech rights of “less wealthy and powerful speakers” to favor the rights of “wealthy donors” and “business corporations.”[94] Even cases expanding the First Amendment are attacked as wrong on expansionist grounds. The scholarly debate is not about whether judges are over-steering at the First Amendment wheel, but rather about the direction of the resulting drift.

This trend could fuel the implicit discomfort that some readers might have with respect to the consequences of Chen, Tushnet, and Blocher’s conclusions. If the freedom of speech does not depend on the use of literal speech, and the judiciary and academy both bring heavy priors to First Amendment questions in favor of both coverage and protection, then governmental interests in a range of cases will be deprived a fair hearing. It also makes finding limiting principles in First Amendment coverage even more pressing. Below, I first point out a few of the questions that a choice-based coverage theory immediately raises, and then attempt to offer a few such principles.

1. Problems with a Choice-Based First Amendment Theory

If the First Amendment protects one’s decision not to speak, then why would it not protect other failures to act that were for expressive reasons? Would one’s failure to file federal income taxes now be protected by the Speech Clause, if the reason for doing so was to conscientiously object to the military effort funded by those taxes? What about a refusal to report to jury duty as an expression of disbelief in the criminal justice system, or a host of other choices to not act that could be ascribed expressive content by those making them? To use one of Tushnet’s examples from the book, one could easily envision small businessmen claiming that their decision to run businesses of a certain kind is intended to communicate a message about income inequality and corporate power (p. 72). (If you don’t believe this is plausible, ask the couple selling raw milk at your local farmer’s market what they think about the milk at your town or city’s largest grocery store.) Similarly, Blocher argues that “[t]he expansiveness of the autonomy conception leaves its defenders with a vast territory to patrol, because nearly any act can be described as a manifestation of individual autonomy” (p. 128).

In addition, a fully realized, autonomy-based rationale for putting non-speech on equal footing with speech for First Amendment coverage purposes could potentially do significant damage to the regulatory state, particularly with respect to compelled disclosures. If a meatpacker is protected from government interferences by its right not to speak, then FDA-imposed labeling requirements might violate the First Amendment. As Jed Rubenfeld wrote more than fifteen years ago, “[p]eople constantly want to violate laws for expressive reasons.”[95]

In a related argument, Tushnet also warns us about rationales for First Amendment coverage that rely too much on speaker intent, claiming them to be “both over- and under-inclusive” (p. 83).[96] With respect to overinclusivity, all manner of nonartistic activities are often intended to express some point, from the street scalper striking his own tiny blow against Ticketmaster to the leather-jacket-wearing loiterer on the street corner hoping to convey his coolness. If First Amendment coverage is triggered by intent to express, then it must follow that it covers all manner of conduct that no reasonable person would put on equal footing with speech, which seems wrong. And with respect to underinclusiveness, he argues that many artists do not intend to express any particular message at all; the “intent” behind much of art, especially nonrepresentational art, is to explore the relationship between shape and form, or following Cage and Rauschenberg, to cause the viewer or listener to refocus their attention in a new way, or simply to make something that has never been made before. However, there is little question that such works are covered. An intent-based rationale is thus underinclusive if it would fail to provide a basis for covering those acts, even those that adhere to artistic genres, which are not intended to express anything at all. This too seems wrong. Hence the problem with intent as the driving force for deciding coverage questions.

1. Setting Autonomy’s Limits

Both the authors of Free Speech Beyond Words and I argue that the tools for First Amendment coverage questions need further development so that in some cases First Amendment protection questions can be avoided entirely. If that is so, what might constitute what we could call a “First Amendment Step 1” case, where a court would find coverage over a particular act or inaction is not present, and thus there is no protection question to which to proceed?[97] If a coverage rule or theory is too vague or too capacious, we wind up at exactly the point at which we began, and which motivated Chen, Tushnet, and Blocher to write their book—assume coverage exists, then let protection doctrine do all the work. But there are some tools with which to solve this puzzle, some of which are already a part of First Amendment doctrine, and some of which flow from taking choice seriously as a justification for First Amendment coverage.

First, with respect to underinclusivity, a commitment to and thoughtful application of choice-based coverage theory can resolve most concerns. First Amendment coverage for art lacking intent to convey a particularized message may run contrary to the dictates of Spence v. Washington, but it fits comfortably within a choice-based coverage rationale. As I have stated, the basis for First Amendment coverage is that the Amendment, like the liberty component of the Due Process Clause, applies to expressive choices, not the outputs of those choices. If output is not outcome-determinative with respect to coverage, then the listener’s role in the coverage inquiry recedes, and the question whether a message would have been understood by its audience is not dispositive. The decision to make art or not, and the lesser included decision to make art that represents real life vs. art that is a technical experiment in form of the type Tushnet describes, is necessarily covered. And the reasons for the government’s infringement on speakers’ exercise of their right to decide whether and how to speak can also bear on coverage questions—an inquiry that can apply not just to compelled speech contexts, but to others. For example, a mandate that “no one shall make nonrepresentational art” when the government has deemed such art to be contrary to American ideals presents a clear ground for constitutional coverage for noncompliance.

At the same time, however, it cannot be the case that every silence is presumed covered ipso facto because silence is one necessary condition for making expressive choices. Inaction is an actualization of liberty, but there is not pure symmetry between action and inaction for constitutional purposes outside the context of the Due Process Clause; there, inaction is protected only in relation to the right at issue.[98] Forced sterilization by the government violates the Fourteenth Amendment’s liberty component, but so does forced procreation.[99] So too with the Fifth Amendment; the relevant inaction there is in the context of declining to respond to interrogation in the interest of avoiding self-incrimination.[100] In both domains, government compulsion or proscription through what Seana Shiffrin terms “legal materials”[101]—“laws, regulations, court rulings, and resolutions”[102]—is what provides the basis for finding that a symmetry between action and inaction exists.

There is also something different about these rights that compels the conclusion that their protections are symmetrical—that they extend to both action and inaction around that right. Blocher writes in another article that “[a] choice right is a right to do both X and not-X, where not-X means refusing to do or accept X.”[103] A particular right is deemed a “choice right” in “light of the purposes” the relevant amendment is “meant to serve”[104]—and if those purposes point to protecting choice, then the symmetry manifests through the proscription of both government prohibition (restraining one from doing X) and coercion (forcing one to do X). In other words, certain rights are “bilateral”—with respect to such a right, “the freedom to do x entails the freedom to not do x.”[105] This is not the same, however, as a “generalized right against coercion.”[106]

The right to freedom of speech is a quintessential choice right. Its primary purpose is to serve not just the expression of one’s views, but also the formation of the views one may decide to express. It protects freedom of mind and thought, not simply freedom of speech.[107] Those interests require bilateral protection. Accordingly, the First Amendment is implicated whenever legal materials interfere with expressive choice, including when the choice is not to speak.

Of course, this conclusion likely proves to some that speaker intent-based coverage rationales are necessarily over-inclusive, a charge made by both Tushnet and Blocher.[108] It is certainly true that all kinds of conduct can be undertaken with the intent to communicate a message. If intent to communicate is the trigger for First Amendment coverage, conduct that has long been thought outside of the Speech Clause’s reach brings the First Amendment into the picture, requiring what Tushnet calls “at least a tiny increment[al increase] in the justification for regulation” (p. 85) in a range of areas. And if, as I have advocated, the coverage question is rooted in and informed by the liberty to choose to decide what to say, including whether to speak or not, then, to use Blocher’s phrase, the territory that autonomy defenders must patrol may have expanded even more. In other words, my proposed solution to the underinclusivity problem compounds the overinclusivity problem.

However, making “the choice of the speech by the self the crucial factor in justifying” constitutional coverage does not necessarily enable a freestanding First Amendment defense to safety regulations, criminal conspiracy, or antidiscrimination law.[109] By asking whether a particular act was primarily intended as communicative, we can keep most of the corpus of criminal law, for example, outside of the territory to patrol. The conclusion that most of the law punishing verbal commission or participation in crimes does not implicate the First Amendment is justified because even though that speech is the product of decisional autonomy, the primary interest in deciding to make it—the speaker’s interest in the communicative act—is to facilitate private gain, usually (though not always) at the expense of another, or to otherwise infringe on another person’s autonomy in a way that the criminal law protects.[110]

In application, Tushnet’s ticket scalper may have decided to scalp over other forms of income because he believes Ticketmaster and Live Nation are evil gougers. My “Yankees Suck” T-shirt hawker at Fenway Park certainly believes the message on his shirts to be both true and deserving of wide dissemination. At the end of the day, however, the primary fruits both intend to come of their labor are not social change, monopoly-busting, or relocations from Manhattan to Boston, but more cash in their respective pockets. Forcing them to get a peddling license, limiting them to certain areas outside the ballpark, or banning them altogether at risk of criminal fine pursuant to generally applicable laws and regulations are thus unproblematic, because their communicative choices to scalp or to sell T-shirts with one message over another are not covered, even though those acts could be argued to have been products of expressive choice. Deciding to wear a T-shirt that contains a message is different, even under a choice-based coverage theory, than deciding to sell that same shirt. Other factors that serve as reliable proxies for speaker intent, such as the form the speech takes and the speaker’s desired audience, could potentially play a much greater role in coverage questions than they do now.[111]

There are also, of course, externalities to constitutional coverage for silence—and some of those externalities can be negative. To find that the First Amendment materially impedes government from forcing individuals to speak, especially in the disclosure context, can adversely impact a range of decisions by third parties, such as employees’ awareness and understanding of their legal rights with respect to safety or other employment terms and conditions.[112] It may be the case that when silence is intended to mislead or obfuscate, it can be treated the same way as verbal speech intended to do the same, on either the coverage or protection end.

Whether it is information about an employee’s right to organize or disclosures to prospective purchasers of a home, covering communicative silence should not frustrate societal goals that are more important than the individual’s right to withhold relevant information.[113] In situations like these, distinctions between the right to not speak and the right to be free from compelled speech can be helpful. With respect to compelled disclosures, courts could differentiate between (1) forcing someone to speak when they would choose to communicate via silence, and (2) “anti-silence” regulatory schemes that recognize those circumstances in which protecting silence can entrench informational asymmetries that have significant costs for prospective listeners.

The point here is not to provide an all-encompassing answer to the issue of First Amendment coverage that fits every case. (For example, if the “intent is primarily to get more cash in my pocket” from above is my proposed non-coverage rule, then most corporate political speech, as well as all of commercial speech, would be outside the First Amendment.) Rather, like Tushnet, Chen, and Blocher, my goal is to get us asking the question. I also hope to show that an autonomy-based rationale for coverage does not necessarily lead to the conclusion that every act should be constitutionally covered so long as an argument exists that it was intended to be expressive.

Conclusion

Law professors are granted the gift of being able to give sustained and intense study to all manner of judicial pronouncements, even the most tossed-off ones. It is therefore fairly easy scholarly work to take apart legal assertions that turn out to be wrong. In Free Speech Beyond Words, Tushnet, Chen, and Blocher take on a much more challenging assignment: to provide support for a legal assertion that is in fact correct. They put to the side the questions of content neutrality, overbreadth, and governmental motive which occupy so much of the current collective thinking about the First Amendment, and focus their efforts on a more fundamental question that, despite its importance, has so far gone largely ignored: why the Amendment’s “freedom of speech” applies to a distorted guitar chord, spilled paint on canvas, and Lear’s limericks.[114]

In the article that provides the framework for the authors’ project, Frederick Schauer writes that “[t]he history of the First Amendment is the history of its boundaries.”[115] Free Speech Beyond Words will stand as an essential signpost in helping academics and judges alike to understand where, and just as importantly how, those boundaries should be set.

 

            [1].    William Nelson Cromwell Professor of Law, Harvard Law School.

            [2].     William M. Beaney Memorial Research Chair Professor, University of Denver Sturm College of Law.

            [3].     Professor of Law, Duke Law School.

            [4].     Associate Dean for Academic Affairs & Associate Professor, Elon University School of Law; Affiliate Fellow, Yale Law School Information Society Project. Thanks to Vincent Blasi, Joseph Blocher, Alan Chen, Heidi Kitrosser, Kate Klonick, Helen Norton, David Pozen, Scott Skinner-Thompson, Alexander Tsesis, Mark Tushnet, Morgan Weiland, and to colleagues at the 2017 Loyola University Chicago Constitutional Law Colloquium and Yale Law School Information Society Project Freedom of Expression Scholars Conference for comments and suggestions. Thanks as well to Britney Boles and Helen Tsiolkas for research assistance.

            [5].    Hugo Ball, Dada Manifesto (1916), https://en.wikisource.org/wiki/‌Dada_‌Manifesto_ ‌(1916,_Hugo_Ball) (last updated June 11, 2017).

            [6].    See Kyle Gann, No Such Thing as Silence: John Cage’s 4’33’ 2–3 (2010); Will Hermes, The Story of 4’33”, NPR (May 8, 2000, 12:00 AM), http://www.npr.org‌/2000/05/08/‌1073885/4-33.

            [7].     Latin for “is silent,” tacet is a musical scoring term meant to indicate that an instrument or voice does not sound.

            [8].     Gann, supra note 6, at 178–80.

            [9].     John Cage: An Autobiographical Statement, John Cage (Apr. 17, 1990), http://www.johncage.org/autobiographical_statement.html; Gann, supra note 6, at 111.

         [10].     Morgan Falconer, Painting Beyond Pollock 168 (2015).

         [11].      Id.

         [12].     Francesca Wilmott, Composing Silence: John Cage and Black Mountain College, Inside/Out Blog (Jan. 3, 2014), https://www.moma.org/‌explore/inside_out/‌2014/01/03/‌composin‌g-silence-john-cage-and-black-mountain-college-3/.

         [13].    Brice Marden, Statement Accompanying Podcast, Robert Rauschenberg: Among Friends, MoMa, https://www.moma.org/audio/playlist/40/639.

         [14].     As critic Douglas Kahn writes, Cage “extended the decorum of silencing by extending the silence imposed on the audience to the performer.” Douglas Kahn, John Cage: Silence and Silencing, 81 Musical Q. 556, 560 (1997). See also Gann, supra note 6, at 19 (4’33” “called upon the audience to remain obediently silent under unusual conditions”).

         [15].     Gann, supra note 6, at 4. See Robert Kostelanetz, Conversing with Cage 70 (2003).

         [16].     James Pritchett, What Silence Taught John Cage: The Story of 4’33”, Rose White Music: The piano in my life (2009), http://rosewhitemusic.com/piano/writings/silence-taught-john-cage/.

         [17].     Michael Nyman, Experimental Music: Cage and Beyond 26 (2d ed. 1999).

         [18].    Kostelanetz, supra note 15, at 70.

         [19].     According to linguistic philosopher J. L. Austin, “A locutionary act has to do with the simple act of a speaker saying something, i.e. the act of producing a meaningful linguistic expression.” Yan Huang, The Oxford Dictionary of Pragmatics 176 (2012).

         [20].     Quoting Hurley v. Irish-Am. Gay, Lesbian & Bisexual Grp. of Bos., Inc., 515 U.S. 557, 568–69 (1995).

         [21].    As Schauer and the authors note, the First Amendment’s boundaries—what it protects—are “far more consequential” than the doctrinal work that comes after an affirmative answer to the coverage question—i.e., whether the speech-affecting regulation is content-based or content-neutral and the like. (p. 2, citing Frederick Schauer, The Boundaries of the First Amendment: A Preliminary Exploration of Constitutional Salience, 117 Harv. L Rev. 1765, 1767 (2004)).

         [22].    See supra note 7. To be sure and as discussed infra, some historical or theoretical approaches are occasionally used to decide coverage questions in First Amendment cases. The test from Spence v. Washington is one such approach. But the Court’s applications of these tools to coverage questions is so intermittent and inconsistent that they may as well say almost nothing at all.

         [23].    U.S. v. Stevens, 559 U.S. 460 (2010).

         [24].     Sorrell v. IMS Health, 564 U.S. 552 (2011).

         [25].     Brown v. Entertainment Merchants’ Association, 564 U.S. 786 (2011).

         [26].     See also Schauer, supra note 21, at 1767 (“[Q]uestions about the involvement of the First Amendment in the first instance are often far more consequential than are the issues surrounding the strength of protection that the First Amendment affords the speech to which it applies. Once the First Amendment shows up, much of the game is over.”).

         [27].    Perhaps this Review’s length proves the point.

         [28].    In Brown, California, which was defending its statute prohibiting the sale of violent video games to minors against First Amendment challenge, acknowledged that “video games qualify for First Amendment protection.” Brown v. Entm’t Merchs. Assoc., 564 U.S. 786, 790 (2011).

         [29].     Justice Alito, joined in concurring in the judgment by Justice Scalia, would have invalidated the California statute on facial vagueness grounds and thus would not have reached the First Amendment coverage and protection questions. Id. at 807 (Alito, J., concurring in judgment).

         [30].     Justice Thomas’ dissent in Brown argued that the original public meaning of the First Amendment could not have understood the “freedom of speech” to include a right to speak to children without their parents’ consent, and thus the Amendment did not cover the right of video game manufacturers to sell their games to children. Id. at 821 (Thomas, J., dissenting).

         [31].    See, e.g., Morse v. Frederick, 551 U.S. 393, 410–11 (2007) (Thomas, J., concurring) (finding that “the First Amendment, as originally understood, does not protect student speech in public schools.”).

         [32].    Though the Court’s foundational First Amendment coverage cases were written at a time when opinions were much less verbose, they similarly spend very little time developing a coverage doctrine that can be applied in subsequent cases. See Chaplinsky v. New Hampshire, 315 U.S. 568, 571-72 (1942) (one paragraph); Spence v. Washington, 418 U.S. 405, 409-11 (1974) (four paragraphs).

         [33].     See Ashwander v. Tenn. Valley Auth., 297 U.S. 288, 347 (1936) (Brandeis, J., concurring); Delaware v. Van Arsdall, 475 U.S. 673, 693 (Stevens, J., dissenting) (labeling the Brandeis opinion in Ashwander “one of the most respected opinions ever written by a Member of this Court”).

         [34].    A. E. Dick Howard, Out of Infancy: The Roberts Court at Seven, 98 Va. L. Rev. In Brief 76, 84 (2012) (discussing the Court’s avoidance of the First Amendment question in favor of a more straightforward vagueness question in Fox Television Stations v. FCC, 567 U.S. 239 (2011)).

         [35].     See, e.g., Bowers v. Hardwick, 478 U.S. 186, 191 (1986) (the Fourteenth Amendment does not cover the “claimed constitutional right of homosexuals to engage in the act of sodomy”), overruled by Lawrence v. Texas, 539 U.S. 558 (2003). The asymmetry I discuss above is doubly strange given that the tiers-of-scrutiny analysis in the substantive due process area largely mirrors the analysis used in Speech Clause cases.

         [36].    District of Columbia v. Heller, 554 U.S. 570 (2008); see also Joseph Blocher and Darrell A.H. Miller, What Is Gun Control? Direct Burdens, Incidental Burdens, and the Boundaries of the Second Amendment, 83 U. Chi. L. Rev. 295, 296 (2016) (quoting Schauer, supra note 21, in noting that per Heller, in cases involving bans on gun possession by felons and the mentally ill, the Second Amendment “just does not show up [at all]”).

         [37].    Spence, 418 U.S. at 410–11.

         [38].    See Jason Heller, A Purplish Haze: The Science Fiction Vision of Jimi Hendrix, noisey (Mar. 14, 2017), https://noisey.vice.com/en_us/article/wnkgen/a-purplish-haze-the-science-fiction-vision-of-jimi-hendrix. The novel refers to a “purplish haze” caused by a “huge moon, hanging dim and violet and malevolent above the horizon.” Philip José Farmer, Night of Light 34 (1966).

         [39].    Chen points out that for many years the Catholic Church banned the musical interval         known as the tritone, which uses a diminished fifth, on the ground “its dissonant sound evoked evil,” and that the Church even labeled it the “Devil’s Interval” (p. 27). He does not note, however, that Hendrix used the Devil’s Interval in the opening to Purple Haze. Keith Shadwick, Jimi Hendrix: Musician 96 (2003).

         [40].    Alexander Meiklejohn, Free Speech and Its Relation to Self-Government 888–89 (1948); Robert Post, Participatory Democracy and Free Speech, 97 Va. L. Rev. 477, 483 (2011).

         [41].    See Nicole Lyn Pesce, Prince’s secret vault of unreleased music could produce albums for another 100 years, NY Daily News (Apr. 22, 2016), http://www.nydailynews.com/entertainment/music/prince-secret-music-vault-produce-albums-100-years-article-1.2611146. The lyrics of Darling Nikki, from Prince’s 1984 multiplatinum, Grammy-winning album Purple Rain, were Tipper Gore’s primary inspiration for founding the Parents Music Resource Center, whom we have to thank for Parental Advisory stickers on the covers of plastic discs that used to contain digitized music called “CDs.” See Censor This: Music Censorship in America, Internet Archive: Wayback Machine https://web.archive.org/web/20030406085225‌/http://www.ge‌o‌cities.‌com‌‌/‌‌fireace_00/pmrc.html (last visited Sep. 1, 2017).

         [42].    Miller v. Civil City of S. Bend, 904 F.2d 1081, 1125 (7th Cir. 1990) (Easterbrook, J., dissenting).

         [43].    This quote is taken from the poet Harold Rosenberg’s essay in ArtNews concerning the significance of the Abstract Expressionist movement in American art. Harold Rosenberg, The American Action Painters, ArtNews (Dec. 1952), http://www.artnews.com/2007/11/01/top-ten-artnews-stories-not-a-picture-but-an-event/.

         [44].    See Amos Barshad, ‘Yankees Suck! Yankees Suck!’, Grantland (Sept. 1, 2015), http://grantland.com/features/yankees-suck-t-shirts-boston-red-sox/.

         [45].    Since this argument is also a critique of the intent-based approach to First Amendment coverage that I propose in Part II of this Essay, I will return to it in Part III.

         [46].    Miller v. California, 413 U.S. 15 (1973).

         [47].    See also Sonya G. Bonneau, Ex Post Modernism: How the First Amendment Framed Nonrepresentational Art, 39 Colum. J. L. & Arts 195, 196 (2015) (“[V]isual art’s main doctrinal residence stood in obscenity law’s backyard, as a vague definitional tautology: art constitutes speech so long as it is not obscene, and speech is not obscene if it’s art.”).

         [48].    See, e.g., Thomas Girst, (Ab)using Marcel Duchamp: The Concept of the Readymade in Post-War and Contemporary American Art, 2 Marcel Duchamp Stud. Online J. 5 (2003), http://www.toutfait.com/issues/volume2/issue_5/articles
/girst2/girst1.html (Marcel Duchamp stating his ready-mades “sought to discourage aesthetics”); Jerry Salz, Idol Thoughts, Village Voice (Feb. 21, 2006), https://www.villagevoice.com/2006/02/21/idol-thoughts/ (Duchamp’s work “provide[s] a way around inflexible either-or aesthetic propositions”).

         [49].    Jennie Gottschalk, Experimental Music Since 1970 1 (2016).

         [50].    Id. at 195.

         [51].    At least one court of appeals has come to the same conclusion with respect to Hurley. See Holloman ex. rel. Holloman v. Harland, 370 F.3d 1252, 1270 (11th Cir. 2004) (the test for expressive conduct post-Hurley asks whether a reasonable person would interpret conduct as expressing “some sort of message, not whether an observer would necessarily infer a specific message.” (emphasis in original)).

         [52].    Calvin Tompkins, Duchamp: A Biography 186 (1998).

         [53].    Philip Hensher, The Loo That Shook the World: Duchamp, Man Ray, Picabia, The Independent (Feb. 20, 2008), http://www.independent.co.uk/arts-entertainment/art‌/features/the-loo-that-shook-the-world-duchamp-man-ray-picabi-784384.html.

         [54].    Obviously, this problem could also run in the other direction. See Richard A. Posner, Art for Law’s Sake, 58 Am. Scholar 513, 514 (1989) (“[A] work highly valued in its time, or for that matter in later times, may eventually come to seem thoroughly meretricious. Artistic value is something an audience invests a work with, and as the tastes of audiences change, so do judgments of artistic value.”).

         [55].    Abby Goodnough, Giuliani Threatens to Evict Museum Over Art Exhibit, N.Y. Times (Sept. 24, 1999), https://partners.nytimes.com/library/arts/092499brooklyn-museum.html?mcubz=1.

         [56].    Id. See Paul Lieberman and Diane Haithman, N.Y. Art Show Gets Scathing Giuliani Review: “Sick Stuff,” L.A. Times (Sept. 24, 1999), http://articles.latimes.com/‌1999/sep/24/news/mn-13513.

         [57].    Jonathan Harris, Federal Art and National Culture: The Politics of Identity in New Deal America 25 (1995); see also Francis V. O’Connor, Federal Support for the Visual Arts: The New Deal and Now 20 (1969) (PWAP standards evinced a “suspicion of anything experimental or controversial”); Jerry Adler, 1934: The Art of the New Deal, Smithsonian Mag. (June 2009), http://www.‌smithsonianmag.com/‌arts-culture/1934-the-art-of-the-new-deal-132242698/ (describing “conservative” nature of PWAP-funded projects).

         [58].    Bonneau, supra note 47, at 203–04 (citing William Hauptman, The Suppression of art in the McCarthy Decade, ArtForum,Oct. 1973, at 48).

         [59].    95 Cong. Rec. 11, 584–85 (1949) (Rep. Dondero delivering a speech entitled Modern Art Shackled to Communism).

         [60].    See Bonneau, supra note 47, at 202 (citing Clement Greenberg, Avant-Garde and Kitsch (1939), reprinted in Clement Greenberg, Art and Culture: Critical Essays 3, 19 (1961)).

         [61].    Lewis Carroll, Through the Looking-Glass, and What Alice Found There 21 (1872).

         [62].     Burroughs described the cut-up method as follows:

Take a page. Like this page. Now cut down the middle and cross the middle. You have four sections: 1 2 3 4. Now rearrange the sections placing section four with section one and section two with section three. And you have a new page. Sometimes it says much the same thing. Sometimes something quite different . . . .

William Burroughs, The Cut Up Method, in The Moderns: An Anthology of New Writing in America 345, 345–48 (LeRoi Jones, ed. 1963).

         [63].    Richard Elliott, The Sound of Nonsense 8 (2018).

         [64].    Flarf poetry’s primary aesthetic is, in the words of its practitioners, “deliberate shapelessness of content, form, spelling, and thought in general, with liberal borrowing from internet chat-room drivel and spam scripts . . . .” The Flarf Files, http://writing.‌upenn.‌edu/‌epc/authors/bernstein/syllabi/readings/flarf.html (last visited June 12, 2018).

         [65].    Wilco, I Am Trying to Break Your Heart, on Yankee Hotel Foxtrot (self-released 2001; Nonesuch Records 2002). One might argue that the expressive content of lyrics like these, like those of Jabberwocky, is rhythmic rather than literal meaning.

         [66].    Oliver Wendell Holmes, Jr., Law in Science and Science in Law, 12 Harv. L. Rev. 443, 460 (1899); see also Letter from Oliver Wendell Holmes to Harold J. Laski (May 9, 1925), in 1 Holmes-Laski Letters: The Correspondence of Mr. Justice Holmes and Harold J. Laski, 1916–1935, 737, 738 (Mark DeWolfe Howe ed., Harv. U. Press 1953) (noting how difficult it is to “think accurately—and think things not words”).

         [67].    Ludwig Wittgenstein, Tractatus Logico-Philosophicus §5.6, at 149 (C. K. Ogden ed. & trans., Harcourt Brace 1922) (emphasis omitted); see id. §3.032, at 43, 45 (“To present in language anything which ‘contradicts logic’ is as impossible as in geometry to present by its co-ordinates a figure which contradicts the laws of space; or to give the co-ordinates of a point which does not exist.”).

         [68].     The rules of this language game certainly depend on the age of the daughter (mine is 8), as well as other contexts that Wittgenstein would deem as relevant to rule-setting.

         [69].    John Greenman, On Communication, 106 Mich. L. Rev. 1337, 1347–48 (2008).

         [70].    Spence v. Washington, 418 U.S. 405, 410–11 (1974); see R. George Wright, What Counts as “Speech” in the First Place? Determining the Scope of the Free Speech Clause, 37 Pepp. L. Rev. 1217, 1238 (2010) (“In the absence of the speaker’s intent to promote some more or less determinate understanding, we may be skeptical that speech in the constitutional sense is present.”).

         [71].    John Stuart Mill, On Liberty 7 (1859).

         [72].    Ball, supra note 5.

         [73].     See, e.g., Obergefell v. Hodges, 135 S. Ct. 2584, 2597 (2015) (“[T]hese liberties extend to certain personal choices central to individual dignity and autonomy, including intimate choices that define personal identity and beliefs.”); Bowers 478 U.S. at 191 (explaining that previous cases “were interpreted as construing the Due Process clause of the Fourteenth Amendment to confer a fundamental individual right to decide whether or not to beget or bear a child” (emphasis added)); Loving v. Virginia, 388 U.S. 1, 12 (1967) (“[T]he freedom to marry, or not marry, a person of another race resides with the individual and cannot be infringed by the State.” (emphasis added)).

         [74].    See W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 632 (1943).

         [75].     Wooley v. Maynard, 430 U.S. 705, 714 (1977) (quoting Barnette, 319 U.S. at 637).

         [76].     See Hurley v. Irish-Am. Gay, Lesbian & Bisexual Grp. of Bos., 515 U.S. 557, 574 (1995).

         [77].    Louis Michael Seidman, Silence and Freedom 9 (2007).

         [78].    Bd. of Regents of the Univ. of Wis. Sys. v. Southworth, 529 U.S. 217, 239 (2000) (Souter, J., concurring in judgment) (describing rights at issue in Hurley, Wooley, and Barnette respectively).

         [79].    Cf. Keller v. State Bar of Cal., 496 U.S. 1, 16 (1990) (forced contribution of a portion of dues to political candidates violates dues-payers’ rights to freedom of association); Abood v. Detroit Bd. of Educ., 431 U.S. 209, 234 (1977).

         [80].    Slacker (Detour Filmproduction 1991).

         [81].     Martin H. Redish, Freedom of Expression, Political Fraud, and the Dilemma of Anonymity, in Speech and Silence in American Law 143 (Austin Sarat ed., 2010); see also id. at 149 (“[I]n an important sense, anonymity is a subcategory of a right not to speak: the choice not to reveal one’s name is a narrower exercise of a broader right to choose to not speak at all, because it represents the decision of the speaker to selectively limit his expression.”).

         [82].    Paul Horwitz, Comment on Chapter 4: Anonymity, Signaling, and Silence as Speech, in Speech and Silence in American Law 181 (Austin Sarat ed., 2010).

         [83].     See David Levine, Screwing the World (1984), Illustration Chronicles, http://illustrationchronicles.com/Screwing-the-World-with-David-Levine (last visited June 13, 2018).

         [84].     Lisa Kern Griffin, Silence, Confessions, and the New Accuracy Imperative, 65 Duke L.J. 697, 704–05 (2016).

         [85].    Id. at 698, 704–05 (second quote quoting Austin Sarat, Introduction: Situating Speech and Silence, in Speech and Silence in American Law 1, 3 (Austin Sarat ed., 2010) (quotation marks omitted)).

         [86].    Seidman, supra note 77, at 2.

         [87].     Id. at 169.

         [88].    Texas v. Johnson, 491 U.S. 397, 400 (1989).

         [89].     Id. at 405.

         [90].     Id. at 406.

         [91].     City of Dallas v. Stanglin, 490 U.S. 19, 25 (1989).

         [92].    Mark Tushnet, Introduction: Reflections on the First Amendment and the Information Economy, 127 Harv. L. Rev. 2234, 2237 (2014).

         [93].     U.S. v. Virginia, 518 U.S. 515, 558 (1996) (accusing the Court of “load[ing] the dice” in constitutional cases in order to reach its desired result).

         [94].    Gregory P. Magarian, Managed Speech: The Roberts Court’s First Amendment 157–58 (2017); James Bradley Thayer, The Origin and Scope of the American Doctrine of Constitutional Law, 7 Harv. L. Rev. 129 (1893). Tushnet here is noting the absence of Thayerian approaches in First Amendment scholarship that pervade other areas of constitutional law—namely, the idea that judges should only strike down a law when its unconstitutionality “is so clear that it is not open to rational question.” Tushnet, supra note 92, at 2244 (quoting Thayer, supra, at 144).

         [95].    Jed Rubenfeld, The First Amendment’s Purpose, 53 Stan. L. Rev. 767, 769 (2001).

         [96].    Other scholars have leveled similar critiques at autonomy-based rationales. See, e.g., Frederick Schauer, The Second-Best First Amendment, 31 Wm. & Mary L. Rev. 1, 7 (1989).

         [97].    See Chevron v. Nat. Res. Def. Council, 467 U.S. 837 (1984). Recent scholarly work has sought to shed more light on the coverage question. See, e.g., Morgan Weiland, Expanding the Periphery and Threatening the Core: The Ascendant Libertarian Speech Tradition, 9 Stan. L. Rev. 1389, 1457–58 (2017) (arguing that coverage questions should focus on whether the speaker autonomy sought to be protected is “thick,” meaning individually held and consonant with other First Amendment interests, such as self-governance or self-realization, or “thin,” meaning assertion of a bare speech right by corporate speakers or other non-natural legal persons).

         [98].    See supra note 73 and accompanying text.

         [99].     Skinner v. Oklahoma, 316 U.S. 535 (1942). The question may not be as simple as this purely symmetrical approach would conclude. See I. Glenn Cohen, The Constitution and the Rights Not to Procreate, 60 Stan. L. Rev. 1135, 1140–42 (2008) (proposing a three-part framework of legal, gestational, and genetic parenthood and arguing the “right to not procreate” would operate differently depending on which of those three contexts applied). For present purposes, I refer to the most general hypothetical form of forced procreation: a law mandating that parents have children, or what Cohen calls gestational parenthood, which proscribes the decision, to use the Court’s words in Eisenstadt v. Baird, “whether or not to bear or beget a child.” Eisenstadt v. Baird, 405 U.S. 438, 453 (1972).

      [100].    See supra notes 84-87 and accompanying text.

      [101].    Seana Shiffrin, A Thinker-Based Approach to Freedom of Speech, 27 Const. Comment. 283, 287 (2011).

      [102].    Id.

      [103].    Joseph Blocher, Rights To and Not To, 100 Calif. L. Rev. 761, 765 (2012).

      [104].    Id. at 766–67.

      [105].    John H. Garvey, What Are Freedoms For? 17 (1996).

      [106].    Joseph Blocher, The Right to Not Keep or Bear Arms, 64 Stan. L. Rev. 1, 54 n.15 (2012).

      [107].    See Schiffrin, supra note 101, at 283 (“[w]e should understand freedom of speech as, centrally, protecting freedom of thought”); Martin H. Redish, The Value of Free Speech, 130 U. Pa. L. Rev. 591, 593 (1982) (free speech serves the interest of “individual self-realization”).

      [108].    For a related overinclusivity-based critique of this aspect of Spence, see Robert C. Post, Recuperating First Amendment Doctrine, 47 Stan. L. Rev. 1249, 1251–53 (1994).

      [109].    Edwin Baker, Human Liberty and Freedom of Speech 52 (1989).

      [110].    See Kent Greenawalt, Speech, Crime, and the Uses of Language 742–56 (1989).

      [111].    See Eugene Volokh, Crime-Facilitating Speech, 57 Stan. L. Rev. 1095, 1145–73 (2005).

      [112].    See generally Helen Norton, Truth and Lies in the Workplace, 101 Minn. L. Rev. 31 (2016).

      [113].    See, e.g., Hill v. Jones, 725 P.2d 1115, 1118–19 (Ariz. Ct. App. 1986) (citing Restatement of Contracts § 161 (Am. Law Inst. 1981), with respect to when a duty to disclose, or as the court calls it “a duty to speak,” arises).

    [114].    See Edward Lear, A Book of Nonsense (1846).

      [115].    Schauer, supra note 21, at 1765.