I

To the extent that the Supreme Court has a theory of speech, it is the following:

1. For “the expression of an idea” to count as speech, it must be “sufficiently imbued with elements of communication.”
2. “An intent to convey a particularized message [must be] present.”
3. “[I]n the surrounding circumstances the likelihood [must be] great that the message would be understood by those who viewed it.”⁵

   Spence v. Washington, 418 U.S. 405, 409–11 (1974).

Brison dismisses this theory outright, saying that it “does not succeed in distinguishing speech from nonspeech conduct” and that “it has proven to be notoriously difficult—some would say impossible—for the Court to distinguish speech from other forms of intentional action.”⁶

But I would not give up so easily. In fact, I shall argue that this account of speech is not just a viable notion for law but also a good one for ordinary understandings of speech and therefore for philosophy too.

Consider the following examples.

For behavior to count as a human action, as opposed to a mere event or happening in the physical world, there is a standard philosophical requirement that there be at least one true description of it as intentional, as intended by an agent.⁷ “If, for instance, Mary is jumping on the sofa and she breaks it, even if she did not break it intentionally, breaking the sofa was her action, because it was intentional under the description ‘jumping on the sofa.’”⁸ In Example 1 there is no consciousness, no intention, and so no conduct in the sense of intentional action. This cough, Brison agrees, “is not an action, since it is not intentional under any description. It is like snoring or anything else one ‘does’ while asleep or otherwise unconscious.”⁹

(This example is designed to show that intentional action is not necessarily communicative.) This time I coughed on purpose, as everyone normally does when they cough (without, perhaps, thinking about it quite so much). Even assuming this cough sounds exactly like the previous one, here I was not, strictly speaking, forced (as in Example 1) to cough; I could have avoided coughing by, for example, holding my breath or bolting from the compartment for fresh air. It is true that I meant (intended) to cough, but normally “a cough is just a cough,” and I “meant” nothing else by it. In fact, if anything I tried to avoid drawing any particular attention to my plight. So here there is intention and conduct but not (I shall argue) communication or speech.

To return to the distinction between actions and events: Here, coughing is an action and not simply (as in Example 1) an event, since it is intentional under the description “coughing on purpose to clear my throat.” (Of course, it could also be correctly described as [unintentionally] “contracting certain throat muscles and setting off certain vibrations in the air.”) But is it an expressive action?

Brison seems to think so: “In deciding not to suppress the cough or leave the room, Y acted in a way that he knew would communicate something to X.”¹⁰

But think what that would mean…. Every time someone coughed they would be communicating or “conveying a message,” if that was “at least a foreseeable consequence of coughing.”¹¹ In fact, in the example above, not coughing would also convey a message—probably even a more obvious one—since I described the alternatives to coughing in pretty dramatic terms (“holding my breath or bolting from the compartment for fresh air”). So, cough or don't cough: either way you can't avoid communicating. Indeed, whatever you do you will be communicating, since this analysis could be extended indefinitely.¹² On Brison's view, it seems that we are pretty much condemned to communicate all the time whatever we do, so long as someone, somewhere might think it meant something, however unintended. (This issue is important because—as discussed in the final section of this essay—the distinction between laws related to speech and laws unrelated to speech depends on the prior distinction between communicative intentional action and noncommunicative intentional action. If “all conduct is expressive,” then neither distinction can be made.)

Brison supplements this analysis with her own example of yawning “when one has decided not to resist the urge to yawn during a talk that has gone on for too long, knowing (but not intending) that the yawn will signal … to the speaker that the talk has gone on too long, whether or not the yawner intends this to be the message communicated.”¹³

Two issues should be distinguished: (a) Was communication intended? (Here Brison wants to say “No.”) (b) Did the actor nonetheless communicate? (And here Brison wants to say “Yes.”) For the reasons developed in this essay, I shall argue that when the first answer is “No” the second answer should be “No” too. But in this particular example I do not get beyond the first answer, for three reasons.

(1) Yawning is not particularly difficult to suppress or conceal (unlike the coughing described above). This fact is well known and makes the unsuppressed or unconcealed yawn all the more significant. (2) Yawning out of boredom has a long, conventional history, well known the world over, and celebrated in stylized depictions since time immemorial in literature, drama, and film. The Oxford English Dictionary even defines yawn in one sense as “denoting something that induces boredom,” and gives as an example: “[They] were very slow in taking the hint conveyed by his violent yawnings, that he was anxious to retire to rest.”¹⁴

(3) Thus “knowing (but not intending) that a yawn will communicate something” is implausible (at least in Brison's scenario); it violates the normal presumption that everyone intends the ordinary consequences of their voluntary actions.¹⁵

(This example is designed to show that intentional action—even to influence others' behavior—is not necessarily communicative.) One might say that here I am trying to “plant an idea” in the smoker's mind without wanting the smoker to realize I am doing so. I am hoping the smoker will “get the message,” and in that sense I am “sending a message”—but without intending that there be any connection between the two. In fact, I hope the connection will not be evident. If I am successful, the smoker will get the message without realizing I was sending a message.

Normally, however, when we speak or communicate, we do not simply try to plant ideas in people's minds. Or at least, even when that is partly our intention, another part is to “announce our intention” of doing so. Otherwise we would simply be trying to induce changes in behavior as one might with subliminal advertising or in training an animal.

Brison disagrees again; she thinks this example is “analogous to someone's burning a cross on (or right next to) a black family's lawn in an attempt to induce the family to move away” (which, as she correctly notes, “communicates a message of terror”).¹⁷

But, again, that is a poor analogy. Everyone (from the Supreme Court on down) knows that burning a cross is communicative. Practically anything one does to or with a cross means something, because the cross is a preexisting symbol with a lengthy religious history. Cross burning itself has a long history and acquired its current meaning of racial hatred more than a century ago: The point of my example was to test the thesis that “influencing people's behavior is not necessarily communicative,” not to show that communication can influence behavior (which is easy and which Brison easily shows). Suppose the laws in my jurisdiction make it a crime to “communicate a message of terror” (to quote Brison).¹⁹ As I am puttering around in my yard late one day I decide to gather up and burn various branches and yard trash that have accumulated there. After I have piled everything into a big heap, some of the branches happen to fall together so as to form a large cross. Thinking nothing in particular about that (other than amusement at the coincidence) I proceed to ignite the pile of debris just as I normally would. Unbeknownst to me, from the perspective of a black family's home not far away, the burning branches show up as a huge flaming cross dramatically framed against the darkening sky. I suppose that “conveying a message … was at least a foreseeable consequence” of my actions, so on these facts, according to Brison, I have committed the crime of “communicating a message of terror.”²⁰ (This would be like my Example 2.)

After serving my prison sentence I return home, determined to seek revenge against my neighbors but equally determined never again to run afoul of the law against “communicating a message of terror.” So I secretly release a tiger near the neighbors' house, just to terrorize them. That should certainly succeed in terrorizing them, but the idea that someone did this on purpose, just to “send them a message,” would not be high on anyone's list of likely explanations. (A much more obvious thought would be that the tiger escaped from a visiting circus and just happened to be in the neighborhood, etc.) Yet according to Brison, “[r]egardless of [my] intentions,” the release of the tiger “nonetheless communicates a message of terror” to my neighbors.²¹

“In spite of the fact that” I do not want them to know that I am “sending a message of terror,” I have “terrorize[d]” them, which is presumably enough for Brison to say that I have once again committed the crime of “communicating a message of terror.”²² (This would be like my Example 3.)

Now there can be no doubt that I am not “just coughing” but am also, and primarily, “expressing my displeasure.” Here there is communication and perhaps even speech, at least in the “symbolic” sense. And here I am not just “sending a message” but am also drawing attention to the fact that I am sending a message.

Now the smoker cannot hope to ignore me; or, at least, ignoring me will mean not just ignoring my cough. It will also mean ignoring the evident fact that I am now intentionally sending a message and expressing my displeasure by coughing.

Brison remains unconvinced. She agrees that this final example is communicative, but she thought the two previous examples were communicative too, so nothing changed with these new facts. The phenomenon of “miscommunication” is supposed to reveal an especially telling objection: “If … communicating anything that one does not intend to communicate [were] impossible … the everyday phenomenon of miscommunication [would be] impossible.”²³

I am not sure what, if anything, turns on the existence of the word miscommunication (or the fact that people say things like “I must have miscommunicated”). I am happy to adopt as my own position the view that “Every miscommunication is itself a communication.” The miscommunication does not hit its intended target, so to speak; but it had a target. It was simply so poorly and artlessly constructed that someone got not the intended message but a confused or muddled or garbled variant of it—along with the fact (and here there was no mistake) that some communication was intended. (The person who says “I must have miscommunicated” does not go on to explain: “I wasn't trying to communicate at all.”) So it will not in any way detract from my analysis to consider miscommunication a form or variant of communication. The real puzzle is not how people get the right message but how they know a message is being communicated at all. (Even if this does not seem puzzling for those involved, explaining it is a genuine puzzle.)

II

It is common ground that there can be communication of unintended meaning. What is subject to debate is whether there can be unintentional communication. Consider the following pairs of sentences:

1. Someone scored a goal. The ball passed through the goalposts.
2. Someone greeted someone. A greeting took place.²⁴

   This example is suggested by MOYA, supra note 8, at 38–39, 49–53.

In the description “Someone scored a goal,” the ball's passing through the goalposts is being viewed as intentional, as intended by an agent. But for someone to score a goal, the alternative description “The ball passed through the goalposts” must also be true, and here the ball's passing through the goalposts is being viewed without regard to intention, as a mere event or happening in the physical world. This event could happen in any number of ways (the ball could be lifted up by a bird or perhaps a gust of wind). An ordinary event (a ball's passing through the goalposts) is an action if there is at least one true description of it as intentional (scoring a goal). And in general, the actions a person performs intentionally can receive other true descriptions under which they are unintentional.

But now consider the description “Someone greeted someone.” Here there seems to be no corresponding event in the physical world that would have to take place in order for “Someone greeted someone” to be true. To add “A greeting took place” sounds redundant or tautologous in a way that “The ball passed through the goalposts” does not. But why? I think it is because with “A greeting took place,” no real alternative is suggested; we are still not viewing the greeting without regard to intention, as a mere event or happening in the physical world. Instead, we are importing intention into the very concept of a greeting—mainly because it makes no sense to view a greeting “without regard to intention, as a mere event or happening in the physical world.” (Would you consider yourself to be “greeted” by a machine or even a trained animal?)

In certain sports and under certain circumstances (e.g., during an actual game and according to the rules), getting the ball to pass through the goalposts means you have scored a goal. But suppose that during a game you kick the ball, and somehow it passes through the other team's goalposts. That still means a goal was scored (and it was still your action since it was intentional under the description “kicking the ball”), but presumably you did not intend to score a goal for the other team. Could something similar happen in the case of a greeting?

Implied in the idea of a greeting is the intention of remaining on good terms with one's acquaintances—let's call it the “general commitment to friendliness.” (This is what the machine and the trained animal lacked; and, indeed, something would be odd about your behavior if you greeted people in the usual ways but then proceeded to be as unfriendly to them as to total strangers.) Since no specific event is required for a greeting, if some event occurs anyway in an unexpected or unintended way, that cannot constitute a “greeting” without the implied intention of friendliness.²⁵

In this sense, a greeting as such is not susceptible of being described as unintentional. There is thus no such thing as an accidental (unintentional) greeting, though someone may of course mistakenly think she has been greeted when no one greeted anyone.²⁶

Consider the difference between offending (which is not necessarily intentional and thus, I shall argue, not necessarily communicative) and insulting (which is).²⁷

The distinction is evident at the terminological level from one of the Oxford English Dictionary's examples for offend: “Wander they in their pleasing darknesse, offended if you shew them light,” and from its definition of insult: “To assail with offensively dishonouring or contemptuous speech or action.”²⁸ The classic bland apology that accepts no blame usually starts out: “I sincerely apologize if I have offended anyone. That was certainly not my intention.” This formulation is preferred by politicians and professional athletes who have made some egregious faux pas or been caught in especially sordid straits. The idea that they would have wanted to communicate anything is generally the farthest thing from their minds. Deepest secrecy is much more to the point.²⁹

Suppose I secretly torture cows in my darkened basement, just for fun. Adherents of the Hindu religion, who consider cows sacred, would no doubt take great offense at this if they were to find out, but I know nothing of that; I am simply pursuing my private hobby. Suppose, however, that secretly filmed videotapes of my cow-torturing rituals arrive (unbeknownst to me) on the shores of India, where the inhabitants proclaim themselves “deeply offended,” as indeed they are. But have I communicated anything to anyone?

Irritated by all this controversy and the ominous threats coming my way, I make a special point of inviting to my next cow-torturing event the camera crews of the Indian News Network, with a direct live feed to New Delhi. Now I have truly “added insult to injury,” as the saying goes.

For another example, suppose I have (finally) been invited to Madam Rothschild's annual garden party, the social event of the year. Obviously I need a new outfit for the occasion, so I pick out a handsome wool sweater with an abstract design featuring black spots of various sizes on a white background, intermingled with striking red blotches. Unbeknownst to me, Madam Rothschild's beloved Dalmatian Trixie has recently met a horrible fate, crushed by a huge tractor-trailer right in front of Madam Rothschild herself, who was understandably devastated. “The sight of poor Trixie lying there, all limp and bloody,” she moaned to all who would listen, “will be with me to my dying day.” She forthwith decreed that all images of Trixie should forever be banned from her mansion and grounds.

Imagine my surprise, then, when I make my grand entrance in my striking new sweater. Madam Rothschild, deeply offended, turns bright crimson and practically swoons on the spot, but not before having me forcibly ejected by her security forces. Stung by the injustice of it all, I bide my time, and a year later I get my chance for revenge when Madam Rothschild inadvertently invites me to her next garden party. Now, I already have the perfect sweater to wear this time, for the perfect insult….

An insult is communicative and (assuming it could be the subject of a legal case) counts as speech for purposes of the First Amendment. Wearing a certain sweater is not speech in any usual sense; but if it is intended to communicate, it may, under the right circumstances (as in the above example), constitute an insult and count as speech for First Amendment purposes.

The difference between insulting and offending, however, is not simply the difference between intending to insult and not (necessarily) intending to offend. Some people would probably take offense even at the use of their religion in a (somewhat irreverent) fictional example illustrating legal theory—and that is foreseeable. According to Brison, my use of such an example would be communicative if their offense was “at least a foreseeable consequence” of my action.³⁰

Does that mean I have also insulted them?

I think more facts would be needed to establish that claim. In general, it would be hard to come up with an example that did not potentially offend someone.³¹

But specific circumstances could make a given example more significant. Perhaps American professors have made a real habit of torturing cows, and the resulting outrage in the Hindu world is well known to all. Perhaps the practice has become such a sore point that even the playful, hypothetical condoning of cow torturing is now (along with, e.g., racism and child molesting) more or less beyond the pale. Under these special circumstances, my use of cow torturing as a casual example could only be a deliberate attempt to dredge up that whole painful history. It would be as if to say: “I know this has been a particularly painful episode for you [since, on my new facts, the outrage is “well known to all”], so for that very reason I am going to reintroduce it.” What was formerly at most an occasion for taking offense now communicates a deliberate and unmistakable insult.

When I insult you, it is essential that you recognize the insult for what it is: a direct communication of ill will from me to you. That recognition is part of the insult. Not only do I want you to feel bad—I also want you to know that I want you to feel bad. And I want that knowledge to be, at least in part, what makes you feel bad. Otherwise, you could feel offended as much as you like, but it might have nothing to do with me or only incidentally relate to me as having provided (perhaps unknowingly) the occasion for your offense. I cannot insult a dog, however much I might want to (even assuming it “distinguishes between being stumbled over and being kicked”).³²

Paul Grice makes this point with two examples that may at first seem puzzling.³³

(1) What is the difference between my showing you a photograph of your spouse in flagrante delicto and my drawing you a picture of the same? (Let's assume the photograph was taken automatically by a security camera.) Once I place the photograph in your hands, it no longer matters whether I or someone else gave it to you or even whether you found it in a drawer. The photograph itself is the best and only necessary evidence; it does “all the talking.” But as my drawing slowly unfolds, my (perhaps somewhat mysterious) intentions are at center stage, and even if the resulting picture is indistinguishable from the photograph, we would be justified in saying I “meant” something in drawing it that we would not say of my handing you the photograph (unless, perhaps, I do so in an unusually prolonged, suspenseful, dramatic way designed to draw attention to my presentation). The drawing is patterned on something in my mind, which in turn could either closely resemble reality or (unlike the photograph) be completely made up. My evident intention that you believe the former is your only basis for believing it. (2) What is the difference between a traffic police officer's gesturing for me to pull over and his standing in the middle of the road, blocking my way? In the latter case I have to stop in any event, whatever the officer's intentions (just as I would for a cow in the road). In the former case the officer gets me to stop only by getting me to recognize that he wants me to stop. Suddenly, his gesture means something: “Pull over!” (Otherwise I would not pull over.)

The defining characteristic of communication is a certain transparency of intentions. The general intention to communicate is understood more specifically as the intention to communicate a particular message. (Where this intention is evident but the underlying message is garbled, we have “miscommunication”; where the intention can be only dimly perceived, we have “doubtful” communication; where it cannot be perceived at all, we have “attempted” [or “failed”] communication; and where there is no such intention, there is no attempt to communicate.) Thus the question “How do we communicate?” quickly turns into the question “How do we make our intentions (to communicate) known?” Where there are no established conventions (like those of language) for doing so, we can rely on one of the most basic forms of meaning: contrast.

The significant “figure” on an otherwise plain “background” is a basic model of contrastive meaning. One of the most useful contrasts for purposes of communication is the contrast between the natural and the “nonnatural.” Doing something in a natural, normal, functional way is not likely to draw any particular attention to its character as a potential communication. Grice gives as an example: “If my college stopped my salary altogether, I should accuse them of ruining me; if they cut it by one pound, I might accuse them of insulting me.”³⁴

In the former case, it makes financial sense for the college to stop Grice's salary; it can save some money (perhaps it is financially strapped, perhaps it will eliminate the whole Philosophy faculty, etc.). Or perhaps the college really does want to “ruin” Prof. Grice. But whatever the motive (intention), it is presumably not primarily to “make a point,” which is precisely what we would say of the latter case, which has no discernable basis in finances or other noncommunicative purposes.

In my Examples 2 and 3 above, coughing in the normal way did not communicate. But upon hearing my stylized, ostentatious, highly exaggerated cough in Example 4, the boorish smoker could only wonder: “Now, why would anyone cough like that?”—unless, perhaps, I was trying to do something more than just cough, that is, trying to send a message by coughing. In doubtful or ambiguous cases, “one way for an actor to indicate that otherwise noncommunicative conduct is being used to convey information is to do the act in such a way as to show that it is obviously not meant to perform its ordinary function, and indeed, that the act is done for no evident noncommunicative reason at all.”³⁵

If you drive toward me at night with your lights off, I might turn my lights off and on rapidly. Seeing that, you might wonder: “Now, why would he do that?—unless, perhaps, he is trying to signal me that … oops! My lights are off.”

But where we have established conventions for communication, these elaborate mechanisms for ensuring the recognition of intention may no longer be necessary.³⁶

When I use the turn indicator on my car (in the usual way), it means “I am about to turn,” and it already includes the implied intention to communicate that fact (and to turn). But suppose I want you to know that the sun is shining. I might invite you into my office, lead you to the window, open the shades, and gaze out into the sunlight myself—all in hopes of getting you to see and realize that the sun is shining. But none of this is equivalent to telling you: “The sun is shining.” Indeed: Included in the conventional use of language is the implied intention to communicate. With a true act of communication I share my intentions with you, and your recognition of those intentions is an essential element of what I intend.³⁸

III

Now, what about the objection: “Any action can be made communicative. Therefore it is impossible for courts to distinguish speech from other forms of intentional action.”³⁹

There is a great deal of truth to the first part, and much of what I have said above shows how actions can be made communicative. But Brison's enthusiasm for this theme exceeds even mine: I shall take this charitably in the way Shakespeare doubtless intended—as metaphorical or figurative or even fanciful—and pass on to better examples. It was particularly easy to find speech at issue in Spence v. Washington because the defendant had taped one preexisting symbol (the peace symbol) onto another (the American flag). Flags, crosses, and other well-known symbols are, one might say, “inherently expressive.” Practically anything one does to or with them means something, because they are already expressive in their own right in multiple historical contexts.

But, as suggested above, other things and actions can be made expressive too, if only we provide them with suitable contexts. Conduct that is not speech in the usual senses but is intended to communicate may, under the right circumstances, count as speech for First Amendment purposes. This applies even to things and actions that are not normally thought of as expressive or communicative at all. Driving too fast and breaking the speed limit, for example, are not normally thought of as expressive. But suppose a certain group takes issue with the speed limits and finds them too restrictive, inconsistent with natural law, or whatever. Suppose further that the group wants to publicize its grievances and draw attention to its concerns; so it organizes a “Day of Speeding,” during which its members will deliberately break the speed limits at designated times and places.

Given this context and with enough public awareness, if you happened by on the “Day of Speeding” and saw a car rushing by at great speed precisely at one of the designated times and places, you might very well think to yourself: “There goes one of those speed limit protestors, trying to make a point.” (Here the speeder's intentional action would be communicative.) Let's assume that on any other day, you would merely have thought: “Wow! That car was really going fast—too fast.” (Here the speeder's intentional action would be noncommunicative.) So the speed limit protestors might very well succeed in communicating by speeding, given the unusual context they have created.

Does this mean that the speed limit protestors have a First Amendment defense to their violation of the law? That our speed limits are in danger of being struck down as unconstitutional? Or that courts would find it difficult here “to distinguish speech from other forms of intentional action”?

The short answer is “No,” and the short reason is that not all conduct intended to communicate counts as speech for constitutional purposes; and not all that does is protected by the First Amendment.⁴¹

So one should first ask “Is it speech?” and then ask “Is it protected speech?” The speeding example is designed to be a case in which courts would and should find that the activity—in the special circumstances described above—counts as speech for constitutional purposes. But that is hardly the end of the inquiry. In the Spence case, the Court went on to inquire into the purposes behind Washington's statute criminalizing “improper use” of the flag. In particular, the Court wanted to see whether this was “a case of prosecution for the expression of an idea through activity” or one “in an environment over which the State by necessity must have certain supervisory powers unrelated to expression.”⁴²

Corresponding to these two possibilities are two very different descriptions of “what happened.” In the speed limit example, someone familiar with the protest plans might say: “There goes someone protesting the speed limit.” But an equally valid description might be: “There goes someone driving very fast” (or more precisely—if it could be measured: “There goes someone driving at 80 miles per hour”). This is presumably the sort of description that someone unfamiliar with the protest plans would give. And indeed, this is presumably the more usual sort of description (unless speed limit protests are becoming epidemic, which, on the above facts, they are not). Most important, this latter sort of description presumably captures what the legislators who wrote the speed limit law were responding to. They wrote a law to solve the safety problem presented by driving very fast, or over 80 miles per hour, or whatever. Although facts could be invented such that the legislators were really trying to squelch speed limit protests, those are not the facts of this example. But they are the facts of the Spence case, where the Court found that “no … governmental interest unrelated to expression has been advanced or can be supported on this record,” and struck down the “improper use” statute.⁴³

This whole approach depends on the prior distinction (imperiled by Brison's analysis) between communicative intentional action and noncommunicative intentional action. A particular intentional action cannot be—at the same time and in the same respect—both communicative and noncommunicative. But if it makes sense to take a broader perspective and consider, separately, what can make a general type of action either expressive or nonexpressive, then both descriptions are possible. This in turn makes possible the legal distinction between laws related to speech and laws unrelated to speech.

Thus there should be nothing surprising or mysterious or paradoxical about the fact that almost anything can be made expressive or communicative; that courts can readily distinguish what does and does not count as speech for constitutional purposes; and that even an activity that concededly counts as speech may yet be restricted or even criminalized, so long as there is also a way to describe it that has nothing to do with speech. Neither description is inherently superior or more accurate or true; but regulation of the meaningful content of speech is presumptively invalid under the First Amendment,⁴⁴

so the law looks to that completely separate, alternative track where things are described most naturally in terms of “fast-moving cars” and the like. If this is truly the descriptive basis of the law, it will not be invalidated merely because some clever and creative soul can turn an otherwise nonexpressive activity into a forum for communication.

Were this not the case, our legal system would indeed be on shaky ground. But as it is, we can all now relax, secure in the knowledge that our speed limit laws will never be subject to constitutional invalidation.