The Beautiful Game, Magical Mands, and Divine Intercession


I’m still alive but I’m barely breathin’
Just prayed to a God that I don’t believe in
— from “Breakeven,” performed by The Script

acknowledgement: adrienne jennings provided helpful feedback on a draft of this post.

Every time I release a new blog post, I mutter to myself something like, “Please please please let somebody actually READ this.”

To be clear, when I say this it’s just me and my laptop, so my verbal behavior — something that’s inherently social, we’re told — has no actual audience. If you take my words at face value, you might think I believe those words can somehow change other people’s probability of reading the post.

I’m not the only one who acts as if verbal behavior has the power to dictate the future. Recently the author of a short essay in The Guardian (“How can you tell if someone’s really an atheist? Watch them at a penalty shootout”) offered up this observation regarding soccer (i.e., futbol, “The Beautiful Game”):

There are no atheists in a penalty shootout. I contend that most fans of the teams involved engage in something approximating prayer.

To explain, in some soccer situations tie matches aren’t allowed. A penalty shootout is a tiebreaker process in which players from each team take turns trying to kick the ball into a goal, challenged only by a goal keeper from the opposing team. The team with the most successful kicks wins the match, so the stakes are high, and given the fever pitch of soccer fandom, it’s (almost) logical that partisans would appeal to the highest possible authority for assistance.

This image has an empty alt attribute; its file name is penaltykick-scaled.jpeg
A soccer penalty kick. Amen.

This kind of behavior isn’t limited to sport. People shout at sluggish stop lights to change already. In a drought they plead with the clouds to drop a bit of moisture (or they may even perform a rain dance, a sort of non-vocal request or command). An acquaintance of mine orders her dog around despite the poor creature having completely lost her hearing two years ago. And there’s a superb YouTube video (known affectionately as “Crazy German Kid“) of a boy in the midst of an extinction burst rather colorfully commanding his computer game to load faster.

What’s fascinating about these instances is that the people involved are unlikely to really think their words have the power to dictate future events. For instance, most people understand that stop lights, computers, and deaf dogs do not respond to human pleas. But we inveigh them anyway. Even many people who believe in an all-powerful deity will acknowledge that the outcome of human sporting contests is not the sort of thing on which one may expect divine intervention (indeed, as religious scholars have noted, if partisans from competing teams pray to the same God with equal fervor for assistance, how would that God even choose a winner?).

Overall, we’re talking about more or less rational people doing a seemingly irrational thing with their verbal behavior. In technical terms, this verbal behavior takes the form of a mand, which in Verbal Behavior Skinner defined thusly:

A mand is characterized by the unique relationship between the form of the response and the reinforcement characteristically received in a given verbal community. It is sometimes convenient to refer to this relation by saying that a mand “specifies” its reinforcement.

But the cases we’re discussing violate the strict definition of a mand in a couple of ways: There is no conventional audience, and there is no history of reinforcement for the particular mand (nor, in a strict sense, can there be, given that the “audience” is incapable of coming under instructional control). Skinner anticipated this problem, suggesting that stimulus generalization might be responsible.

An example of extended stimulus control is seen when people mand the behavior of dolls, small babies, and untrained animals. These “listeners” cannot possibly reinforce the behavior in characteristic fashion. Nevertheless, they have enough in common with listeners who have previously provided reinforcement to control the response, at least when it shows appreciable strength…. We acquire and retain the response Stop! because many listeners stop whatever they are doing when we emit it, but as a result we may say Stop! to a car with faulty brakes or to a cue ball which threatens to drop into a pocket of the pool table.

There are mands which cannot be accounted for by showing that they have ever had the effect specified or any similar effect upon similar occasions. The speaker appears to create new mands on the analogy of old ones.

Several studies verify that mand repertoires can generalize to new situations (see here and here and here). It can be a good thing when mands generalize — for instance, when a child who learned to ask Daddy for food can do so with any adult. But in some cases, of course, the generalization is to a false audience, that is, to something (“dolls, babies, and untrained animals”) incapable of being under instructional control. In Skinner’s words:

This sort of extended operant may be called a magical mand…. Flushed with our success under favorable reinforcing circumstances, we set out to change the world without benefit of a listener.

To be clear, magical mands are a spillover effect from the forces that shape up practical mands (ones that actually work), but as Caio Miguel has pointed out, this isn’t always garden-variety stimulus generalization, in which the setting for a new mand is similar to that in which manding previously was reinforced. Mands are always driven by some motivating operation — in casual terms, we mand when we “want something” — so novel (and nonsensical) mands can reflect generalization based on similarity of  MOs. A child who successfully mands for food may spontaneously do so for water if food and water deprivation are experienced similarly. An adult who has verbally directed the movement of human beings (“Go to the corner and take a right”) may attempt to do so for a soccer ball (“Go into the goal!”) [In both cases, speaking loosely, the person wants something to go somewhere].

In writing Verbal Behavior, Skinner explained a lot about mands, both magical and practical, but what he couldn’t have anticipated is the complex derived stimulus relations that behavioral research began to illuminate starting in the 1970s. Since then Relational Frame Theory has guided research showing how novel mands (including those deemed “magical”) might arise quite separately from simple generalization (see here and here). For anyone interested in fully understanding the mand repertoire that stuff is worth a read because it shows that the emergence of novel mands is far more complicated than Skinner’s 16-page introduction to it implies.

There’s one more critical wrinkle about novel mands, one that Skinner did anticipate:

In moments of sufficient stress, the speaker simply describes the reinforcement appropriate to a given state of deprivation or aversive stimulation. The response must, of course, already be part of his verbal repertoire as some other type of verbal operant.

Those motivating operations mentioned earlier — the more powerful they become, apparently, the more likely a mand repertoire is to propagate into some context where it hasn’t previously been reinforced. Hence people seem most likely to verbally “dictate the future” when the stakes are high. When I’m not in a hurry, for instance, I tend not to bark orders at a leisurely stoplight. But let me be late for class, and I will verbally rip that light a new metaphorical asshole. In neither case, of course, does the stop light care.

The soccer article I mentioned referenced a man who under normal circumstances was entirely nonreligious — for instance, he never, ever, under normal circumstances, asked an unseen deity to alter the future on his behalf. But the man freely “concedes he resorted to prayer when brain tumours put his life in danger.”

Religion is a fascinating crucible for evaluating “extended” mands.

Regarding the presence or absence of an audience for prayer, well, to some extent that is in the eye of the beholder. To a nonbeliever, spoken prayers are merely sound waves cast into a void. Any presumed audience is a figment of the believer’s imagination. Yet I think that nonbelievers sometimes trivialize, and therefore oversimplify, the experience of the believer. To the nonbeliever, prayer is a classic case of magical manding in part because of the absence of an audience. And yet it might be said that, where verbal behavior is concerned, it matters less that a communication partner exists than that one is perceived.

Skinner himself demonstrated as much in the 1930s with the “verbal summator,” which presented non-language speech sounds that listeners often “heard” as coherent statements. More to the point, in the 1960s “Eliza,” an early chatbot programmed to mimic therapist verbal behavior, fooled many people into thinking they were interacting with a human therapist via computer interface. Today, “audience simulators,” like Apple’s Siri and Amazon’s Alexa, are everywhere, which begs a critical question. When a human converses with Siri or Alexa, is this verbal behavior? Consider a specific example: When I tell Alexa to turn out the kitchen light, there is an audience (just not a human one), and my instructions arise partly from a history of reinforcement (Alexa usually does what I ask her to). How is that different from human-specific manding?

When religious believers pray they believe they have an audience. Moreover, they believe the audience is capable of answering prayers because experience says this sometimes happens. After all, sometimes one’s prayers are followed by events that seem related to them (penalty kicks sometimes go into the goal!). Yes, the nonbeliever will call this adventitious reinforcement, but from a verbal behavior perspective it doesn’t matter whether it’s adventitious reinforcement or divine intercession. From the believer’s perspective, a prayer was answered, and an answer requires an Answerer. In this sense prayer satisfies the two prerequisites of the mand: an audience and a suitable history of reinforcement. To me, that makes prayer a practical mand… unless, of course, it comes from a nonbeliever.

Which brings us back to that man who freely “concedes he resorted to prayer when brain tumours put his life in danger” and raises an interesting question about motivating operations. If the motivation for manding is strong enough, does that affect the tendency to believe in an audience that may not actually exist? In radical behaviorism, belief is behavior, and therefore malleable by the same forces that alter physical behavior.


  1. In Verbal Behavior, Skinner provides an account of literary behavior that is not too different from what I’ve described here. A writer writes for, and often mands (e.g., “Call me Ishmael”) to, an audience that is not present at the time of writing and in fact may never be encountered. In terms of the momentary control of writer verbal behavior, then, this audience is strictly hypothetical. Unlike with prayer, however, the writer’s entreaty to the reader may never be reinforced (adventitiously or otherwise). It might therefore be suggested that prayer is sometimes less “magical” than literary manding. What Skinner gets right is that literary composition, like prayer, is often the product of strong states of deprivation (“Evidently the lyric poet needs many things and needs them badly.”). Strong motivating operations, as mentioned above, can promote the generalization of practical mand repertoires to “magical” circumstances.
  2. The most interesting form of verbal behavior in the universe, cursing, can have mand properties when it is directed at an “impossible” audience (that car in front of you that’s traveling 10 mph below the speed limit; the hammer that just accidentally struck your thumb; etc.). But cursing is multiply determined, so f@#k it, I’ll save this topic for a future post.
  3. Check out this fun but challenging treatise on functional similarities and dissimilarities between verbal and nonverbal events.
  4. I hope it is obvious that I’m not taking a stand on the ontological status of anyone’s preferred deity. My goal in mentioning prayer is simply to show that, functionally, it’s less different from everyday practical manding than might otherwise be assumed. Therefore, hypothetical reader, please please please don’t brand me a religious bigot!


“Breakeven” was written by Andrew Marcus Frampton, Daniel John O’Donoghue, Mark Anthony Sheehan, and Stephen Alan Kipner. Lyrics © BMG Rights Management, Sony/ATV Music Publishing LLC, Universal Music Publishing Group