When my son was about two, one of his favourite games was roaring. I would be the bear, or the lion, or some nameless monster, and I would roar at him, and he would shriek with delight and demand that I do it again. And again. And again. I was the performer and he was the audience, and his delight was the payment that kept the show running. Occasionally, though, I got it wrong. A roar too convincing, a lunge a little too real, and the shrieks became tears. The game collapsed, and it took a loving cuddle to mend.
What interests me is not that he was frightened, which is easy enough to understand, but that most of the time he wasn't, or not more than he could handle. He knew perfectly well what was going on, and he wanted the roar to be frightening, but not too frightening. He wanted, in other words, to enjoy a pretence without being overwhelmed by the feelings it elicited. That is a remarkable thing to want, and I think it may be closer to the heart of what makes us human than we might suppose.
The anthropologist Gregory Bateson noticed something similar in a zoo in 1952. "I saw two young monkeys playing," he wrote, "i.e. engaged in an interactive sequence of which the unit actions or signals were similar to, but not the same as those of combat." The playful nip, as he famously put it, denotes the bite. Both monkeys have to be in on it, or the play fight becomes a real one. Bateson called these "metalinguistic (but not verbalized) rules", which is an odd way to gesture towards the pre-linguistic. The gesture is important, but the terminology works against it.
Still, the observation itself is sound, and it is not confined to monkeys. Many dogs "bow" before play: the front end drops as if to spring, and then waits. Marc Bekoff found that they often bow again straight after a bite or a tumble that might have gone too far. It is tempting to say the bow means "still playing", but we needn't suppose anything so grand. A behaviour reliably followed by play comes to be expected to bring play, and that is all that's required. Many primates have a play face that works the same way as a contextual cue. Animals' play has its own reliable preludes and repairs, without anyone needing to treat them as signals, and good science avoids reading in more than the behaviour requires, whether linguistic rules, metalinguistic ones, or indeed any rules at all.
So play, with its preludes and repairs, is not what makes us unusual. Something else is.
Here's a wee clue. In a 2018 study of preschool children watching cartoons, Caspar Addyman and his colleagues found that children laughed eight times as much when watching with others as when watching alone, and smiled nearly three times as much. But they rated the cartoons as no funnier. The extra laughter wasn't a report of added amusement. It was for each other. Why else would anyone go to the pub to watch a match they could just as easily watch at home? Delight, it turns out, is something we give, and something we give back. Long before applause, which is a late and rather conventional invention, there was the gleeful posture, the giggle, the leaning in and the expectant smile. Visible delight is a strange sort of currency. It is minted freely and in abundance, it has no exchange rate, and it lasts only as long as the pleasure that earns it. But it changes hands all the same.
Some years ago I argued that value is not something we find in things but something that appears in practices of exchange: two mismatched appetites meeting, each with something the other wants. Entertainment fits this exactly. The performer offers skill and effort; the audience pays in delight and encouragement. And the encouragement does more than pay. It steers. The laugh (or even just the continued attention) tells the performer what worked, the performer refines it, the audience asks for more, and round it goes. That is how skill gets refined far beyond anything practical need would ever demand. Nobody needed my roar to be better. My son simply wanted to hear it again, and I gave it willingly, with pleasure.
Now consider two close contenders. Bonobos are famously more tolerant and playful than their chimpanzee cousins. They play far more as adults and use play to defuse tension, and some researchers argue that they have, in effect, domesticated themselves by selecting against aggression. Dogs, meanwhile, were domesticated by us, bred for fifteen thousand years or more for their attentiveness to human beings. They follow our pointing better than chimpanzees do. They read our delight and work to earn more of it, repeating the trick that got the laugh. And yet neither, so far as anyone has shown, sits as an audience. A bonobo plays, but does not sit back to watch others play for the pleasure of watching. A dog performs for our delight, but never returns the favour. Show a dog a game and it wants to join in fully. It cannot help itself. The exchange runs one way.
My son, by contrast, was roaring back at me in a calibrated interchange. Michael Tomasello calls this role-reversal imitation: the child who is tickled learns to tickle. Performer and audience are not fixed positions but seats that can be swapped, and to swap them you have to have sat in both. The performer knows what delights because they have been delighted. The audience knows what to encourage because they have had a go themselves.
But there is something more in the audience's seat, and I think it is the hinge of the whole thing. To be an audience is to refrain. To hold back from joining in, from intervening, from rushing the stage. To sit still, to watch from the sidelines, to peek, to steal a glance. This is not passivity; it is an achievement, and a hard-won one. Anyone who has taken a small child to the theatre knows how hard it is to sit still while something wonderful is happening, and pantomime is so beloved partly because it licenses the bursting out: "He's behind you!" Stendhal tells of a soldier in Baltimore who, in 1822, shot the actor playing Othello to save Desdemona. Whether or not it actually happened, the story is told because it's a clear failure of restraint: the spectator who could not stay a spectator, the human version of the dog that has to join in.
It's worth adding that this is not the same as an animal watching from cover. A cat in the long grass holds perfectly still, but it holds still in order to pounce. Its restraint is a means to an action of which it is a fully realised part. The audience holds still with nothing to follow but delight. The cat waits to strike; the audience waits only to see. Once again, it's the difference between participation and spectatorship.
Why does this matter? Because only from outside a frame can you grasp it as a frame. The participant lives inside the game. The spectator sees it as a game, with its occasions and roles and turns, its beginnings and endings, its traditions, its rules. And once you can see the frame, you can start to see something else: that what a thing is can depend not on what it is like but on where, when and how it relates to its wider context. A pretence of anger can look almost exactly like the real thing. What makes it a pretence isn't in the behaviour; it's in the setting, the occasion, the shared understanding of everyone present. Pretence is where we first learn that context can do work that often pays a hefty dividend.
Symbols need exactly this. Nothing about the word "cat" ties it to cats, and no amount of looking at the word will confirm it. A symbol holds only because those who use it can rely on one another to keep it in place, without checking it each time against the thing itself. That isn't resemblance worn thin. It's a different footing altogether.
This, I suspect, is why the development from mimicry to symbols so often looks like a gradual and continuous slide. It's easy to imagine resemblance thinning out, gesture by gesture, until only convention remains. But a thinner resemblance is only ever a sketchier match. What symbols need is something else: the ability to hold back, to wait, to carry on without confirmation. And that is just what audiences and performers teach each other. The audience learns to withhold its delight: to sit silent in the dark, to save its appreciation for a bigger payoff. The performer learns to carry on regardless, trusting that the pleasure is there even when nothing shows it. And these are only the crude beginnings. Much of what an audience feels, being impressed, moved or edified, has no behavioural expression of its own at all, which is presumably where applause comes in: one conventional gesture standing in for all of them. Sometimes the performer's reward arrives long after the room has emptied, as a review, a pay cheque or an invitation to perform again. By then delight has become something else entirely: a token, a sum, a promise. Knowing that you are bringing pleasure with no evidence of it in front of you is, I suspect, something no other animal does. A dog that gets no reaction soon gives up the trick. A storyteller facing a silent room keeps going. The pleasure is no less real for being inconspicuous. Getting there takes creatures who can stand outside a frame, and that takes creatures who can sit and watch.
It also gives us good reason to be cautious about attributing symbols to other species, and not out of any prejudice about brains. A dog that comes running at the word "walkies" is answering a cue, not sustaining a symbol. Dogs have intelligence, attentiveness, a shared form of life and millennia of selection in our favour, and they still have only one side of the exchange. What they lack is not a faculty but a practice: the practice of sitting on both sides of the frame.
We are, I think, uniquely an audience species. We gather in thousands to watch other people kick a ball, sit in the dark to watch shadows move across a screen, and laugh more at cartoons when we laugh together. It is easy to dismiss all of this as mere entertainment. But the word has an older meaning, of holding something between us, keeping it in mind, giving it room. To entertain an idea is to hold it without being taken in by it, to watch it from the side and see what it does. That is exactly what my son learned to do with a roar.
