Roman Laughter in Latin and Greek


The study of Roman laughter is in some ways an impossible project. That is partly what makes it so intriguing, so special, so enlightening, and so worthwhile. As I hope I have made clear already (perhaps too clear for the tastes of some readers), the laughter of the past is always likely to frustrate our most determined efforts to systematize and control it. Anyone who—with a straight face—claims to be able to offer a clear account of why or how or when Romans laughed is bound to be oversimplifying. But in the inevitable confusion (in the mess left in laughter’s wake), we still learn a lot about ancient Rome and about how laughter in the past might have operated differently. This is a subject (like many, to be honest, in ancient history) in which the process of trying to understand can be as important and illuminating as the end result.

But process isn’t everything, and we should not entirely accept defeat before we begin. Whatever the tricky problems that I have been enjoying so far, there are also some striking and relatively straightforward observations to be made about how laughter works in the Latin language and in Latin literature. In fact, to investigate Roman laughter is to engage with some of the most basic and familiar words in Latin (those that even the rawest beginner is likely to have encountered), as well as some rather more recondite vocabulary. It also involves exploring some of the less-trodden byways of Latin literature, as well as throwing fresh light on some of the most canonical Latin texts we have.

One of most important of these observations concerns the Latin vocabulary of laughter. It is only a slight exaggeration to say that there is just one word in Latin for “laughing.” In modern English, we are used to a range of subtly nuanced (even if elusive) terms for laugh: from chuckle andchortle through giggle, titter, and snigger to howl and guffaw—not to mention such related words as grin, beam, smile, and smirk. Ancient Greek too has a wide range of laughter vocabulary, from the standard gelan and its compounds through variants such as kanchazein (a more robust form) and sairein (e.g., Commodus’ grin; see p. 6) to the delightfully onomatopoeic kichlizein (not far from our giggle) or meidian (often translated as “smile”). In Latin we are dealing, for the most part, with just the word ridere, its compounds (adridere, deridere, irridere, and so on), and its various cognates as adjectives and nouns (risus, “laughter”; ridiculus, “laughable”). All of these signal some form of audible, physical reaction or gesture broadly and recognizably akin to laughter as we know it. Dictionary definitions and some modern critics try to calibrate these variants precisely, from deridere, for example, signaling derision to irridere ridicule or laughing at. Yet the terms are almost certainly much less fixed, referentially, than such definitions imply.1

The confidence with which it is often assumed, for example, that adridere always refers to supportive laughter or, pejoratively, flattery, is quite misplaced. True, sometimes it does: Ovid tells his learner lover to make a good impression by joining in the laughter (adride) whenever his would-be girlfriend laughs; the hallmark of comic toadies is “to offend no one and be a total yes-man” (adridere omnibus); and Horace uses the word in the context of sympathetic laughter.2 But it is certainly not always so supportive, as phrases such as “laughing savagely” (saevum adridens) make absolutely clear.3 In fact, in another passage of Terence’s Eunuch, Gnatho exploits the potential double entendre of the word when he reflects on his life as a scrounger and his relationship with the (rather dim) guys who are his meal tickets: “I don’t set out to make them laugh at me, but actuallyeis adrideo and compliment their wit at the same time.” The joke here turns on the possible slippage in the phrase eis adrideo between “I flatter them” and “I laugh at them.” Is Gnatho merely toeing the subservient line, or is he hinting to the audience that he has the upper hand in dealing with the likes of Thraso? Who, in other words, is laughing at whom? The ambivalence is half-seen and half-missed by one late antique commentator, who wrote simply that Terence had used “arrideo instead of irrideo.”4

Some modern critics have been even more confident than this in suggesting which Latin word should be used where, even inserting the “correct” term where necessary. One glaring case concerns the text of an epigram of Martial. The poem is a squib addressed to one Calliodorus, who fancies himself a great jester and so dinner party guest, and according to the manuscript tradition includes the phrase omnibus adrides. The most recent editor, with staggering self-confidence, has simply replaced this with omnis irrides. Why? Because, he explains, “adrides must mean either ‘you smile at approvingly’ . . . or ‘you please.’ . . . Neither fits Calliodorus. . . . The word for his activity can only be irrides.”5 Such rewriting is the price you have to pay if you want to preserve neat linguistic boundaries.

Beyond ridere and its linguistic family, there are few Latin alternatives. Occasionally, words such as renidere (shine out) do metaphorical duty for some shades of laughter or facial expression (renidere is, more or less, “to beam”).6 Rictus can refer (unflatteringly) to the open mouth or gaping jaws that are inevitably part of the laughing process, as well as to the bared teeth of an animal.7 Elsewhere, cachinnare or (more commonly) the noun cachinnus can be used for a particularly raucous form of laughter or for what we might call “(a) cackle.” As one late Roman grammarian, Nonius Marcellus, put it, it had been used to signify “not just laughter [risus] but a stronger sound.”8 The words have a catchy onomatopoeic ring but again are harder to pin down than dictionary definitions imply and prove resistant to the very precise classification that we might like to impose on them.

It is true that a contrast between cachinnare and (mere) ridere is sometimes more or less spelled out. Cicero, for example, at one point in his broadside against Verres, the infamous governor of Sicily, turns to attack Verres’ nasty sidekick Apronius, for humiliating a supposedly upstanding member of the Sicilian elite; Cicero pictures a banquet at which “his fellow guests laughed [ridere], Apronius himself cackled [cachinnare].”9 Likewise, in what was effectively his manifesto poem, the satirist Persius was clearly trying to outdo his predecessor Horace in describing his own reaction to the folly of the world as cachinnare, not Horace’s gentler ridere.10

However, the word is not always so loaded, so aggressive, or so loud. It is the pleasant sound of laughter (cachinni), along with wine, wit, and a pretty girl, that sums up the atmosphere of a friendly party at the poet Catullus’ house; it is the laughter of disbelief (cachinnasse) with which, in Suetonius’ biography, Vespasian’s grandmother reacts to the unlikely omen that her grandson will become emperor; and it is the furtive giggles of servant girls (furtim cachinnant) laughing at their mistress behind her back.11 What is more, metaphorical usage too reflects that range. Cachinnareand cachinni, both verb and noun, are used to evoke the sound of water—from the pounding of the ocean to the gentle rippling of Lake Garda.12 Cackles or giggles or ripples? We should always hesitate before assigning too rigid or precise a value to Latin terms for “laughing” or “laughter.”


So far I have not pointed to a word that corresponds to our own smile. I mean that curving of the lips that may, or may not, be a preliminary to a fully vocalized laugh—but is independently one of the most powerful signifying gestures in the modern Western world. From “Smile, please” to smiley faces, it underpins for us all kinds of human interaction, signaling warmth, greeting, wry amusement, disdain, affection, confidence, ambivalence, and much more. It is hard for us to imagine social life happening without it, yet it is hard to find a Latin equivalent.

In ancient Greek the position appears somewhat simpler. The word meidiaō may be much more distant from our smile than that standard translation implies. In Homer and other early writers, meidiaō can also be a sign of hostility, aggression, or superiority, and in general it seems to be treated as a gesture of the face as a whole rather than just the lips.13 But as Halliwell shows, it does overlap in part with our usage, notably because unlike laughter, and like our “smiling,” it makes no noise (or as he more carefully puts it, “It is impossible . . . to show that meid- terms ever imply vocalisation”).14 In Latin there is no specific term of that sort. When Virgil evoked the “smiling” gods of Homer, he often fell back on another compound of ridere, that is subridere, which technically means a “suppressed or muffled laugh,” even a “little laugh.”15

Renidere (to beam) can also, metaphorically, signal a silent facial expression that seems akin to a smile. This is how the poet Catullus has Egnatius famously reveal his urine-cleaned teeth: “Egnatius . . . renidet.” And Robert Kaster, in exploring the world and the text of Macrobius’ Saturnalia,has not only translated the word as “smile” but also suggested that these “smiles” play a particular role in articulating the learned discussion that is staged in the dialogue. Phrases such as “Praetextatus smiled” (Praetextatus renidens) tend to greet an ignorant, out-of-place comment by some (usually inferior) participant in the discussion, and they invariably herald a pronouncement by an expert “which admits no contradiction.” Kaster is an acute observer of the structure of this late antique debate and of the hierarchies within it. But it is far less clear than he suggests that this “beaming” is a close match for our own category of grandly supercilious smiling—those “gestures of magnificent condescension,” as he puts it.16

Other, more discursive, metaphorical uses of the word outside Macrobius—admittedly often centuries earlier than the Saturnalia—are varied but revealing. Catullus certainly likens the expression (renidet) to laughing, but Egnatius’ determined display of his white teeth is an absurd form of laughter (risus ineptus) and so is itself laughable. In Ovid, renidens is (twice) the expression of foolish optimism on the face of young Icarus, in Livy it is that of the boastful trickster, and Quintilian also uses it of a misplaced sign of pleasure (intempestive renidentis).17 Repeatedly, as with the Greek meidiaō, the emphasis is on the facial expression as a whole (hilaro vultu renidens, renidenti vultu, renidens vultu18), not specifically the lips—as is also once made explicit in Macrobius: vultu renidens.19 For the most part, the common defining feature of this gesture seems to be the facial “glow” (of confidence, whether well-placed or misplaced) rather than the oscular curve, or “smile” as we know it.20

So did the Romans smile? At the risk of falling into the trap of overconfident classification that I have been criticizing, my working hypothesis is “by and large, in our terms, no.” But that is not (simply) for linguistic reasons, and it needs to be argued rather carefully. The cultural significance of smiling may be reflected in, but is not wholly circumscribed by, language. Several modern European languages (English and Danish, for example—like ancient Greek) have separate word groups, from separate linguistic roots, that distinguish “smile” from “laugh.” Others (notably the Romance descendants of Latin) do not. Reflecting those Latin roots, modern French uses sourire for “smile,” just as Italian uses sorridere (both derived directly from subridere; respectively cognate with the French rire and the Italian ridere). Yet both of these modern cultures have an investment in the social significance of smiling, as distinct from laughter, no less intense than that of (for example) their modern Anglo-American counterparts.

Nonetheless, the linguistic patterns of Latin do seem to accord with other negative hints which suggest that smiling was not a major part (if a part at all) of Roman social semiotics. Only the most hard-line ethologists, neuroscientists, and their followers hold to the human universality of such facial gestures—whether in form, type, or meaning.21 Crucially important for me is that we find in Roman literature none of those distinctions between smiling and laughing drawn by the likes of Lord Chesterfield (for whom a silent smile was a sign of decorum, in contrast to “loud peals of laughter”),22 and—whatever is going on in Macrobius—we see no clear evidence that smiling as such was a significant player in Roman social interactions in general. “Keep smiling!” and the like were sentiments unheard of in Rome, so far as I can tell, and as Christopher Jones has shown, two Romans meeting in the street were likely to greet each other with a kiss, where we would smile.23

Of course, arguments from silence are always perilous, especially when the process of spotting the smile is necessarily an interpretative one. But it is hard to resist the suggestion of Jacques Le Goff that (in the Latin West at least) smiling as we understand it was an invention of the Middle Ages.24 This is not to say that the Romans never curled up the edges of their mouths in a formation that would look to us much like a smile; of course they did. But such curling did not mean very much in the range of significant social and cultural gestures at Rome. Conversely, other gestures, which would mean little to us, were much more heavily freighted with significance: Caesar scratching his head with one finger, which would now indicate no more than an annoying itch, could give Cicero the hint that Caesar posed no danger to the Roman Republic.25

There is an important lesson in this. It has become standard practice when translating not only subridere but also ridere itself and its other cognates into English to use the word smile where it seems more natural to us than laugh (even some famous lines of Virgil have been the victim of this tendency; see pp. 84–85). This has a doubly misleading effect. It tends to give smiling a much bigger presence in Roman cultural language than it deserves—or ever had. And in offering an apparently “better” translation, it tends to erode the potential foreignness of Roman patterns of laughter, to make them look increasingly like our own. To be sure, we cannot absolutely prove that there was no strong and meaningful Roman tradition of smiling that lurked underneath the general rubric of ridere. We need to remain alert to that possibility. But we should also resist the easy temptation to reconstruct the Romans in our own image. So even where laugh may seem awkward, I shall use it as the first option in translating ridere and its compounds and cognates: that is not to say that even the English word laugh captures exactly what the Romans meant by ridere, but it is certainly less misleading than smile. And that awkwardness is, after all, part of the historical point.


We are not simply dealing with the poverty in the Latin vocabulary of laughter compared with the richness of (say) Greek, or with a simple lack of cultural discrimination in classifying laughter’s various forms. We are dealing with a different richness of vocabulary and perhaps with a significantly different set of cultural priorities. For however few the Latin terms for laughter may be, the terms for what may provoke it—in the forms of jokes and witticisms—are legion. To list just some: iocus, lepos, urbanitas, dicta, dicacitas, cavillatio, ridicula, sal, salsum, facetiae. We can no more define the precise difference between dicacitas and cavillatio than we can define how exactly chortle differs from chuckle. But the contrast with the Greek range of vocabulary—which is overwhelming dominated by two words for joke, geloion and skōmma—is striking.26 Whatever the origin and history of these terms (on which see further chapter 5), their range and variety point to a Roman cultural concern with the provocation of laughter and with the relationship between the laugher and whoever prompted the laughter (both joker and butt).

Interestingly, Roman popular sayings also seem to reflect these priorities. Proverbs and slogans about laughter are common in modern English-speaking culture: “He who laughs last laughs longest,” “Laugh and the world laughs with you” (or, to quote a Yiddish proverb, “What soap is to the body, laughter is to the soul”). Overwhelmingly, they treat laughter (and its effects) from the point of view of the person who laughs. Romans also sloganized laughter, but much more frequently these slogans stressed the role of the joker rather than the laugher (“It’s better to lose a friend than a jest,”27 “It’s easier for a wise man to stifle a flame within his burning mouth than keep his bona dicta [wit or quips] to himself”28) or focused on the relationship between the laugher and the object of their laughter or on questions of who or what was an appropriate target for a jest (“Don’t laugh at the unfortunate”29). To put this another way, where most modern theory, and popular interest, is firmly directed toward the laugher and to laughter’s internal coordinates, Roman discussions tended to look to the human beings who caused laughter, to the triangulation of joker, butt, and laugher—and (as we shall see in the next chapter) to the vulnerability of the joker, no less than of the person joked about.


One of the pleasures of tracking down Roman laughter is that it leads to some extraordinary—surprising and even startling—works of Latin literature still somewhat off the beaten track, unfamiliar even to most professional classicists. We find all kinds of glimpses into Roman laughter in some unexpected places, and there is no shortage of them. They include long discussions that broach, directly or indirectly, the question of what makes people laugh, reflect on the protocols and ethics of laughing, or use laughter as a marker of other cultural values at Rome. No discussion of laughter is ever neutral.

So, for example, laughter features as one diagnostic of the emperor’s mad villainy or perverse extravagance in the biography of the third-century CE emperor Elagabalus—which belongs to that strange, partly fictional, partly fraudulent, but hugely revealing collection of imperial lives known as the Augustan History (or Historia Augusta—the history, that is “of the emperors,” Augusti).30 In what is almost a parody of a pattern that we shall see repeated in the lives of earlier emperors in less tendentious accounts (see chapter 6), Elagabalus outdid his subjects in laughter as much as in everything else. In fact, he sometimes laughed so loud in the theater that he drowned out the actors (“He alone could be heard”)—a nice indication of the social disruption caused by gelastic excess. He also used laughter to humiliate. “He had the habit too of inviting to dinner eight bald men, or else eight one-eyed ones, or eight men with gout, or eight deaf men, or eight with particularly dark skin, or eight tall men—or eight fat men, in their case to raise a laugh from everyone, as they could not fit on the same couch.” It was not so much the mad replication that caused the laughter but rather his slapstick exposure of the victims’ fatness. There was a similar comic style in his experiment with a Roman prototype of whoopee cushions: “Some of his less prestigious friends he would sit on airbags, not cushions, and he had these deflated while they were dining, so that the men were often suddenly found under the table in the middle of their meal.”31 This is a combination of power, dining, laughter, and practical jokes to which we shall return.

An even richer discussion that often goes unnoticed (or is merely pillaged for some of the individual jokes it contains) is found in the second book of Macrobius’ Saturnalia. Writing in the context of a highly learned, late antique subculture, Macrobius (through the scripted contributions of his various characters) offers the closest thing we have from the ancient world to an extended history not so much of laughter but of joking, and, indirectly at least, he reflects on different styles of jokes and on the nature and importance of “old jokes.”

The scene is simple. In keeping with the lighthearted atmosphere of the festival, the Saturnalia, that provides the dramatic context of the work, each of the discussants in turn picks a joke from the past to recount to the others (Hannibal and Cato the Elder are the earliest Roman “jokers” cited, though—true to type—the Greek character in the discussion, Eusebius, contributes a quip from Demosthenes, and the Egyptian Horus picks an epigram of Plato’s).32 This leads on to a rather more systematic anthologizing of the quips of three historical characters—Cicero, the emperor Augustus, and his daughter, Julia—and occasionally to wider reflections on laughter.33 In part, Macrobius’ account matches the standard historical template, with its emphasis on antiqua festivitas and the fearlessness, if not the rudeness, of the jokers of earlier times.34 But it also carefully shows what hangs on the choice of a favorite joke and how that choice may relate to character. Predictably, it is one of the uninvited guests, the oddball bully Evangelus, the man most concerned to undermine the atmosphere of literary high culture, who chooses the joke about sex; the buttoned-up grammarian Servius can hardly bear to tell a joke at all and in the end settles for a dry piece of wordplay.35

The final section of their discussion turns, significantly, to another key institution of Roman laughter: mime (in Latin, mimus). This particular form of dramatic display was not, as its name in English might suggest, a silent affair, dependent on gesture alone, but a performance with words, sometimes improvised, sometimes scripted, and both male and female actors. Its precise character and history are much less understood than modern textbook accounts sometimes suggest, as is its precise relationship to another ancient genre—pantomime. But two features are clear. First, mime could sometimes be very bawdy, and our genteel debaters of the Saturnalia are careful to stress that they will not actually bring the mimes into their banquet, only a selection of the jokes—so avoiding the bawdiness (lascivia) but reflecting the high spirit (celebritas) of the performances.36Second, it was the one and only cultural form at Rome whose primary, perhaps even sole, purpose was to make you laugh. So Roman writers repeatedly stressed—and that was the message blazoned on the tombstones of some mime actors.37

I shall later argue (see pp. 167–72) that the hilarity so strongly associated with mime is one aspect of the more general importance of imitation and impersonation in the production of Roman laughter, from actors to apes. But Macrobius’ discussion already gestures in that direction with a series of stories about the competition between two pantomime actors, Pylades and Hylas, to present convincing imitations of mythical characters. In the cleverest of these, the audience is reported to have laughed at Pylades, who was playing the mad Hercules, because he was stumbling around “and wasn’t maintaining the manner of walking appropriate to an actor.” He took off his mask and berated them: “Idiots,” he said, “I’m playing the part of a madman.” In a nice twist, the audience turns out to have been laughing at a man for what they imagined was a bad piece of acting, when in fact it was a perfect example of (laughable) impersonation.38

Sometimes it is not a lengthy discussion, such as Macrobius’, but just a couple of unnoticed words in some little-read text that can shed unexpected light on the operations and significance of laughter in Roman culture. The collected volumes of Roman oratorical exercises that go under the general title of Declamations have recently attracted some keen scholarly attention, but even so they are still relatively underexploited. A combination of rhetorical training and after-dinner entertainment, these exercises usually started from a fictional (or at least fictionalized) legal case, on which the learner orators or celebrity after-dinner speakers would take different sides, for defense or prosecution. The collections gathered together some of these cases, along with excerpts from particularly notable speeches by famous rhetorical showmen; they represent, in a sense, both a manual of models to imitate and a compilation of oratorical “greatest hits.”39

One telling example, from the collection compiled by the elder Seneca in the early first century CE, concerns a (fictionalized) version of the case of Lucius Quinctius Flamininus, who was expelled from the Senate in 184 BCE for inappropriate conduct while holding office.40 Several shorter and slightly different variants survive elsewhere in Latin literature,41 but the declamation centers on the relationship between Flamininus and a prostitute, whom—in his infatuation—he had taken with him when he left Rome to govern his province. At dinner there one evening, she remarked that she had never seen a man’s head cut off, so to please her, Flamininus had a condemned criminal executed right in front of her in the dining room. Then, in the fictionalized world of the declamation, he was accused of maiestas (often translated as “treason” but better as “an offense against the Roman state”).42

The oratorical highlights focus not on the rights and wrongs of the execution of the criminal as such (the man had, after all, been condemned to death anyway) but on its context. The declamation is in fact a treasure-house of Roman clichés on the proper separation of the official business of state from the pleasures of ludic entertainment and the jocular world of the dinner party. Many of the quoted speakers found snappy ways of summing up this underlying issue. Taking “the forum into a feast” (forum in convivium) was no better than taking “a feast into the forum” (convivium in forum), quipped one. “Have you ever seen a praetor dining with his whore in front of the rostra?” asked another, referring to the raised platform in the Forum from which speakers traditionally addressed the Roman people.43

Held up for specific criticism is the fact that the executioner was drunk when he killed the man and that Flamininus was wearing slippers (soleae), both signs of private pleasure rather than official duty. But another marker of transgression lies in the “jokes” being made of the serious business of state. An execution has been turned into “a dinner table joke” (convivales ioci), Flamininus is himself accused of “joking” (ioci), and the woman is said to have been “making fun” (iocari) of the fasces, the symbols of Roman power. In fact, according to one of these rhetorical reenactments of the terrible scene, when the unfortunate victim was brought into the room, the prostitute laughed (arridet)—not, as the translation in the Loeb Classical Library has it, with very different implications, “smiled.”44 There is, I suspect, a sexual resonance here; laughter was often associated with ancient prostitutes, so it is exactly what you might expect this, or any, whore to do.45 But more than that, the single word arridet (emphatically at the end of the sentence) underlines the irruption of gelastic frivolity into the world of state business.46

What happened next, however, brings into focus a different role of laughter in the social interaction around this dinner table. The whole occasion is written up in decidedly melodramatic terms (we are asked to imagine at one point that the unfortunate criminal misreads the scene as the preliminary to a pardon and actually thanks Flamininus for his mercy). But what did the other guests do once the execution had been carried out? One man wept, one turned away, but another laughed (ridebat)—“to keep in with the prostitute” (quo gratior esset meretrici).47

This is laughter provoked by something quite different from the jokes of Macrobius. Jocular and (transgressively) ludic though the laughter of this whole scene may be, there are no verbal quips to prompt the outbursts. We see instead the laughter of (inappropriate) pleasure on the part of the woman and the laughter of flattery, or (to put it more politely) of social alignment, on the part of another dinner guest. This is another example of that nexus of signals implied by a laugh—from pleasure to approval to outright sycophancy—to which we shall return.


The study of laughter does not merely reanimate some less-known works of Latin literature; it also encourages us to look again, through a different lens, at some of the most canonical. We have already glanced at Horace’s Satires and at Catullus. There are many more cases where laughter plays a role, sometimes disputed, in the most famous Latin classics to have survived from the Roman world: from Ovid’s Art of Love, with its parodic set of instructions to young women on how to laugh,48 through Virgil’s reference to Venus’ laugh, which enigmatically seals the discussion between her and Juno at the beginning of Aeneid 4 (and with it the fate of Dido),49 to the opening of Horace’s Art of Poetry, where he lists the kinds of representational incongruities that would, he claims, make anyone laugh (“If a painter wanted to put a horse’s head on a human neck . . . would you be able to keep your laughter in?”).50

The most famous, and controversial, of all such references to laughter, however, is the especially puzzling end to Virgil’s puzzling fourth Eclogue. This poem was written around 40 BCE, against the background of promising attempts—fruitless as they proved in the long term—to secure peace in the civil war between Octavian (the future emperor Augustus) and Mark Antony. It heralds the coming of a new golden age for Rome, embodied in or brought about by the birth, imminent or recent (the chronology is vague), of a baby boy. Virgil celebrates this baby in messianic terms (hence the title “Messianic Eclogue” often given to the whole poem)—“the boy under whom . . . a golden race shall rise up throughout the world” and so on. But who was the baby? This has been a major source of dispute for centuries, with suggestions ranging from the yet unborn child of either Octavian or Mark Antony (both of whom turned out, inconveniently, to be girls) through a purely symbolic figure for the return of peace to Jesus—whose birth, this idea goes, Virgil was unwittingly prophesying.51 But almost equally controversial has been the significance of the last four lines of the poem (60–63), which address the baby and focus on the “laughter” (risus) exchanged between him and his parent(s). What is thisrisus, and whose risus is it anyway?

Once more, the details of the argument focus on exactly what the Latin author wrote and how accurately the medieval manuscripts, on which we rely, reflect that. The main issue comes down to the origin and direction of the “laughter” and depends on the difference of just a few letters. The crux is this. In the poem’s final couplet, was Virgil thinking of the risus of the baby, directed either to his parenti (singular, dative case, presumably his mother52) or to his parentes (plural, accusative case, meaning mother andfather)? Or did he mean that the risus of the parentes (here a nominative case) was directed at the baby? And what hangs on this? The argument is technical and ultimately, let me warn you, inconclusive—and it involves Latin words that to the innocent eye are identical (or almost so), even if they point to significantly different interpretations. But it is also very instructive and well worth pursuing in all its intricacy. For it puts laughter right back into the heart of a debate about one of the most classic of all classical texts while exposing the pitfalls of not reflecting carefully enough on the linguistic rules and cultural protocols of Roman laughter.

All the main surviving manuscripts run:

    Incipe, parve puer, risu cognoscere matrem

(matri longa decem tulerunt fastidia menses);

incipe, parve puer: cui non risere parentes,

nec deus hunc mensa, dea nec dignata cubili est.

Literally, this means “Begin, little boy, to recognize your mother with risus (to your mother ten months [of pregnancy] have brought long distress); begin, little boy: he on whom his parents have not risere, no god thinks worthy of his dinner table, no goddess worthy of her bed.” The idea (frankly “enigmatic” as it is53) must be that the starry, divine future of the child depends on his parents’ warmth for him now, reflected in their risus toward him.

But most modern editors of the poem have thought this so enigmatic, not to say unconvincing, that they have chosen to adjust the text in order to change the nature of the interaction described. Instead of having the parents (parentes) direct their risus toward the baby (cui), they have the baby (qui substituted for cui) directing his risus toward his parent—that is, his mother (parenti). On this reading, the interaction of the final two lines runs as follows:

    Incipe, parve puer: qui non risere parenti,

nec deus hunc mensa, dea nec dignata cubili est.

Or, “Begin, little boy: those who have not risere on their parent, no god thinks worthy of his dinner table, no goddess worthy of her bed.” In other words, it is what the baby himself does that paves the way for his future greatness.

There are some strong reasons for making these changes. In general, the revised text seems to make better sense. For one thing, the phrase “Begin, little boy” seems to demand some action on the part of the baby, not—as our manuscript reading would have it—on the part of the parents. For another, the idea that the entirely “natural” response (risus) of the parents to their child should be prophetic of his future seems hard to fathom. What is more, although there is no direct support for it in any of the manuscripts of Virgil, this does seem to be much closer to the text that Quintilian had in front of him just a century or so after Virgil wrote—as we know, because he refers to this particular passage in discussing a tricky point of Roman grammar.54

But whether these changes are correct or not (and I doubt that we shall ever firmly settle this), the questions here also turn the spotlight on to laughter—or more precisely, on to what difference thinking harder about laughter might make to our understanding of the text. For critics of these lines tend to fall back on a series of overconfident assumptions about the linguistic and social rules that governed Roman risus—and on all kinds of claims about what ridere and risus can (or must) mean. This is a place where we find many false certainties about Roman laughter on show.

So, for example, there is an alternative and less drastic emendation in line 62—which retains the idea that it is the risus of the baby but changes just one letter of the manuscript version. It replaces cui with qui but keeps the plural parentes found in the manuscripts, to read “qui non risere parentes.” Assuming that parentes is in the accusative case, this would mean “those who have not risere at their parents.” It is, at the very least, an economical solution, but it has often been rejected on the grounds that “rideo with the accusative can only mean ‘laugh at’ or ‘mock’” (and so would suggest, ludicrously, that the baby here was ridiculing his parents). In fact, that is simply false; as the most careful critics have conceded, there are numerous examples in Latin of ridere being used with an accusative object in an entirely favorable sense.55

From a different angle, many scholars have seized on Pliny’s statement that human children do not laugh until they are forty days old—except for Zoroaster, who laughed (risisse) from the moment he was born. In this way, they argue, through his hints at supernaturally precocious laughter, Virgil is claiming divine status for the child. Maybe. But the fact is, we have no idea how old Virgil’s baby is meant to be, we have no idea how widespread in the Roman world Pliny’s factoid about the chronology of laughter was, nor does the closest parallel passage (as we shall shortly see) provide any justification for that religious interpretation.56 There have also been firm (and conflicting) views expressed on whose risus is meant earlier, in line 60 (risu cognoscere matrem, or “to recognize your mother with risus”). Must this be the risus of the baby, in recognition of his mother? Or could it be her risus, which allows the baby to recognize her?57 The Latin is, of course, consistent with either (or indeed both simultaneously).

Perhaps more important, though, underlying almost all recent interpretations of these lines we can detect a decidedly sentimental tinge. Even one of the most hardheaded Latinists, Robin Nisbet, suggests that the scene’s “humanity” (whatever he means by that) is a good indication that “a real baby is meant” rather than some abstract symbol of peace and prosperity, and some critics, even when they are not arguing for a prophetically Christian reading of the text, evoke a scene that is frankly closer to an image of the adoring Virgin Mary and baby Jesus than to anything we know from pagan Rome.58 This sometimes chocolate-box tone is underpinned by what has become the standard translation of risus and ridere here, “smile” rather than “laugh”: “Begin, little boy, to recognize your mother with a smile.”59 It conjures up a picture of the loving smiles that bind mother and son and resonate powerfully in our understanding of babies and parenthood. How misleading is this?

So far I have avoided this issue, by keeping largely to the Latin terms. But not only should “smile” never be the translation of first resort for ridere; in this case there is also a clear suggestion in one of Virgil’s closest predecessors for this scene that a vocalized laugh is definitely meant. Virgil most likely drew and adapted this scene from Catullus, who in his wedding hymn for Manlius Torquatus imagines the future appearance of Torquatus junior, a baby sitting on his mother’s lap, stretching out his hands to his father, and “sweetly laughing to him with his little lips half open” (dulce rideat ad patrem / semihiante labello).60 This is not the curved lips of a silent smile; it is a laugh, and that is what we should think of in the Virgilian scene too.

It is perhaps easier for those not so embedded in the traditions of Virgilian scholarship to see the wider possibilities here, and their different perspectives can be instructive. For modern theorists of literature and psychoanalysis who have reflected on the role of laughter as a metaphor of communication, this passage has had a particular importance, even if it has rarely been discussed at length. Georges Bataille, for example, referenced Virgil’s words in a famous essay on the subject. “Laughter,” he wrote, “is reducible, in general, to the laugh of recognition in the child—which the following line from Virgil calls to mind.”61 Julia Kristeva, likewise, hinted at the scene described by Virgil when she theorized the crucial role of laughter in the relationship between mother and baby and in the baby’s growing sense of its own “self.”62 These ideas found an echo in the work of the cultural critic Marina Warner, who commented directly on the final lines of Eclogue 4 in the course of a more general discussion of (in her words) “funniness.” She had no difficulty in translating Virgil’s ridere as “laugh” and in seeing a point to that laughter: “‘Learn, little boy, to know your mother through laughter.’ Did he [Virgil] mean the child’s laughter? Or the mother’s? Or, by omitting the possessive, did he want his readers to understand that recognition and laughter happen together at the very start of understanding, identity, and life itself?”63

This is a radically different type of reading from those I have just reviewed. Many classicists would, I suspect, be reluctant to follow Warner, still less Bataille or Kristeva, and this is not the place for a lengthy discussion on the strengths and weaknesses of their arguments.64 But at the very least, in interpreting this contested passage so differently and in their conviction that we are dealing with vocal laughter, they offer a powerful reminder of how dangerous it is to assume that we know how Latin risus worked—let alone to impose some version of “baby’s first smile” on the culture of ancient Rome.


Roman laughter was not, however, merely laughter in Latin. So far in this chapter I have focused on Latin literature, but already by the second century BCE, Rome had a bilingual literary culture, in which laughter could be debated and discussed in both Latin and Greek.

In fact, both incidents of Roman laughter that I chose to discuss in the first chapter of this book are classic examples of this kind of linguistic and literary bilingualism. The first (pp. 1–8) describes an incident that took place in the Colosseum at Rome, in a fearful and funny standoff between the emperor Commodus and a group of the Roman political elite; it was taken from a history of Rome written in Greek by a Roman senator whose original home was in the Greek-speaking province of Bithynia, in what is now Turkey. The second (pp. 8–14) was taken from a Latin comedy originally performed in the second century BCE at (almost certainly) a religious festival in the city of Rome. But—in a form of literary syncretism long debated by scholars of Greco-Roman comedy—it was in fact a Romanized adaptation and conflation of two plays by the late fourth-century Athenian dramatist Menander. Neither of these survives beyond some fragmentary snatches recovered from Egyptian papyrus and excerpts quoted by later authors, but, from even the few passages we have, it is clear that some of the funny lines I discussed earlier go back, with adjustments, to one of Menander’s plays.

The question is not whether these two stories deserve their place in an exploration of Roman laughter. Of course they do: each in its different way unfolds within a Roman institutional framework, and each is told by a “Roman” writer (Dio a Roman senator, Terence probably an enfranchised ex-slave). But they raise the question of where we might want to draw the line. There is in particular a vast amount of surviving literature written in Greek in the period of the Roman Empire, when the Greek world was under Roman political and military control—from the satires of Lucian to the lectures of Dio Chrysostom and the boy-gets-girl novel (Leucippe and Cleitophon) by Achilles Tatius, not to mention the biographies and philosophy of Plutarch, the histories of Dio and Appian and Dionysius, or the wearisome hypochondria of Aelius Aristides and the interminable (fascinating to some) medical treatises of Galen. Does it all count as Roman? Does “Roman” laughter potentially include the laughter of the whole Roman Empire, from Spain to Syria? What is the difference between Greek and Roman laughter? I have already pointed to some mismatches in the vocabulary of laughing and jesting in the Latin and Greek languages. How far does that indicate significant cultural differences that we should be taking into account?

These reflections gesture toward a lively, wider debate among historians and archaeologists about the very nature of “Roman” culture. Complex as this debate has become, one simple question largely sums it up: what do we mean by that superficially unproblematic adjective Roman (whether “Roman laughter” or “literature,” “sculpture” or “spectacle,” “politics” or “pantomime”)? Which Romans are we talking about? The wealthy literate elite? Or the poor, the peasants, the slaves, or the women? And even more to the point, are we thinking of the term geographically, chronologically, or more integrally linked to political and civic status or to distinctive norms of behavior and culture? Can, for example, an intellectual treatise written in Greek by an Athenian aristocrat in the second century CE count as Roman because Athens was then part of the Roman Empire? Would it be more convincingly Roman if the Greek writer was (like Dio) simultaneously a Roman senator or if we knew that the work was read and debated by Latin speakers in Rome itself?

There are, of course, no right answers to these questions. The most influential recent studies have insisted on disaggregating any unitary notion of “Roman” culture while also arguing against any simple progressive model of cultural change across the ancient Mediterranean.65 No one would now think of the early city of Rome as a cultural vacuum that was gradually filled, in a process neatly labeled “Hellenization,” thanks to its contacts with the Greek world. (The Roman poet Horace would, I suspect, have been horrified to discover that his words “Captured Greece took captive its rough conqueror” would be dragged out of context and turned into a slogan for the simple inferiority of Roman versus Greek culture.66) Likewise, few historians would now characterize growing Roman influence in the West as a straightforward process of “Romanization”—or, alternatively, think in terms of a clear standoff between “Roman” cultural forms and those of the more or less resistant “natives.”

Instead they point to a shifting cross-cultural multiplicity of “Romannesses,” formed by an often unstable series of cultural interactions summed up in a range of sometimes illuminating, sometimes overseductive, sometimes (I fear) quite misleading metaphors, such as constellation, hybridity, creolization, bilingualism, or crossbreeding.67 In fact, in some of the most radical work, even the basic descriptive language of ancient cultural difference and ancient cultural change in the Roman Empire seems to have been turned inside out and upside down. So, for example, in Andrew Wallace-Hadrill’s wonderfully heady study Rome’s Cultural Revolution, the very opposition between Roman and Greek (Hellenic) culture is drastically subverted. That is to say, Wallace-Hadrill offers a series of powerful arguments for seeing Rome as a prime engine of “Hellenization,” “Hellenization” as one aspect of “Romanization,” and ultimately “Roman” influence as a driver behind the “re-Hellenization” of the Hellenic world itself!68

These vertiginous issues inevitably lurk in the background of any book such as this one. But my most pressing questions are rather narrower and more manageable. For a start, we have to face the fact that we have almost no access whatsoever to the culture of laughter among the nonelite anywhere in the Roman world. Whether the style of “peasant laughter” really was as different from that of the urban elite as we often imagine, who knows? (We shouldn’t forget that the supposed lustiness of the peasant can be as much an invention of the sophisticated city dweller as an accurate reflection of the gelastic life of simple peasant society.)69 In any case, to study “Roman laughter” is now necessarily to study laughter as it is (re)constructed and mediated in a range of elite literary texts. The question is: which ones, and particularly which ones of those produced in Greek or partly rooted in the Greek world? Is there a line to be drawn? Where? Does Plutarch—Greek essayist, priest at the sanctuary of Delphi, and avid student of “Roman” culture—belong in this book, in Stephen Halliwell’s Greek Laughter, or in both? Are we in danger of confusing “Greek” with “Roman” laughter? And how much does it matter?

There can be no hard-and-fast rules. Recent critical approaches to the Greek culture of the Roman Empire have stressed many different, sometimes contradictory, aspects: its emphatically Hellenic (even “anti-Roman”) coordinates, its active role in the reformulation of the very categories of “Greek” and “Roman” or in supporting the political and social hegemony of Rome over Greece, and so on.70 In practice, the modern dividing line between “Greek” and “Roman” has sometimes come down to little more than subject matter (if the work in question is about Rome, it tends to be treated as Roman; if about Greece, then it’s seen as Greek—despite the fact that the bifocal, Greco-Roman perspective of Plutarch and others makes nonsense of that procedure). Perhaps even more often, to be honest, it comes down to the territorial divisions of the modern academy. On the one hand, scholars of classical Greek literature tend to embrace and interpret this material as somehow an extension of their territory (it is, after all, written in “their” language and constructively engages with its classical Greek predecessors). Many Roman cultural historians, on the other hand, would claim it as part of their remit (it was written in “their” period and often gestures directly or indirectly to the power structures of the Roman empire). The truth is, there is no safe path to be trodden between seeing this literature in terms of (on the one hand) being Greek or (on the other)becoming Roman—to conscript the titles of two of the most influential modern contributions to this whole debate.71

I shall proceed with some very basic methodological guidelines in mind. First, that the “Greek” and “Roman” cultures of laughter in the period of the Roman Empire were simultaneously both foreign to each other and also so mutually implicated as to be impossible to separate. Simply by virtue of language, some sense of cultural difference could always be mobilized. We have to imagine, for example, that when Virgil had his text of Homer in front of him and was considering how he would reflect the Greek wordmeidiaō in his own epic (see p. 73), he necessarily pondered on the different senses of Greek and Latin words for laughter and what might hang on them. And we caught a glimpse (on p. 78) of paraded ethnic preferences in joking among the elite diners at Macrobius’ Saturnalian dinner party: Greek, Egyptian, and Roman. We certainly need to keep alert for hints of cultural difference. But for the most part, there is little to be gained (and much to be lost) by attempting to prize apart the gelastic culture of imperial literature, still less by distributing these culturally multifaceted texts on one side or the other of some notional “Roman”/“Greek” divide (Plutarch’s Roman Questions in, Leucippe and Cleitophon out; Apuleius’ Latin version of the story of “Lucius the Ass” in, the parallel Greek version out). Elite Romans, wherever in the empire they lived, learned to “think laughter” in debate with both Greek and Latin texts. We are dealing, in large part at least, with a shared literary culture of laughter and “laughterhood,” a bilingual cultural conversation.

My second guideline serves to limit that very slightly. If we do imagine Roman imperial culture as a conversation (to add, I confess, yet another metaphor to those of hybridity, constellation, and the rest), I have chosen to concentrate on those literary works written in Greek where we can most confidently point to an explicitly Roman side in that script, rather than merely a generalized sociopolitical Roman background. That is sometimes through characters clearly labeled as Roman being featured in a dialogue (as we find, for example, in Plutarch’s Table Talk) or through specifically Roman subject matter and context (such as the names, currency, and events that form part of the background to the gags in the late antique “jokebook” the Philogelos, or “Laughter lover”).

What is striking is how powerful the Roman intervention in that conversation can be. In fact, as we shall now see, some of the traditions of laughter that may appear superficially to be more or less pure “Greek” turn out to be much more “Roman” than we usually assume. Sometimes we find that what we take as notable traditions of classical Greek laughter are very largely constructions of the Roman period. Occasionally we find that the Greek idiom of laughter adapts to ideas and expressions that are distinctively Latin. And when—conversely—Roman authors take over Greek jokes, we have evidence for the creative adaptation of the original material for a Roman audience. Here again, Terence’s Eunuch—with Gnatho the sponger, Thraso the soldier, and the joke about the young Rhodian—offers a nice glimpse of the “Romanization” of Greek laughter and the archaeology of a Roman joke while introducing some of the bigger issues of the final section of this chapter.


The comedies of Plautus and Terence have long provided revealing instances of the intricacy of Roman engagement with Greek culture—and the philological work of Eduard Fraenkel in the 1920s underpins many discussions of this.72 The plays are explicitly drawn from Greek models, but the dramatists actively reworked the “originals” into something significantly different, with a new resonance in the Roman context. For example, whatever its Greek source (which is still debated), Plautus’ Amphitruo closely engages with that most distinctive of all Roman celebrations: the triumphal procession, held in honor of military victory. Plautus in fact comes close to adapting whatever his (Greek) original was into a comic parody of the origins of the (Roman) triumph.73

In Terence’s Eunuch, this creative adjustment goes right down to the individual jokes, so adding a further twist to the scenes of laughter that I looked at in the first chapter—and an important coda to my treatment there. The prologue of the play states clearly that its models were two late fourth-century plays of Menander: The Eunuch and The Toady (Kolax), from which the characters of the soldier and the sponger/flatterer (or toady) were drawn. We have, from various papyrus scraps and quotations, more than a hundred lines of The Toady, and these confirm that the characters of Gnatho and Thraso went back to that source (even if they were known by different names in Menander’s play).74 In fact, a brief snatch of dialogue, quoted by Plutarch, seems likely to have been the inspiration for one of the exchanges between the two that I quoted in chapter 1—a classic example of a willfully misleading explanation for an outburst of laughter. This, as we saw (p. 11), is Terence’s version:

Gnatho: hahahae

Thraso: What are you laughing at?

Gnatho: At what you just said, and at that story about the guy from Rhodes—whenever I think about it.

And this, to judge from Plutarch (who is discussing the problems of dealing with flatterers), is the “original” passage in The Toady, which Terence took over. The sentiment is strikingly similar, and the words are attributed to the sponger/flatterer of the title:

    I’m laughing when I think about the joke

You made against the Cypriot.75

Whether that explanation for laughter was as wickedly misleading in Menander’s play as in Terence’s, we do not have enough information to say (though Plutarch’s claim that the toady was “dancing in triumph” over the soldier with these words suggests that it was). But one thing seems certain: in each play there was some comic reference back to an earlier joke—yet the exact terms of that joke were different. In The Eunuch, it was a joke about the Rhodian boy (“chasing after delicacies”). In Menander, it is some (lost) gag about a “Cypriot”—perhaps, as some critics have proposed, connected with the old Greek saying about Cypriot bullocks eating dung (so all Cypriots are “shit eaters”).76

If so, we can only guess what lay behind Terence’s change. Perhaps the Cypriot bullock joke was simply not part of the Roman repertoire and was likely to fall flat in front of Terence’s first audience. Perhaps he entirely rewrote the joke to make a topical allusion to Roman political relations with Rhodes. But maybe Terence changed only the nationality of the quip’s antihero (the boy chasing the delicacies), from Cypriot to Rhodian; after all, in his Eunuch, the desired girl came from Rhodes, and maybe there was an intentional link. If so, that would give a deeper resonance, for the more learned members of the Roman audience, to the idea that it was an old joke (see p. 13). In fact, it was so old that it went back not just to Livius Andronicus but (plus or minus the Cypriot–Rhodian switch) to the age of Menander in the fourth century BCE. Here, in other words, the Greek inheritance was not merely adjusted to a different comic context; it was turned into an integral part of the Roman joke itself.


Classicists have long tussled with the ways that Roman writers reinvigorate (or recycle) their Greek predecessors, pointing to a characteristic combination of similarity and difference found throughout Roman (re)-use of Greek cultural forms, right down to the laughs. But they more rarely look at the relationship from the other side. To conclude this chapter, and to think more about potential “Roman” aspects of “Greek” laughter, I am taking a cue from Andrew Wallace-Hadrill and from Tony Spawforth, who have both argued for a wide-ranging cultural impact of Rome on the Greek world (from the style of lamps made in Roman Athens to the “cultural comportment” of the imperial Greek elite).77 Some of the traditions often assumed to be those of classical Greece owe a lot in various ways to the cultural conversations of the (Greco-)Roman Empire.

One of the most memorable symbols of Greek laughter is the fifth-century BCE philosopher Democritus, from the northern Greek city of Abdera—who has gone down in history as “the laughing philosopher,” celebrated in that role not only in antiquity but also by modern artists and writers as diverse as Peter Paul Rubens and Samuel Beckett. Often paired with Heraclitus (his opposite—“the weeping philosopher”), Democritus crops us time and again in ancient writing in his iconic role as “the laugher” (or as the “laughter expert”).78 When, for example, Cicero is settling down in On the Orator to a discussion of the role of laughter in oratory and wants to brush aside the impossible question of what laughter actually is, he writes, “We can leave that to Democritus”;79 others tell how Democritus’ mockery of his fellow countrymen gave him the nickname Laughing Mouth or made him, as Stephen Halliwell has put it, the “patron saint” of satiric wit (“Democritus used to shake his sides in perpetual laughter,” wrote Juvenal, even though there was much less in his day to provoke ridicule—no flummery, no togas with purple stripes or sedan chairs).80

But by far the richest account of Democritus’ laughter is found in what is, in effect, an epistolary novella comprising a series of fictional letters written in Greek, exchanged between the citizens of Abdera and the legendary Greek doctor Hippocrates—now preserved among the writings associated with Hippocrates (spuriously, in the sense that almost certainly none are from his own hand).81 In this story, the Abderites (who have their own cameo part to play in the history of laughing and joking, as we shall see in chapter 8) are increasingly concerned about the sanity of their famous philosopher, for the simple reason that he was always laughing, and at the most inappropriate things. “Someone marries, a man goes on a trading venture, a man gives a public speech, another takes an office, goes on an embassy, votes, is ill, is wounded, dies. He laughs at every one of them,”82 they write in their exasperation to Hippocrates, asking him to come to Abdera to cure Democritus. The doctor agrees (and the novella includes some comic touches among the preparations—from transportation to arrangements for his wife during his absence). But as we learn from the letters, when he encounters the patient, he soon discovers that Democritus is not mad at all: he is rightly laughing at the folly of humanity (“You think there are two causes of my laughter—good things and bad things. But I laugh at one thing—mankind”83).

En route to this (happy) conclusion, there is plenty of opportunity for the various parties to offer their views of what laughter is for. In fact, the novella is one of the most extended philosophical treatments of laughter to survive from the ancient world. But what I want to underline here is that there is no evidence whatsoever for any particular association between Democritus and laughter before the Roman period. The earliest reference we have to this connection is that casual aside in Cicero, while the Hippocratic novella is almost certainly to be dated to the first century CE, several centuries after the deaths of both of its protagonists.84 Democritus’ own writing, so far as we can reconstruct it, was principally concerned with theories of atomism and a much more moderate ethical stance than the “absurdist” position that the novella implies. How or why he had been resymbolized by the first century CE in these very different terms, we can only conjecture.

We find a broadly similar pattern in another significant symbol of Greek laughter—that is, the tradition of distinctively “Spartan” laughter. Sparta is the only city in the ancient world, outside the realm of fiction (see pp. 181–83), where there was said to have been a statue, even a shrine and a religious cult, of Laughter; it was attributed to the mythical lawgiver Lycurgus.85 Moreover, the boot-camp atmosphere of classical Sparta is supposed to have included a prominent role for laughing and jesting. The young Spartiates were said to learn both to jest and to endure jesting in their “common messes” (sussitia), and the Spartan women were supposed to ridicule those young men who failed to meet the standards of the training system.86 The surviving references to Spartan quips and witticisms emphasize their down-to-earth frankness, even aggression (such as the retort of the lame Spartan fighter who was laughed at by his peers: “Idiots, you don’t need to run away when you fight the enemy”87). Tempting as it may be to use this evidence to fill in some of the many gaps in what we know of classical (fifth- and fourth-century BCE) Spartan culture,88 the fact is that it all comes from writers of Roman date—principally, but not only, Plutarch. It must in part reflect a nostalgic construction of Spartan “exceptionalism,” with these supposed “primitive” traditions of laughter being used, retrospectively, to mark out the oddity of the Spartan system.89

Of course, in both these cases we should be careful not to overclaim. We would get a very odd view of ancient history if we assumed that no traditions existed before the first surviving reference to them (“absence of evidence is not evidence of absence,” as the old inferential cliché goes). It would be implausible to imagine that, in his casual aside, Cicero invented Democritus’ connection with laughter; much more likely he was referring (with what degree of knowledge is not clear) to some preexisting commonplace. On the evidence we have, it is impossible to be certain exactly when the popular metamorphosis of Democritus—from atomist to laugher—took place.90 There is certainly a deeper prehistory to the traditions of Spartan laughter too: Plutarch, in fact, cites a third-century BCE source for the “shrine of Laughter,” and many of those anecdotal quips attributed to famous Spartans of the past may well have had an even earlier origin.91 Yet the fact remains that—selected, adjusted, and embellished as they must have been—the traditions about Democritus and the Spartans have come down to us in the literature of the Roman Empire. In a scholarly world in which historians have tried to push so many traditions back to the glory days of classical Greece, it is important to remember that many of the details, the interrelationships, the cultural nuances (even if not the entire traditions themselves) are the product of the Greco-Roman imperial world.

One final example gives us a nice glimpse of the two-way traffic in “laughter culture”—not only from Greece to Rome but also from Rome to Greece. One of the slogans of British eighteenth-century urbanity was “Attic salt”—the traditions of elegant wit particularly associated with ancient Athens. The same Lord Chesterfield who so disdained “audible laughter” was a tremendous advocate of this particular style of jest, as he wrote to his long-suffering son: “That same Attic salt seasoned almost all Greece, except Boeotia; and a great deal of it was exported afterward to Rome, where it was counterfeited by a composition called Urbanity, which in some time was brought to very near the perfection of the original Attic salt. The more you are powdered with these two kinds of salt, the better you will keep, and the more you will be relished.”92 Poor Lord Chesterfield could not have been more wrong in his chronology, or in suggesting the transmission of “Attic salt” from Greece to Rome. It is true that Roman writers admired Athenian wit: they saw it as a form to be imitated, and in their cultural geography of wit they put the Athenians in prize position, followed by the Sicilians and then the Rhodians.93 But so far as we can tell, the idea of wit as salt (sal) was originally a Roman idea, defined in Latin and part of a range of Roman cultural tropes that (as we shall see) linked jesting and laughing to the sphere of dining and the repertoire of cooking. “Attic salt” was not a Greek term, but it was the Romans’ way of describing their own construction of Athenian wit.

No Athenians, so far as we know, ever congratulated themselves on their “Attic salt.” In classical Greece, the word hals (salt) was not part of the terminology of jesting. Eventually, however, the idea did spread eastward. Some Greeks of the Roman period apparently adopted, incorporated, and maybe adjusted this characteristically “Roman” perspective on laughter. In the second century CE, we find Plutarch referring to the wit of Aristophanes and Menander as hales—their “little pinches of salt.”94 We should make sure not to underestimate the Roman aspects of that often inextricable mixture that is the Greco-Roman culture of laughter.

• • •

It is to various aspects of that inextricable mixture that we now turn. The issues that I have been discussing in these first four chapters underlie the explorations in the second part of this book of particular aspects of Roman laughter and of some of the distinctive characters who have a particular role to play in the “laughterhood” of Rome. We shall encounter laughing emperors, plenty of monkey business, and some passable jokes—but first the funniest man in the Roman world, Marcus Tullius Cicero, and some of his fellow orators. There have been several excellent studies of uses of wit and laughter in the Roman courtroom, but I shall focus on the dilemmas confronting the joking orator trying to raise a laugh from his audience in order to expose some of the ambiguities and anxieties of the culture of laughter in ancient Rome.

FIGURE 1. Frans Hals, The Laughing Cavalier (1624). This painting—which we now take for granted as an image of a laughing man—raises the question of how confidently we can identify laughter in the art of the past.

FIGURE 2. Mosaic—“Beware of the dog”—from the House of the Tragic Poet, Pompeii (first century CE). How can we decide if this image was intended to make visitors laugh?

FIGURE 3. Bronze statuette of an actor with an ape’s head (Roman date). This nicely symbolizes the overlap between the mimicry of actor and of monkey.

FIGURE 4. A boy with a performing monkey, from an original painting (first century CE) in the House of the Dioscuri, Pompeii. The ape becomes an actor.

FIGURE 5. Parody of Aeneas, escaping from Troy, with his father and son—with ape heads (from an original painting, first century CE, from Pompeii).

FIGURE 6. Rembrandt’s self-portrait as Zeuxis (c. 1668). Notice the painting of the old lady in the background.

If you find an error or have any questions, please email us at Thank you!