Texts and Abbreviations

In the notes, I have followed the conventions of L’Année philologique for abbreviating the titles of periodicals. For titles of ancient works, I have given a fairly full version or used the abbreviations of the Oxford Classical Dictionary (third edition). In a few cases where it is standard practice and there can be no confusion (e.g., Catullus or Livy), my references omit the title entirely. All translations, unless otherwise indicated, are my own. I have used standard editions of ancient texts—Oxford Classical Texts, Teubners, or recent Loebs—but have pointed to different manuscript readings where significant. For modern works with a potentially misleading discrepancy between the date of the edition I have cited and the date of first publication, I have indicated both, in this form: Hobbes 1996 [1651].

Other abbreviations are as follows:


L’Année épigraphique: Revue des publications épigraphiques relatives à l’antiquité romaine. Paris, 1888–.


Anthologia Latina, ed. A. Riese et al. Leipzig, 1894–1926.

Anec. Graeca

Anecdota Graeca, ed. I. Bekker. Berlin, 1814–21.


Anthologia Palatina, in The Greek Anthology, Loeb Classical Library, ed. W. R. Paton. London, 1916–18.


Corpus Glossariorum Latinorum, ed. G. Goetz et al. Leipzig, 1888–1923.


Corpus Inscriptionum Latinarum. Berlin, 1863–.


Die Fragmente der Vorsokratiker griechisch und deutsch, 11th ed., ed. H. Diels and W. Kranz. Zurich and Berlin, 1964.


Groningen Colloquia on the Novel. Groningen, 1988–.


Grammatici Latini, ed. H. Keil. Leipzig, 1855–80.


Inscriptions de Délos. Paris, 1923–.


Inscriptiones Latinae Selectae, ed. H. Dessau. Berlin, 1892–1916.

Jacoby, FGrHist

Die Fragmente der griechischen Historiker. Berlin, Leiden, 1923–.


A Latin Dictionary, ed. C. T. Lewis and C. Short. Oxford, 1879.


A Lexicon of Greek Personal Names, ed. P. M. Fraser et al. Oxford, 1987–.


Lexicon Iconographicum Mythologiae Classicae. Zurich, 1981–.

New Pauly

Brill’s New Pauly, ed. H. Cancik, H. Schneider, and M. Landfester, English trans. ed. C. Salazar and F. G. Gentry. Leiden, 2002–10.


Oxford Latin Dictionary, ed. P. Glare. Oxford, 1982 (rev. 2012).


Poetae Latini Minores, ed. A. Baehrens. Leipzig, 1879–83 (rev. F. Vollmer).


Oxyrhynchus Papyri. Egypt Exploration Society. London, 1898–.


Pompei, pitture e mosaici, ed. G. Pugliese Carratelli. Rome, 1990–99.

Rerum memorandarum Lib.

F. Petrarca, Rerum memorandarum Libri, ed. G. Billanovich. Florence, 1945.


Remains of Old Latin, Loeb Classical Library, ed. E. H. Warmington. London and Cambridge, MA, 1935–40.



1. The poem is titled “Invocation of Laughter” (1909): “. . . O laugh out laugheringly / O, belaughable laughterhood—the laughter of laughering laughers . . .” This translation is from www.russianpoetry.net, a project of Northwestern University’s Department of Slavic Languages and Literatures. It is also featured in Parvulescu 2010, 1–4.


1. Dio 73(72).18–21 gives a full account of these spectacles (20.2 notes the plans to fire into the crowd, in imitation of Hercules’ attack on the Stymphalian birds); Hopkins and Beard 2005, 106–18, describes the arrangement of the audience and conventions of the proceedings (including on this occasion).

2. Herodian 1.15.

3. Dio 73(72).21.

4. On his name, see Roxan 1985, no. 133; Gowing 1990. Dio was probably a few years under forty at the time, hence my young.

5. Dio 73(72).23 (the timetable of composition); Millar 1964, 1–40.

6. Dio 73(72).21.

7. Carter 1992, 190. This essay is a wonderful attempt to redefine the “giggle” as a mechanism of female power (rather than as the trivializing laughter of “girls” and a sign of their powerlessness). See further p. 157.

8. Anec. Graeca 1.271. The erotics of κιχλίζειν and its association with prostitutes are clear in the numerous examples collected in Halliwell 2008, 491. But it is a more complicated word (and sound) than is often acknowledged; see, for example, Herodas 7.123, which describes it as “louder than a horse”—hardly a “giggle” in our terms (despite the onomatopoeia). Jeffrey Henderson 1991, 147, points to other (erotic) associations.

9. The Greek insistently repeats the words: κἂν συχνοὶ παραχρῆμα ἐπ’ αὐτῷ γελάσαντες ἀπηλλάγησαν τῷ ξίφει (γέλως γὰρ ἡμᾶς ἀλλ’ ου λύπη ἔλαβεν), εἰ μὴ δάφνης φύλλα, ἃ ἐκ τοῦ στεφάνου εἶχον, αὐτός τε διέτραγον καὶ τοὺς ἄλλους τοὺς πλησίον μου καθημένους διατραγεῖν ἔπεισα, ἵν’ εν τῇ τοῦ στόματος συνεχεῖ κινήσει τὸν τοῦ γελᾶνἔλεγχον ἀποκρυψώμεθα (Dio 73[72].21.2). In alluding (with no details) to a story of laughter defying all attempts to restrain it, Aristotle (Eth. Nic. 7.7, 1150b11) writes of people “bursting out in a flood of laughter” (τὸν γέλωτα ἀθρόον ἐκκαγχάζουσιν).

10. Dio 9.39. Dionysius of Halicarnassus’ account of the same incident (Ant. Rom. 19.5) also features Tarentine laughter and shit, but it is the bad Greek, rather than the funny clothes, of the ambassadors that provokes the mirth. For a further example of Dio, as an eyewitness, using laughter as a response to the bathos of imperial power, see 74(73).16.

11. Despite the brave optimism of J. R. Clarke (2003; 2007, 109–32), who attempts to exploit visual images to access the world of “ordinary” people’s laughter; see further above, pp. 57–59.

12. Hopkins 1983, 17 (my italics).

13. Critchley 2005, 79.

14. It is hard to capture elegantly in English the potential slippage between something or someone who is laughable in the sense of “capable of raising a laugh” and something or someone who is laughable in the sense of “ridiculous.” Where it seems particularly important, I highlight the issue with a hyphen: laugh-able. The more pronounced ambiguity in the Latin ridiculus is discussed on pp. 102–3, 125.

15. τὴν δὲ κεφαλὴν τὴν ἑαυτοῦ σεσηρὼς ἐκίνησεν (Dio 73[72].21.2). The word is discussed by Halliwell 2008, 521, 533nn12–13.

16. Suetonius, Calig. 27; Seneca, De ira 2.33; discussed on p. 134.

17. These paragraphs touch on a view of laughter commonly associated with Mikhail Bakhtin; see further pp. 59–62. Critchley 2005 offers a brisk critique of Bakhtin, on which I draw here, and, in so doing, usefully headlines Slavoj Žižek’s critique of Umberto Eco’s The Name of the Rose (with its strident claims that totalitarianism offers no place for laughter) and Žižek’s (semiserious) arguments that Eastern-bloc totalitarianism was anyway itself always “a joke”; see especially Žižek 1989, 28–30. Semiserious or not, Žižek encourages us to think of a much more diverse engagement between laughter and political power.

18. A wall painting from the Villa San Marco at Stabiae captures this scene (Barbet and Miniero 1999, vol. 1, 211–12; vol. 2, plate 12.4), and Dio’s reference to Hercules and the Stymphalian birds (see above, n. 1) suggests that the emperor’s antics were seen in mythic terms. But maybe we should not press this too far; the truth is that the canonical image of Perseus with the head of Medusa held high in one hand and sword in the other is in large part a creation of the Renaissance (with Benvenuto Cellini’s statue from the Piazza della Signoria in Florence a key inspiration).

19. For example, Hopkins 1983, 16–17; Dunkle 2008, 241.

20. We should be alert to (at least) two senses of the English phrase laugh at. In the weaker sense, “What are you laughing at?” is more or less synonymous with “Why are you laughing?” (“I’m laughing at the jokes”). In the stronger sense, it represents something more aggressive (“I’m laughing at Commodus”). This is not unlike the range of the Latin “Quid rides?” (as in the passage of Terence discussed on pp. 11, 14).

21. For Romans laughing at the bald, see pp. 51, 132–33, 146.

22. The complexities of Dio’s account are well noted by Hekster 2002, 154–55.

23. The precise details of the history of Roman games (ludi) and the development of theatrical performances within them are complex, and in part obscure; see F. H. Bernstein 1998; F. H. Bernstein 2011; Beard, North, and Price 1998, vol. 1, 40–41, 66–67; vol. 2, 137–44. Manuwald 2011, 41–55, reviews the festival contexts of theatrical performances.

24. Beacham 1991, 56—85 (on stages and staging); Manuwald 2011, 55–68 (Temple of the Great Mother, 57); Goldberg 1998 (specifically on the Temple of the Great Mother and comic performances of the second century BCE).

25. Hunter 1985 is a sane introduction; Marshall 2006 includes an up-to-date discussion of masks (126–58); with Manuwald 2011, 79–80. For masks, or not, in mime, see above, p. 168.

26. We rely here on the possibly unreliable account of Suetonius, Poet., Terence 2 (and we must assume that the “repeat performance” refers to the first production).

27. Barsby 1999 and Brothers 2000 are helpful discussions of the play as a whole.

28. Another manuscript version of the didascalia ascribes the first performance to the Ludi Romani (Barsby 1999, 78)—which would (sadly) rule out any direct connection between the representation of the eunuch in the play and the original performance context. The cult of Magna Mater was a complex amalgam, parading both Roman and disconcertingly foreign elements (such as castration); on these representational and other complexities, see Beard 1996.

29. Gnatho himself had already paraded that insincerity a couple of hundred lines earlier (249–50), in a double entendre on his life as a sponger, discussed on pp. 71–72.

30. My translation of this line (“Dolet dictum inprudenti adulescenti et libero,” 430) follows Donatus’ commentary and those more recent critics and translators (such as Barsby [1999, 164]) who see Gnatho flattering Thraso, by offering (mock) sympathy for the young Rhodian.

31. TH. una in convivio / erat hic, quem dico, Rhodius adulescentulus. / forte habui scortum: coepit ad id adludere / et me inridere. “quid ais” inquam homini “inpudens? / lepu’ tute’s, pulpamentum quaeris?” GN. hahahae. TH. quid est? GN. facete lepide laute nil supra. / tuomne, obsecro te, hoc dictum erat? vetu’ credidi. TH. audieras? GN.saepe, et fertur in primis. TH. meumst. GN. dolet dictum inprudenti adulescenti et libero. PA. at te di perdant! GN. quid ille quaeso? TH. perditus: / risu omnes qui aderant emoriri. denique / metuebant omnes iam me. GN. haud iniuria.

32. TH. ego hinc abeo: tu istanc opperire. PA. haud convenit / una ire cum amica imperatorem in via. TH. quid tibi ego multa dicam? domini similis es. GN. hahahae. TH. quid rides? GN. istuc quod dixti modo; / et illud de Rhodio dictum quom in mentem venit.

33. Donatus on Eun. 426; see also Eugraphius on Eun. 497.

34. GLK 6.447.7 (Marius Plotius Sacerdos); see also 1.419.7 (Diomedes, “hahahe”), 3.91.3–4 (Priscian, “ha ha hae”), 4.255.31 ([Probus], “hahahae”), 6.204.23 (Maximus Victorinus, “haha”). The minor textual variants in the manuscript tradition do not alter the main point (or sound). The recognition of laughter sounds in Greek texts is complicated by the fact that the simple substitution of a smooth for a rough, aspirated breathing turns a ha ha ha into an ah ah ah! Possible instances of laughter scripted in Greek comedy are discussed (and largely rejected) by Kidd 2011, with full reference to earlier bibliography, back to late antique and medieval critics who saw the problems that the presence or absence of aspiration caused.

35. One enterprising seventeenth-century systematizer, “un astrologue Italien, nommé l’Abbé Damascene,” attempted to classify the variants in these sounds and relate them to the different temperaments, hi hi hi indicating melancholics, he he he cholerics, ha ha ha phlegmatics, and ho ho ho hotheads; cited in Dictionnaire universel françois et latin, vol. 5 (Paris, 1743), 1081. Kidd 2011, while acknowledging some version of ha ha ha as a possible means of representing laughter in Greek, points also to such variants as αἰβοιβοῖ and ἰηῦ.

36. From Johnson’s Life of Cowley, first published in a collected edition of 1779–81 (see now, conveniently, Lonsdale 2009, 33); it is an exaggeration because Johnson is referring to not only the sound but also the cause of laughter (a universalizing claim that this book will dispute).

37. Fraenkel 1922, 43–45 (2007, 32–35) offers the most significant variant interpretation—“You are a hare: you go after tasty food” (or in its weaker form “Du suchst dir pulpamentum wie ein Hase,” “You look for pulpamentum like a hare”)—which Fantham 1972, 80, follows but Wright 1974, 25–27, convincingly rejects.

38. Barsby 1999, 163. I stress “Donatus’ text,” as the version of his commentary that has come down to us is a very mixed tradition, including Donatus’ own discussion and his compendium of earlier scholarship on the play as well as later additions and glosses incorporated in the process of transmission (Barsby 2000; Victor 2013, 353–58).

39. “Vel quod a physicis dicatur incerti sexus esse,” Donatus, Eun. 426. Frangoulidis 1994 shows more generally how the themes of the exchanges between Thraso and Gnatho look forward to later scenes in the play.

40. Cicero, De or. 2.217; see p. 28.

41. Freud 1960 [1905]; his analysis in terms of “displacement” (86–93) seems particularly relevant here. The idea of incongruity is characteristic of (among others) the “General Theory of Verbal Humor” (GTVH), as developed in Attardo and Raskin 1991 and Attardo 1994. They stress, in a much more nuanced way than my crude summary suggests, how the sequence of interpretative dilemmas and their resolution construct a joke.

42. On Freud and the physicality of laughter, see pp. 38–39, 40.

43. SHA, Carus, Carinus, Numerianus 13.3–5.

44. For possibly older Greek antecedents, see pp. 90–91.

45. See p. 4.

46. Festus, p. 228L; Diogenes Laertius 7.185. See pp. 176–78 for further examples and discussion.

47. Interestingly, Donatus (Eun. 497) sees Thraso’s question (“What are you laughing at?”) and the whole exchange in terms of the soldier’s desire to elicit flattery for his wit from the sponger (as at 427). Although the commentary reflects on the point of Parmeno’s joke and its exaggeration of Thraso’s status (495), it does not canvas this as a possible prompt for Gnatho’s hahahae.

48. Goldhill 2006 discusses these issues well; Bakhtin 1986, 135, by contrast, claims (at least in relation to carnival laughter) that “laughter only unites” (see further pp. 60–62). Billig’s stress on laughter and “unlaughter” (2005, 175–99) is also useful here.

49. Sharrock 2009, 163–249, discusses other aspects of “tired old jokes” (with a nice analysis of this particular exchange at 164–65). In general, recent discussions of laughter, ancient or modern, have tended to underplay its learned, practiced, or habitual aspects.

50. This idea of the self-reflexivity of laughter is a major theme throughout Halliwell 2008.

51. I have taken all these examples from good recent translations of The Eunuch: Radice 1976, Brothers 2000, and Barsby 2001. A particularly rich selection of laughter insertions (from with a smile to digging him in the ribs) can be found in the Loeb translation of Plautus, Nixon 1916–38.

52. A vague “dozen or so” because emendation can add to the total: Plautus, Poen. 768, Pseud. 946, 1052, Truculentus 209, and conjectured at Mil. 1073; Terence, An. 754, Haut. 886, Hec. 862, Phorm. 411, as well as Eun. 426, 497. The fragment of Ennius is quoted by Varro, Ling. 7.93 (= Ennius, frag. 370 Jocelyn; ROL1, Ennius, unassigned fragments 399); the mention of a shield has encouraged the (unnecessary) assumption that the original context was tragic. I have not included in my total here nine instances of scripted laughter (hahahe) in the Querolus, an anonymous version of Plautus’ Aulularia probably composed in the early fifth century CE, nor the glosses of grammarians. But they would not point to any significantly different conclusion.

53. Other instances imply other emotions: for example, disbelief at Plautus, Pseud. 946, or relief at Truculentus 209—which, together with Pseud. 1052, encouraged Enk (1953, vol. 2, 57–58) and others to reinterpret the (ha)hahae as merely an exhalation, the Latin equivalent of “phew,” a classic scholarly attempt to normalize Roman laughter.

54. This was widely reported in the British media: e.g., the Daily Mail (www.dailymail.co.uk/news/article-1085403/Jim-Bowen-brings-worlds-oldest-jokebook-London-stage—reveals-ancestor-Monty-Pythons-Dead-Parrot.html) and the BBC (http://news.bbc.co.uk/2/hi/7725079.stm).


1. De or. 2.235; the words are put in the mouth of the lead character in this part of the dialogue, Julius Caesar Strabo. I am lightly paraphrasing “Strabo” ’s list of questions: “Quid sit ipse risus, quo pacto concitetur, ubi sit, quo modo exsistat atque ita repente erumpat, ut eum cupientes tenere nequeamus, et quo modo simul latera, os, venas, oculos, vultum occupet?” (The text is uncertain: did Cicero imagine laughter taking over the blood vessels, venas, or cheeks, genas? See p. 116.) Quintilian (Inst. 6.3.7) follows Cicero’s disavowal: “I do not think the origin of laughter has been satisfactorily explained by anyone—though many have tried” (“Neque enim ab ullo satis explicari puto, licet multi temptaverint, unde risus”). For Cicero the jokester, see pp. 100–105.

2. De motibus dubiis 4 (erections), 10.4–5 (laughter), with Nutton 2011, 349.

3. Pliny, HN, praef. 17, proclaims the array of facts; for his encyclopedic project in general, see Murphy 2004; Doody 2010.

4. 7.2, 7.72. See pp. 35, 83–84.

5. 11.198.

6. 11.205 (“sunt qui putent adimi simul risum homini intemperantiamque eius constare lienis magnitudine”). Pliny may be referring to removal (as he notes here that an animal can continue to live if its spleen is removed because of a wound), but elsewhere (23.27) he refers to drugs that reduce the size of the spleen. Serenus Sammonicus (PLM21.426–30) and Isidore (Etym. 11.1.127) agreed with, or followed, Pliny in stressing the role of the spleen in laughter.

7. 7.79–80.

8. 24.164. For the identification with cannabis, see André 1972, 150: “Très certainement le chanvre indien (Cannabis indica, variété de C. sativa L)”; “crowfoot” is the suggestion of L&S, the OLD being more guarded with “a plant yielding a hallucinatory drug.”

9. 31.19; Ramsay 1897, 407–8. For the springs on the Fortunate Islands, see Pomponius Mela 3.102.

10. 11.198. For the Greek tradition of such laughter, see Aristotle, Part. an. 3.10, 673a10–12, and Hippocrates, Epid. 5.95. How far this was clearly or systematically distinguished from the tradition, attested even earlier, of the “sardonic smile” or grimace of pain is a moot point; see Halliwell 2008, 93n100, 315.

11. Praef. 17; the first book of the HN consists entirely of a list of contents of books 2 to 37, with the authorities consulted for each.

12. 31.19 (“Theophrastus Marsyae fontem in Phrygia ad Celaenarum oppidum saxa egerere”). Usually assumed to be derived from Theophrastus’ lost work De aquis; see Fortenbaugh et al. 1992, 394–95 (= Physics, no. 219).

13. Aristotle, Part. an. 3.10, 673a1–12.

14. De usu part. 1.22 (Helmreich) = 1, pp. 80–81 (Kuhn); discussed further above, pp. 165–67. For issues of Galen’s dissection and his views on the homology between animal and human, see Hankinson 1997.

15. Mor. 634a–b (= Quaest. conviv. 2.1.11–12).

16. De or. 2.236 (“Haec enim ridentur vel sola vel maxime, quae notant et designant turpitudinem aliquam non turpiter”); Quintilian, Inst. 6.3.7.

17. De or. 2.242 (mimicry), 2.252 (“pulling faces,” oris depravatio), 2.255 (the unexpected), 2.281 (“incongruous”); for further discussion of Cicero and incongruity, see p. 117. See also Quintilian, Inst. 6.3.6–112; like Cicero, Quintilian (6.3.7) stresses the different ways that laughter is stimulated, from words to action and touch.

18. De or. 2.217: “‘Ego vero’ inquit ‘omni de re facilius puto esse ab homine non inurbano, quam de ipsis facetiis disputari.’” It is even closer to the modern cliché if we emend the text (as many have) to read facetius for facilius (“more wittily than wit itself”).

19. Though, conceivably, in opting for “crowfoot” L&S had in mind Ranunculus sardous (the “Sardinian buttercup” or “laughing parsley”), a member of the crowfoot family that is said (by, e.g., Pausanias 10.17.13, though not by Pliny, HN 25.172–74) to produce a sardonic grin.

20. Fried et al. 1998.

21. Plato (Resp. 5.452d–e, and Phlb. 49b–50e) expresses a view of laughter as derisory; in general, as Halliwell 2008, 276–302, makes clear, Plato has much more to say about laughter than is usually recognized.

22. One influential recent source of this is Skinner 2004, which, as its title hints, explicitly equates Aristotle with “the classical theory of laughter” and has telling references throughout to “Aristotle’s theory” (141) or even “Aristotelian theory in its most blinkered form” (153). See also Skinner 2001 and 2002 for similar, though not identical, versions of the argument. Other references to Aristotle as a systematic theorist, or to one or both of his two main laughter “theories,” include Morreall 1983, 5; Le Goff 1997, 43; Critchley 2002, 25; Taylor 2005, 1. Billig 2005, 38–39, is a rare discordant view, describing Plato and Aristotle as offering “scattered observations” rather than “theories.”

23. The classic attempt to find the Greek sources of Cicero’s account of laughter in De or. 2 is Arndt 1904, esp. 25–40, identifying Demetrius of Phaleron as the principal influence. Greek sources also play a major part in the discussion of Cicero in Grant 1924, 71–158. More recently, along similar lines, Freudenburg claimed, “It is quite clear that the Hellenistic handbook writers on rhetoric, followed by Cicero, made no significant advance upon Aristotle’s theory of the liberal jest” (1993, 58). “The liberal jest” is the hallmark of the witty gentleman (eutrapelos); see above, p. 32.

24. To parody Whitehead 1979 [1929], 39, with its famous claim that the “European philosophical tradition . . . consists of a series of footnotes to Plato.”

25. For example, McMahon 1917; Cantarella 1975.

26. Eco 1983. Not all critics have admired The Name of the Rose. For Žižek (1989, 27–28), “there is something wrong with this book” (“spaghetti structuralism” as he nastily calls it) and its views on laughter. Laughter, in Žižek’s worldview, is not simply “liberating” or “anti-totalitarian” (his words) at all, but often “part of the game” of totalitarianism.

27. Skinner 2008 (my italics). Classicists have been known to write in similar, if slightly less confident, terms; see, for example, Freudenburg 1993, 56.

28. Janko 1984, revisited by Janko 2001. Now in the de Coislin collection in the Bibliothèque Nationale (hence its modern name), the Tractatus was once part of a monastery library on Mount Athos. The sections most directly concerned with laughter are 5–6; some of their observations are very close to those found in a preface to manuscripts of Aristophanes and clearly belong to the same tradition—whatever that is.

29. Eloquent are, for example, Arnott 1985 (nicely summarizing, at 305, the much earlier conclusion of Bernays 1853, to the effect that the Tractatus was “a miserable compilation by a pedantic ignoramus”); Silk 2000, 44 (“Janko’s rewarding study . . . tends to evade its striking mediocrity”). Nesselrath (1990, 102–61) carefully argues against a direct Aristotelian connection but produces Theophrastus out of his hat. Halliwell 2013 (reviewing Watson 2012) is a succinct denunciation of the Tractatus.

30. Silk 2000, 44.

31. I owe much here to the view of Aristotle’s theory of tragedy in Silk 2001. Note especially 176: “Aristotle’s theory (indeed, his treatise [Poetics] as a whole), nevertheless, enjoys the reputation of a coherent argument, and not merely a series of brilliant, but loosely connected, aperçus. What is responsible for this? The answer, I suggest, is not the findings of Aristotle’s scholarly interpreters (whose very public disagreements about this, that and the other point of doctrine tell their own story), but rather the constructive—or constructional—use made of Aristotle in post-Aristotelian theories of tragedy (and/or other serious drama), for which Aristotle’s theory of tragedy is a given, and for which it is characteristically constructed as a coherent given.” Although I would probably lay more responsibility at the door of Aristotle’s modern “scholarly interpreters” (as Silk himself does in the footnote to this passage), the role of Renaissance historians and modern “laughter theorists” (from the Renaissance on) seems to me crucial in the retrospective construction of “the Aristotelian theory of laughter” too. For an even more trenchant view of the general incoherence of the Poetics, see Steiner 1996: “As I listen, endlessly, to debates on the Poetics . . . I am prepared to wager that the young man who took notes at Aristotle’s lecture was sitting very near the door on a very noisy day” (545n5).

32. A “theory of laughter” would also imply the definition of laughter as an independent field of inquiry. Despite a range of (lost) treatises “on the laughable” (περὶ τοῦ γελοίου) and despite intense ancient speculation on many aspects of laughter, it is not clear that laughter was ever so defined in antiquity; see Billig 2005, 38–39. The distinction drawn here between “ideas (or even theories) about” and a “theory of” is a crucial one, and my choice of expression throughout this book will reflect that importance.

33. Eth. Nic. 4.8, 1127b34–1128b9, a passage that has prompted very different reactions from critics: subtle and sophisticated for Halliwell 2008, esp. 307–22; muddled (“it slides from tautology to tautology”) for Goldhill 1995, 19. Halliwell 2008, 307–31, provides a useful point of departure, with bibliography, for all the passages I discuss here.

34. Part. an. 3.10, 673a6–8: τοῦ δὲ γαργαλίζεσθαι μόνον ἄνθρωπον αἴτιον ἥ τε λεπτότης τοῦ δέρματος καὶ τὸ μόνον γελᾶν τῶν ζῴων ἄνθρωπον (not De Anima 3.10 as Bakhtin 1968, 68, has it). See further Labarrière 2000 (which does not, for me, rescue the passage from the charge of circularity).

35. Eth. Nic. 4.8, 1128a30; earlier in the passage (1128a4–7), Aristotle characterizes “buffoons” as those who do not avoid giving pain to the butts of their jokes (τὸν σκωπτόμενον). Much modern criticism of Aristotle’s view on comedy has focused on his attitude to Aristophanic Old Comedy and the personal attacks on individuals it contains (see, for example, Halliwell 1986, 266–276, esp. 273, which critiques that focus; M. Heath 1989); this may not be unrelated to his views on laughter, but it is not my concern here.

36. Poet. 5, 1449a32–37: μίμησις φαυλοτέρων μέν, οὐ μέντοι κατὰ πᾶσαν κακίαν, ἀλλὰ τοῦ αἰσχροῦ ἐστι τὸ γελοῖον μόριον. τὸ γὰρ γελοῖόν ἐστιν ἁμάρτημά τι καὶ αἶσχος ἀνώδυνον καὶ οὐ φθαρτικόν, οἷον εὐθὺς τὸ γελοῖον πρόσωπον αἰσχρόν τι καὶ διεστραμμένον ἄνευ ὀδύνης.

37. Rh. 2.12, 1389b10–12: καὶ φιλογέλωτες, διὸ καὶ φιλευτράπελοι· ἡ γὰρ εὐτραπελία πεπαιδευμένη ὕβρις ἐστίν. Note that Aristotle does not say that “wit” is the only way that those who are fond of laughter demonstrate this fondness—rather that those who are fond of laughter will also be witty.

38. The very nature of theater raises one problem about the location of the potential pain. The tacit assumption seems to be that the pain would be that of the actors in their comic masks, at whom the audience laughs. But why would they, whose job it was to provoke laughter, have been liable to pain in the face of it? A similar point is made, in relation to Aristophanes, by Sommerstein 2009, 112.

39. Goldhill 1995, 19; the issue is made even more loaded by the fact that the Nichomachean Ethics itself is addressed to the πεπαιδευμένος (Eth. Nich. 1.3, 1094b22–25).

40. Exactly how far Aristotle is presenting laughter as derisive in this passage is debatable. It depends in part on how far you imagine that his τὸ γελοῖον carries the derisive sense of the Greek word καταγελᾶν (“to laugh down” or “scoff at”). It is certainly true that Aristotle in the Poetics appears to offer a genealogy of comedy from aggressive satire, but the implications of this for laughter as a whole are less clear. Malcolm Schofield has usefully suggested to me that we might see the Aristotelian witty gentleman as a “tease” who gently makes fun of someone’s faults, in such a way as to give pleasure rather than pain—complicated by the fact (as Aristotle notes in Eth. Nic. 4.8, 1128a27–8) that people vary in what they find pleasurable or painful.

41. As is well known, the image of laughter in Greek literature is much more varied, nuanced, and (sometimes) gentle than derision. One classic example is the parental laughter of Hector and Andromache (Homer, Il. 6.471) when baby Astyanax takes fright at the sight of the plume of his father’s helmet.

42. Rh. 1.11, 1371b34–35; the words are bracketed in, for example, Kassel 1976, following Spengel 1867—and the exclusion tentatively supported by Fortenbaugh 2000, 340. Fortenbaugh 2000, with 2002, 120–126, provides a useful discussion of the topic.

43. David, In Isagogen 204.15–16: “Other animals too are capable of laughter, as Aristotle says in the History of Animals about the heron” (ἔστι καὶ ἄλλα ζῷα γελαστικά, ὥσπερ ἱστορεῖ ὁ Ἀριστοτέλης ἐν τῇ ∏ερὶ ζῴων περὶ τοῦ ἐρωδιοῦ). How exactly we are to explain this claim (mistake, misremembering, or subsequent loss of the relevant passage of Aristotle) is unclear.

44. Porphyry, Isagoge 4 (κἂν γὰρ μὴ γελᾷ ἀεί, ἀλλὰ γελαστικὸν λέγεται οὐ τῷ ἀεὶ γελᾶν ἀλλὰ τῷ πεφυκέναι), trans. Barnes 2003. Other writers of Roman imperial date to make this claim include Quintilian, Inst. 5.10.58; Clement, Paedagogus 2.5. The fact that in the second century CE, Lucian (Vit auct. 26) explicitly associates this claim with a character representing Peripatetic philosophy may, but does not necessarily, mean that it originated with Aristotle or his immediate successors (there were plenty of “Peripatetic philosophers” in the Roman Empire). See further, Barnes 2003, 208–9n22.

45. Ménager 1995, 7–41 (on the history of this idea from antiquity to the Renaissance); Screech 1997, 1–5. On Jesus, see Le Goff 1992. In the canonical gospels of the New Testament, Jesus never laughs; he does so repeatedly in the fragmentary Gnostic “Gospel of Judas” (see Pagels and King 2007, 128, arguing that his laughter always introduces the correction of an error).

46. Physiology of laughter: Pliny, HN 11.198; Aristotle, Part. an. 3.10, 673a1–12; babies: Pliny, HN 7.2, 7.72; Aristotle, Hist. an. 9.10, 587b5–7. On Zoroaster, see Herrenschmidt 2000; Hambartsumian 2001.

47. In the context of his discussion of metaphor at Rh. 3.11, 1412a19–b32 (often wrongly cited, for obvious reasons, as Rh. 3.2; see, for example, Morreall 1983, 131).

48. Leeman, Pinkster, and Rabbie 1989, 190–204, offers the most recent detailed discussion of the possible sources (arguing for a mixture of Greek and Roman); 188–89 discusses cavillatio, dicacitas, and facetiae. Fantham 2004, 186–208 (quotation on 189), with Corbeill 1996, 21–22nn13–14, a sharp, up-to-date account.

49. Halliwell 2008 is especially good on philosophical views of laughter: esp. 271–76 (Pythagoreanism), 276–302 (Platonic Socrates), 302–7 (Stoicism), 343–71 (Democritus), 372–87 (Cynicism).

50. I am referring here to Stein 2006 (the use of slapstick in a hookworm eradication campaign); Janus 2009 (the scripted laughter in Joyce interrupts the traditionally “silent reading” of the novels); Lavin and Maynard 2001 (comparing survey centers where interviewers are “prohibited” from laughing in the course of an interview with those where they are not); Kawakami et al. 2007 (drawing distinctions between and dating the occurrence of spontaneous vs. social laughter in infants).

51. Chesterfield 1774, vol. 1, 326–32, esp. 328 (letter of 9 March 1748), reprinted in D. Roberts 1992, 70–74, esp. 72; see further above, pp. 60, 66–67.

52. W. Lewis et al. 1914, 31 (“We only want Tragedy if it can clench its side-muscles like hand on it’s [sic] belly, and bring to the surface a laugh like a bomb”); Cixous 1976 (“She’s beautiful and she’s laughing,” 885; “rhythm that laughs you,” 882). The essays of Baudelaire 1981 [1855] and Bataille 1997 [1944] have been influential in many of the most radical approaches to laughter. The rich tradition of laughter in feminist writing, from fiction to psychoanalysis, is a major theme of Parvulescu 2010, esp. 101–18, to which Lessing 1962 (a feminist novel, in which laughter is a major player) would be an important addition (see, briefly, Scurr 2003). For a different strand of modern feminist use of laughter (in relation to a Latin text), see above, pp. 84–85.

53. Morreall 1983, 4–37; Critchley 2002, 2–3; more skeptically, Halliwell 2008, 11. Lippitt 1994; 1995a; and 1995b offer a clear, critical introduction to each theory in turn.

54. Hobbes 1969 [1640], 42; Sudden Glory is the title of Sanders 1995.

55. Ludovici 1932, 98–103; Gruner 1978, 43; R. A. Martin 2007, 44–47 (a useful summary). Quotation: Rapp 1951, 21.

56. Rh. 3.11, 1412a31.

57. Kant 1952 [1790], 196–203, quotation on 199; Bergson 1911, esp. 12–38; Raskin 1985; Attardo and Raskin 1991; Attardo 1994 (put into a classical perspective by N. Lowe 2007, 1–12). For those unfamiliar with the old English joke about the door, it plays on the aural ambiguity between the noun jar (a storage vessel, often of glass) and the adjective/adverb ajar (meaning “slightly open”).

58. Deckers and Kizer 1974; Deckers and Kizer 1975; Nerhardt 1976; Deckers 1993; with a useful overview in R. A. Martin 2007, 68–70. The question of whether the subjects in this case may (also) be laughing at the experimenters is rarely raised!

59. Spencer 1860.

60. Phil. 2.39: explaining that as the army camp was “full of care” (plena curae), the jokes served “to relax their minds” (animis relaxantur—the verb can indicate a release from pressure)—though I may be trying to push this too far, when Cicero is thinking much more generally of the role of joking as a break from the cares of war. Corbeill 1996, 185–89, discusses the jokes made on this occasion.

61. Freud 1960 [1905] (“The hearer of the joke laughs with the quota of psychical energy which has become free through the lifting of the inhibitory cathexis,” 201). Experimental psychology does not confirm what Freud’s argument seems to imply: that the more repressed you are, the more you will laugh at a dirty joke (Morreall 1983, 32).

62. M. Smith 2008 rightly criticizes the preoccupation of most laughter theorists with uncontrollable laughter. Ruch and Eckman 2001 is typical in its classification of outbursts of laughter into “spontaneous” on the one hand and “contrived” or “fake” on the other (the terms themselves are a giveaway). The recent neurological work of Sophie Scott and her colleagues has been much more interested in “social” as well as uncontrollable laughter, tracing differences and similarities in the response of the brain to laughter of different types. See, for example, McGettigan et al. 2013; S. Scott 2013.

63. Scruton in Scruton and Jones 1982 offers useful observations about the range of and exclusions from modern studies of laughter (“It is not laughter, but laughter at or about something, that interests the philosopher,” 198); likewise Parvulescu 2010, 3–4 (“Most ‘theories of laughter’ are not concerned with laughter”).

64. Morreall 1983, 30, points to the difficulty in Freud’s view of the conversion of psychic into physical energy, as does, rather more elegantly, Cioffi 1998, 264–304, in his discussion of Wittgenstein’s critique of Freud (“Imagine a world in which, like ours, people laughed at jokes, but unlike ours did not know what that were laughing at until they discovered the unconscious energic processes hypothesised by Freud,” 277). Richlin 1992a, 72, sums up some of the basic problems with the Freudian account succinctly: “That the pleasure consists in relief, in the released pressure of a lifted inhibition, does not describe the feeling of a laugh very well.” Earlier generations of modern laughter theorists were more concerned to link the physical “symptoms” of laughter to its cause: Laurence Joubert, for example, traces laughter to a physical reaction of the heart, contracting and expanding in response to conflicting emotions of joy and sorrow (Joubert 1980 [1579], 44–45). Gatrell 2006, 162–67, traces the eighteenth-century reaction against such physical explanations.

65. See, for example, Gruner 1997, 131–46 (in which a groan in response to a pun is an admission of defeat). Note Baudelaire’s pointed dismissal of the theory as a whole: “Laughter, so they say, comes from superiority. I should not be surprised if, in face of this discovery, the physiologist himself were to burst out laughing at the thought of his own superiority” (1981 [1855], 145). On the general problems with all-encompassing theories of “amusement,” see Scruton in Scruton and Jones 1982, 202.

66. Freud 1960 [1905], 248–54. Perhaps the most problematic aspect of this very problematic argument is Freud’s claim that in the process of ideation more energy is expended on a large movement than on a small one.

67. Berger 1997, 29–30 (incongruity); Sanders 1995, 249. Quotation: Bergson 1911, 18.

68. Morreall 1983, 16 (incongruity)—though Morreall adds, “Because it [the incongruity theory] did not fit in with the superiority theory of his Poetics and Nichomachean Ethics, he never developed it”; Atkinson 1993, 17–18 (relief).

69. Hobbes 1996 [1651], 43; Skinner 2001, 445–46; Skinner 2002, 175–76; Skinner 2004, 162–64.

70. Richlin 1992a, 60 (psychosocial dynamics); Goldhill 2006, 84 (not knowing what we are laughing at); Corbeill 1996, 4–5 (tendentious vs. innocent).

71. Le Goff 1997, 46–47, briefly discusses the question of how far laughter can be reduced “to a single phenomenon.”

72. Douglas 1971, 389. Embedded in her remarks is also the assumption, standard at least since Bergson (1911, 12), that laughter is essentially social, that you cannot laugh alone (hence the canned laughter on television programs). For Pliny, see above, p. 25. I say probably because in some circumstances and in some cultures, belching too can straddle the divide of nature and culture and be taken as meaningful. The other action to which Pliny refers in this passage, spitting, is different again: this is always communication, and not a natural bodily eruption.

73. Aristotle, [Pr.] 35.7, 965a18–22, though the next passage of this compilation (almost certainly put together in Peripatetic circles over a long period from the third century BCE on) claims that people are ticklish only in the armpits. Joubert 1980 [1579], 86, identifies the skin between the toes as a prime site for tickling.

74. Provine 2000, 99–127; R. A. Martin 2007, 173–76. One much disputed theory of tickling—the so-called Darwin-Hecker hypothesis—suggests that there is much more in common between tickling and humor than is usually allowed: both produce laughter by very similar neural processes involving the same region of the brain (Darwin 1872, 201–2; Panksepp 2000, but see C. R. Harris and Christenfeld 1997; C. R. Harris and Alvarado 2005).

75. In my experience, one particularly sadistic version involves contriving an excuse to remove a child from the room—when s/he returns, all the other children are uproariously laughing. Soon enough the returner will join in the laughter, and at that point s/he faces increasingly aggressive questions from the others on what s/he is laughing at—until tears are the result.

76. Lautréamont 1965 [1869], 5.

77. Nietzsche 2002 [1886], 174–75; 1990 [1886], 218.

78. Douglas 1971, 387.

79. Turnbull 1961 (quotation on 45); the mountain people (the Ik) are the subject of Turnbull 1973. Ballard 2006 and Boyer 1989 offer critiques of Turnbull’s general approach to the Pygmies. “Subjective, judgmental and naïve” are the words of Fox 2001 (referring specifically to Turnbull’s treatment of the Ik).

80. On Chesterfield, see pp. 36, 60, 66–67.

81. For example, Catullus 64.284; Lucretius 1.8. The etymology of ridere is uncertain, but the Greek γελᾶν (laugh) may have a root in the idea of brightness and luster, and it is not inconceivable (though unlikely) that the poets are making a scholarly allusion to that in their usage. On γελᾶν, see Halliwell 2008, 13n33, 523, for a sensible discussion, with bibliography.

82. Darwin 1872, 120–21, 132–37, 198–212; with Davila-Ross et al. 2011 (as just one example of up-to-the-minute investigations of ape laughter). Dogs: Douglas 1971. Rats: Panksepp and Burgdorf 1999; Panksepp 2000.

83. Panksepp and Burgdorf 1999, 231, briefly discusses the opposition.

84. Scruton in Scruton and Jones 1982, 199.


1. Herzen’s remark (2012 [1858], 68) is quoted by, among others, Bakhtin 1968, 59; Halliwell 2008, vii; and Le Goff 1997, 41.

2. Le Goff 1997, 41, usefully highlights this distinction between protocol and practice.

3. Published as Thomas 1977. In French, the work of Jacques Le Goff has been similarly programmatic; see Le Goff 1989. Thomas’s original talk was given as the Neale Lecture in English History at University College London on 3 December 1976. He started by suggesting that Sir John Neale, in whose honor the series had been founded, would have thought laughter an “ill-defined and even unhistorical” topic of research. The idea that one’s predecessors or more senior colleagues would disapprove of the subject is something of a cliché among historians of laughter. Saint-Denis (1965, 9) complained that his university authorities had found the topic so distasteful that they refused even to publish a summary of his course of lectures—“Le rire des Latins”—in their Revue des Cours et Conférences; even in the 1990s, Verberckmoes 1999, ix, said much the same.

4. Plutarch, Mor. 633c (= Quaest. conviv. 2.1.9). Cicero, De or. 2.246, likewise puts a joke against a luscus (a man blind in one eye) in the category of the “scurrilous”; predictably, the emperor Elagabalus (SHA, Heliog. 29.3) enjoyed making a joke of all kinds of people with bodily “peculiarities,” from the fat to the bald and the lusci (see p. 77). Plutarch’s protocols might suggest that the joking songs of Caesar’s soldiers (Suet., Iul. 51) should be seen as relatively good humored.

5. Dio Chrysostom, Or. 32.1 (ἐπειδὴ παίζοντες ἀεὶ διατελεῖτε καὶ οὐ προσέχοντες καὶ παιδιᾶς μὲν καὶ ἡδονῆς καὶ γέλωτος, ὡς εἰπεῖν, οὐδέποτε ἀπορεῖτε), 32.56 (“as if you’d been hitting the bottle”—ἐοίκατε κραιπαλῶσιν).

6. Tacitus, Germ. 19. This passage already hints at some of the complexities in understanding the sense of the apparently simple word ridet, which I will explore in more detail. “Laughs off”—in the sense of “takes as a joke”—seems attractive here and accords with the phrase that follows (nec corrumpere et corrumpi saeculum vocatur, “and to corrupt or be corrupted is not put down to ‘the times we live in’”). But as many recent critics have emphasized (for example, Richlin 1992a), “ridicule” in traditional Roman culture could be a powerful weapon against deviance. My hunch is that Tacitus is being (as often) even smarter than he seems and is querying not merely contemporary Roman corruption but some of the most traditional mechanisms (here ridicule) through which Rome had policed its morality. (But see above, pp. 105–8, 120–23, on the tendency of modern scholarship to overemphasize the aggressive, policing functions of Roman laughter.)

7. Twain 1889, 28–29.

8. Leeman, Pinkster, and Rabbie 1989, 259.

9. Murgia 1991, esp. 184–93.

10. Inst. 6.3.100, the Latin text of D. A. Russell in the Loeb Classical Library (similar to that printed in the Teubner text, ed. L. Radermacher).

11. “Hopelessly ungrammatical” because mentiri is a deponent verb, used in the passive voice, whereas mentis is an active form. There is a little more logic to some of these changes than I have perhaps made it appear: mentis, for example, might be a (not unparalleled) manuscript conflation of an original me[n] ex te metiris.

12. Murgia 1991, 184–87, includes further and fuller arguments for his changes.

13. The impossible obicientibus arbore demands some change. Murgia would reasonably claim that it is easier to see how his version (barbare) rather than the more usual obicienti atrociora could have been corrupted into the garbled text of the manuscript (arbore). But he has not convinced other textual critics (for example Russell, whose Loeb text of 2001 notes but does not follow Murgia). Murgia’s emendation of the main joke entails other changes to earlier sentences. The phrase “Umis quoque uti belle datur” introduces the story in the manuscripts of Quintilian. Umis makes no sense whatsoever. It is usually emended to “Contumeliis quoque . . .” (“Insults also can be neatly used”—“I suppose this emendation must be right,” Winterbottom 1970, 112); Murgia suggests “Verbis quoque . . .” (“Words/quips also can be neatly used”).

14. Fontaine 2010.

15. On one occasion, for example, he claims that Varro (Ling. 9.106) already in the first century BCE was working from a faulty text of Plautus that had missed the joke (Fontaine 2010, 29); if so, there are interesting implications for the transmission of jokes within the Roman world itself. But it may not be so. Even assuming that Fontaine’s reading is the correct version of what Plautus wrote, Varro’s text—as Fontaine concedes—may have been “fixed” by a later editor to bring it into line with what by then had become the standard reading.

16. Rud. 527–28; Fontaine 2010, 121–23. He goes on to suggest a pun elsewhere in the play on the word algor (cold), as if it were a verbal form meaning “to gather seaweed.” Sharrock 2011 discusses this particular suggestion and Fontaine’s overall approach.

17. The telling phrase of C. W. Marshall, on the jacket of Fontaine 2010.

18. In arguing in this way, I am not unaware of the strand of research (stretching back to Darwin 1872) that claims there are natural physiological facial expressions of emotion—a strand that some art historians have recently exploited. David Freedberg, for example, has drawn on the research of Paul Ekman and others to argue for clearly identifiable expressions in works of art (see Freedberg 2007), yet as he himself admits, problems and controversies remain, and it is certainly not enough to assert, as he does (33–34), “A comparison of the terrible images shown on Al-Jazeera of Margaret Hassan immediately prior to her execution in 2004 and earlier photographs of her smiling leaves one with no doubt at all about the possibility of identifying constants of emotional expression. The fear and the cheerfulness are instantly and indisputably identifiable as such.” Here I would stress only that, even if we were to accept a “natural” relationship between expression and emotion, an artistic representation is a very different matter—while in any case, laughter is not itself an emotion or even necessarily the product of emotion (or, as Parvulescu 2010, esp. 6–9, would have it, “a passion”).

19. Quotations from Stewart 2008; Goldhill 2008; Cohen 2008; R. D. Griffith 2008.

20. For example, M. Robertson 1975, vol. 1, 101–2, and Trumble 2004, 14–15, see it as a form of animation; Giuliani 1986, 105–6, combines animation with beauty (at the start of a more complex discussion that includes the Gorgon’s “grimace,” 105–12); Yalouris 1986 canvasses the idea of aristocratic contentment. On smiling in general, see above, pp. 73–76.

21. The best survey of these debates is Halliwell 2008, 530–52, which also discusses ancient descriptions (including some of the Roman period) of art that refer to laughs and smiles (notably several in [the older and younger] Philostratus’ ecphrases of painting: e.g., Philostratus mai., Imag. 1.19.6, 2.2.2, 2.2.5; Philostratus min., Imag. 2.2, 2.3). The theoretical implications of the Gorgon’s expression are central to Cixous 1976 (see above, pp. 36–37).

22. Trumble 2004, l–liii; quotation from Wallace Collection 1928, 128. Schneider 2004 discusses medieval images of laughter, including the famous sculpture of the Last Judgment at Bamberg Cathedral, with Jesus between the Blessed and the Damned. This account makes clear what a fine line there is between the ecstatic smiles of the Blessed and the grimaces of the Damned. The Mona Lisa offers another puzzle, debated by Freud, John Ruskin, Bernard Berenson, and many others; reviewed by Trumble 2004, 22–29. So too, as Le Goff points out (1997, 48–49), do images of the story of Isaac. Though laughter is fundamental to that story (and the name Isaac means “laughter”), “if one looks at representations . . . one finds no attempt to represent the laughter.”

23. J. R. Clarke 2007.

24. J. R. Clarke 2007, 53–57. It is tempting to link this (as Clarke does) with the laughter headlined by Petronius, Sat. 29, even though the coordinates are rather different. There a man falls down in astonishment at seeing a lifelike painting of a dog at the entrance to Trimalchio’s house, and his friends laugh at him (not at the dog!); the passage is minutely analyzed by Plaza 2000, 94–103. As a further example of a funny double take, Clarke offers (52) the story of the contest in illusionism between the painters Zeuxis and Parrhasius (Pliny, HN 35.65–66); though laughter is not explicitly mentioned here, it does link with another story of Zeuxis, which I discuss on pp. 72–73.

25. The idea of laughter as apotropaic is a major theme in Clarke’s book (esp. 63–81). In my view (see, e.g., Beard 2007, 248, and above, p. 146), this term explains much less than many scholars like to think and raises more problems than it solves. Do we really imagine that the entranceway to the bijou House of the Tragic Poet was a place of liminality haunted by the evil eye?

26. Ling 2009, 510.

27. Thomas 1977, 77 (my italics). Likewise Le Goff 1997, 40 (“Attitudes to laughter, the ways in which it is practised, its objects and its forms are not constant but changing. . . . As a cultural and social phenomenon, laughter must have a history”); Gatrell 2006, 5 (“Studying laughter can take us to the heart of a generation’s shifting attitudes, sensibilities and anxieties”).

28. Chesterfield 1774, vol. 1, 328 (letter of 9 March 1748); reprinted in D. Roberts 1992, 72.

29. He references in particular the French version of Elias 1978—whose original German text, Über den Prozess der Zivilisation (1939), had not yet been translated into English. It is no coincidence that one of Elias’s essays, left unfinished and unpublished at his death, was on laughter; it is discussed by Parvulescu 2010, 24–26.

30. All quotations from Thomas 1977.

31. Bakhtin 1968.

32. Pan’kov 2001.

33. Le Goff 1997, 51, rightly stresses that Bakhtin was only the most famous of a large group of Soviet scholars working on laughter in the mid-twentieth century; see also (in German translation) Lichačëv and Pančenko 1991.

34. Even some of Bakhtin’s warmest admirers concede this. See, for example, Stallybrass and White 1986, 10: “It is difficult to disentangle the generous but willed idealism from the descriptively accurate in passages like these. Bakhtin constantly shifts between prescriptive and descriptive categories in his work.”

35. Gatrell 2006, 178 (chapter title).

36. This chronology is sketched in the first chapter of Bakhtin 1968, 59–144; quotations on 72, 107, 119.

37. Burke 1988, 85 (reviewing four books heavily dependent on Bakhtinian analysis, including Stallybrass and White 1986, and briefly surveying the reception of Bakhtin in the West). For the enthusiastic adoption of Bakhtin by some critics of classical literature and art, see, for example, Moellendorff 1995; Branham 2002; J. R. Clarke 2007, 7–9; and below, nn. 46–47.

38. Pan’kov 2001, 47.

39. Critiques (or critical developments) of aspects of Bakhtin’s treatment of carnival run into thousands. I have found particularly useful Davis 1975, 97–123, and Stallybrass and White 1986, esp. 6–19 (on the simultaneously radical and conservative aspects of carnival), with Chartier 1987 (on the discourse of nostalgia in the culture of carnival); Le Roy Ladurie 1979 (on carnival’s violence); M. A. Bernstein 1992, 34–58 (on its potential savagery, with important reflections on earlier, Nietzschean models of carnival and their ambivalence); J. C. Scott 1990, 72, 172–82 (stressing the apparent acquiescence of the people in the elite script of carnival); Greenblatt 2007, 77–104 (on the relationship between Rabelais’s text and “real” laughter); Silk, Gildenhard, and Barrow 2014, 121–24 (from a classical starting point).

40. Gatrell 2006, 161.

41. For a brief introduction to the festival and a review of the literary evidence, see D’Agostino 1969; Scullard 1981; Graf 1992, 14–21 (on the etiology and the ritual).

42. Frazer 1913, 306–411; Nietzsche 1986 [1878], 213; Nietzsche 2002 [1886], 114. M. A. Bernstein 1992, 34–35, emphasizes the underlying pessimism of Nietzsche’s account. Frazer was predictably most concerned with drawing a connection between the “Saturnalian king” and his motley crew of dead, divine, and priestly kings. This connection was, he believed, supported by the puzzling Acts of Saint Dasius, which claims (in what is probably a Christian fantasy) that the Saturnalian king in a military garrison on the Danube c. 300 CE was killed at the end of his thirty-day “rule.” See Cumont 1897; Musurillo 1972, 272–75; Versnel 1993, 210–27.

43. Bakhtin 1968, quotations on 7, 138, 70, 14. Bakhtin’s stronger claim of a literally unbroken continuity between the Saturnalia and medieval carnival (8) has generally been viewed more suspiciously (Nauta 2002, 180).

44. Versnel 1993, 136–227, reflects many of these claims (from a partially Bakhtinian perspective); “exuberant gorgings . . .” is his phrase (147), echoed in Minois 2000 (“les orgies des saturnales,” 65). See also Bettini 1991, 99–115; Champlin 2003, 150–51 (at the Saturnalia “within the miniature republic of the household, slaves might act as magistrates and judges,” 150); Dolansky 2011 (495: “Normative codes of behavior were reversed, with masters waiting upon slaves who enjoyed the right to drink to excess and chide their masters”).

45. There is no firm evidence for the precise dating of the Apocolocyntosis. Nauta (1987, 78–84) lays out the arguments and inferences (such as they are) that might point to a specifically Saturnalian date as an introduction to his Saturnalian reading of the text (focusing on laughter and the inversion of norms). Branham 2005 discusses at length Bakhtin’s stress on “Menippean satire”—the genre of the Apocolocyntosis.

46. Gowers 2005, 60, puts both Sat. 2.3 (Damasippus) and Sat. 2.7 (Davus) in a Saturnalian frame (“The topsy-turvy festival of the Saturnalia . . . allows two speakers . . . freedom of speech . . . to remove the smug mask Horace manufactured in Book 1”). Sharland 2010, 261–316, is a particularly hard-line Bakhtinian reading of the Saturnalia and a hard-line Saturnalian reading of Sat. 2.7. See, e.g., 266: “True to the customs of the Carnival, and its predecessor the Saturnalia, a lowly character (in this case, Davus) has been elevated to the position of ‘king’ figure, and is allowed to ‘reign’ temporarily’; 268: “Through its inversions and reversals, Carnival (and Saturnalia) characteristically juxtaposed opposites, matched incompatibles, and joined odd couples.”

47. The classic discussion of comedy as an inversionary Saturnalian genre is Segal 1968 (e.g., 8–9, 32–33), though its inspiration is more Frazer (8) than Bakhtin; the position is reiterated in Segal 2001, 149 (in which Bakhtin has a walk-on part on 8). For other carnivalesque readings, see, for example, Bettini 1981, 9–24; Gowers 1993, 69–74 (a more subtle connection between the textual banquets of Plautine comedy, carnivalesque consumption, and the Saturnalia). Other students of Roman comedy have been dubious about a carnivalesque or Bakhtinian reading, or about some aspects of it: for example, Manuwald 2011, 149; McCarthy 2000, 17–18, esp. n. 26 (deploying Bakhtinian theory but questioning its social “optimism”).

48. Part of the Saturnalian spirit is captured in the illustration accompanying the month of December in the fourth-century CE Calendar of Philocalus, which shows a man, wearing tunic and cape, standing beside a gaming table—with some game (of the edible sort) hanging up behind him. Stern 1953, 283–86, with planches 13 and 19.2.

49. There was feasting and drinking, yes, but no evidence of gross bingeing in the style of carnival. Not surprisingly, it is hard from the scanty material we have to get a clear idea of levels of consumption: Seneca, Ep. 18 (a curmudgeonly letter on how far the philosophical elite should join in the Saturnalia), talks vaguely of luxuria and of dining hilarius (in a jollier fashion); Aulus Gellius 2.24.3 refers to sumptuary laws covering the occasion (but sumptuary legislation is no guide to levels of real excess); SHA, Alex. Sev. 37.6 suggests that this particularly mean emperor splashed out on just a pheasant for Saturnalia. Gowers 1993, 69–74 stresses the consumption of pork as a carnival dish. Exactly how drunk Cato’s slaves would have got on the rations he prescribed for the Saturnalia (Agr.57) is anyone’s guess. Assuming the text is correct, he suggests that the most generous ration for a month’s wine amounted to just under a liter a day per head. Additionally, slaves should be allowed an extra ten liters on the Saturnalia and Compitalia (separately or combined is unclear). Ten liters of modern-strength wine consumed on a single day would indeed suggest Bakhtinian excess, but we are probably dealing with wine of lower strength, and it might not have amounted to much more than double rations if consumed over the duration of both festivals.

50. Apoc 4.3; the emperor’s dying words are reported as “O dear I think I’ve shat [concacavi] myself.”

51. Of course, the Saturnalia is a self-consciously elite work, full of wit, upper-class jokes, and ludic learning, embedded in one version of the academic culture of the fifth century CE. But its wit is in fact not so different from the style of Saturnalian wit we find elsewhere. For references to riddles and puns, see AL 286; Aulus Gellius 18.2, 18.13.

52. Macrobius, Sat. 1.12.7, 1.24.23.

53. Seneca, Ep. 47.14, contra Champlin 2003, 150, which relies on almost certainly faulty modern punctuation. Contra Versnel 1993, 149, Dio 60.19 refers to slaves adopting not the “roles of their masters” but the “clothes of their masters.”

54. Tacitus, Ann. 13.15; discussed by Champlin (2003, 150–53) in the context of his wider claims that there was a “Saturnalian style” to the reign as a whole. Tacitus certainly is suggesting that having Nero on the throne was like being ruled by “Saturnalicius rex.”

55. Accius apud Macrobius, Sat. 1.7.36–37 (= ROL2, Accius, Annales 2–7): the masters prepare the meal, but it is eaten together; Macrobius, Sat. 1.11.1; SHA, Verus 7.5 (slaves and masters eating together, at Saturnalia and other festivals); Macrobius, Sat. 1.7.26 (licentia). Note also the slogan on the Calendar of Philocalus (see n. 48), “Now, slave, you can play/gamble with your master.” Bakhtin and many modern accounts tend to use the ideas of inversion and equality interchangeably, but in fact they represent two crucially different forms of festal transgression.

56. Pliny’s famous account of not spoiling his household’s fun at the Saturnalia (Ep. 2.17.24) oozes paternalism. (A casual reference of his to the Saturnalia in Ep. 8.7 no doubt reflects traditions of Saturnalian free speech, but I am not convinced that it should be seen in quite the inversionary terms that Marchesi 2008, 102–17, imagines.)

57. Fairer 2003, 2.

58. See above, n. 28. Chesterfield’s advice is often assumed (by, e.g., Morreall 1983, 87) to be fairly typical of eighteenth-century preoccupations with the control of laughter. True, it is not unparalleled; see, for example, the advice of Pitt senior to his son (W. S. Taylor and Pringle 1838–40, vol. 1, 79). But as Gatrell (2006, 163–65, 170, 176) makes clear, Chesterfield’s published views were extreme and, in any case, represented an insistence on the control of laughter that can be found at other periods. Chesterfield was also more complicated than he is given credit for—a renowned wit, of (by the standards of the day) grotesque appearance, and celebrated prankster (see Dickie 2011, 87).

59. Thomas 1977. His tactic (as his choice of words indicates: “lingers,” “among the common people,” “continued in villages,” etc.) is to reconcile the differences by implying that more isolated regions or those below the elite took longer to adopt the new protocols.

60. A phrase supposedly uttered by Queen Victoria but as historically perilous as Lord Chesterfield’s advice, for even more reasons: it is not clear that Victoria ever said this or—if she did—in response to what. Vasey 1875 is a truly thoroughgoing, much less well-known, and sometimes hilarious agelastic treatise. “The conclusion is unavoidable, that the absurd habit of laughing is entirely occasioned by the unnatural and false associations which have been forced upon us in early life” (58) gives a flavor.

61. This theme runs throughout Chartier 1987.

62. Much recent work on eighteenth-century laughter and other forms of “sensibility” is alert to this nexus of complexity. In addition to Gatrell 2006 and Dickie 2011, Klein 1994 is an illuminating study. There are, of course, subtle variations on these generalizations. As Ruth Scurr alerted me, the laughter of the French revolutionaries was defined as more innocent than the contrived and vicious laughter of the royal court (see, for example, Leon 2009, 74–99). Some modern celebrations of the relaxation of comic censorship in print and onstage might seem to point in the opposite direction, but the celebration of the freedom of public expression of coarseness is different from the celebration of increasing coarseness itself.

63. Fam. 9.15. This is, in fact, a more puzzling passage than my quotations suggest. If the text as we have it is broadly correct (which it may well not be), Cicero included his home region of Latium among the foreign influences. But as Shackleton Bailey (1977, 350) asks, “How can Cicero of Arpinum equate Latium with peregrinitas?” The overall sense is clear, though the details are irrecoverable. As we shall see in the next chapter, Cicero’s rhetorical treatises are more equivocal about the propriety of old-fashioned festivitas.

64. Livy 7.2; Horace, Epist. 2.1.139–55. The passage of Livy—which offers a brief, multistage account of the origins and development of dramatic festivals at Rome—has been intensely debated (on its meaning, sources, and reliability); for a review, see Oakley 1997, 40–58. In the third stage, the performers are said to give up uttering crude compositions akin to Fescenninus versus (presumably the jesting banter characteristic of Livy’s second stage). Horace’s genealogy envisages the rustics bantering until the Fescennina licentia became so nasty that it had to be controlled by law. For the disputed etymology of Fescennine—from the name of an Etruscan town or from fascinum—see Oakley 1997, 59–60.

65. Gowers 2005; Gowers 2012, 182–86, 199–204 (with review of earlier work); Oliensis 1998, 29.

66. The title of the second chapter of Saint-Denis 1965; the first reflects a similar style: “Jovialité rustique et vinaigre italien.” See also Minois 2000, 71: “Le Latin, paysan caustique.”

67. Macrobius attributes some Fescennini to the emperor Augustus (Sat. 2.4.21); otherwise, as Oakley (1997, 60) rightly insists, the only institutional context attested in the late Republic and early empire is the wedding ritual (Hersch 2010, 151–56); whether or not the term should be applied also to the ribald, joking verses sung at a Roman triumph—as Graf (2005, 201–2), along with many others, implies—is far less clear.

68. Conybeare 2013 is a major study of laughter focused on biblical and theological texts, Jewish and Christian, to which readers frustrated by my limitations are warmly directed!


1. The OLD, for example, offers “to smile at, upon or in response to” for arridere/adridere and “to laugh at, mock, make fun of” for irridere; ridere with a dative suggests “to laugh as a sign of goodwill.” The etymology of ridere is obscure, despite occasional attempts to relate it to the Sanskrit for “to be shy” or to the Boeotian form κριδδέμεν (a variant of γελᾶν, “laugh”).

2. Ovid, Ars am. 2.201; Terence, Ad. 864; Horace, Ars P. 101.

3. Silius Italicus 1.398; another decidedly sinister use of arridere (Seneca, Controv. 9.2.6) is discussed on pp. 79–80. It most likely indicates mocking laughter at Cicero, De or. 2.262.

4. Eun. 249–50; Priscian in GLK 3.351.11 (= Inst. 18.274). Most modern translators and critics who have rightly focused on this passage (e.g., Damon 1997, 81; Fontaine 2010, 13–14) have also missed the full nuance, whichever way they choose to translate adridere.

5. Martial, Epigram. 6.44: “omnibus adrides, dicteria dicis in omnis: / sic te convivam posses placere putas” (ll. 3–4, as the manuscripts have it); the typical sting in the tail turns out to be the man’s fondness for oral sex. For the emendation, see Shackleton Bailey 1978 (quotation on 279, my emphasis—and he goes on: “Since that compound does not take a dative in classical Latin, omnibus must become omnis”); this reading is now incorporated in his Teubner edition of 1990 and repeated in the Loeb Classical Library edition of 1993. For critical discussion of the emendation and Shackleton Bailey’s interpretation of the poem, see Grewing 1997, 314; Nauta 2002, 176–77.

6. Catullus 39, passim; Tacitus, Ann. 4.60 (a more sinister context).

7. Ovid, Ars am. 3.283 (advising girls not to display immodici rictus while laughing); Lucretius 5.1064 (of dogs); see further p. 159.

8. Nonius Marcellus 742 (Lindsay): “non risu tantum sed et de sono vehementiore vetustas dici voluit.”

9. Verr. 2.3.62. That at least is Cicero’s highly colored presentation of the scene (he admits that Apronius’ uproar is only extrapolated from his laughter at the trial).

10. Persius 1.12; see 1.116–18 for an explicit comparison with Horace.

11. Catullus 13.5; Suetonius, Vesp. 5.2; Lucretius 4.1176.

12. Nonius Marcellus 742 (Lindsay) quoting Accius (= ROL2, Accius, Tragoediae 577) on the pounding of the ocean—the text is not entirely certain, and on another reading cachinnare could refer to the screeching of a seabird; Catullus 31.14 (of the ripples of Lake Garda), 64.273 (“leviter sonant plangore cachinni”). There is a curious set of relations here with aspects of the Greek laughter lexicon. Γελᾶν, in Greek, is commonly used for the behavior of the sea. Cachinnare matches (even if it is not directly derived from) the Greek καχάζειν, which does not appear to be used metaphorically for the sound of water, though the very similar Greek word καχλάζειν (with a lambda) is a regular term for “splashing.” It is tempting to think that this pairing lies somewhere behind Catullus’ play withcachinnare (or perhaps καχάζειν and καχλάζειν are not as separate as modern lexicography likes to make them).

13. M. Clarke 2005 is a useful recent review of relevant material stressing the unfamiliarity of the Greek semantics of “smiling”: see also Lateiner 1995, 193–95; Levine 1982; Levine 1984. For the stress on the face: Sappho 1.14; Hom. Hymn 10.2–3 (note that, very unusually, Homer, Il. 15.101–2, has Hera laughing “with her lips”).

14. Halliwell 2008, 524, part of a longer, careful discussion (520–29) of Greek laughter terminology and its physical referents, though apart from this appendix, μειδιῶ has hardly a mention in the book.

15. For example, Virgil, Aen. 1.254 (see also Homer, Il. 15.47); Servius Auct. (ad loc.) quotes a parallel passage from Ennius, which uses ridere rather than subridere: Ennius, Ann. 450–51 (ROL) = 457–58 (Vahlen).

16. Catullus 39. In Kaster 1980, 238–40, the key examples are Sat. 1.4.4, 1.11.2 (quoted), 3.10.5, 7.7.8, 7.9.10, and 7.14.5 (translated accordingly in his edition of Macrobius for the Loeb Classical Library), but note also 1.2.10 (involving the whole face) and 7.3.15 (accompanying an apparent insult), neither of which quite match. Kaster is, I suspect, too keen to find smiles in both Macrobius and the texts he uses for comparison. He refers, for example, to the smiles of Cicero’s dialogues “as an instrument of amused debate and rejoinder,” but the Ciceronian passages he cites refer explicitly to a variety of “laughing” (ridens, adridens, etc.). I am relieved that König 2012, 215–26, has(independently) similar reservations over details in Kaster’s argument on smiling, although for different reasons.

17. Catullus 39.16; Ovid, Ars am. 2.49; Ovid, Met. 8.197; Livy 35.49.7; Quintilian, Inst. 6.1.38 (renidentis a plausible emendation for the manuscript residentis).

18. Apuleius, Met. 3.12; Valerius Flaccus 4.359; Tacitus, Ann. 4.60.2.

19. 1.2.10.

20. This is obviously made more complicated by the fact that os, oris (occasionally used with renideo, as at Ovid, Met. 8.197) could refer to the face or the mouth.

21. I am thinking here of the work of such scholars as Paul Ekman (1992; 1999) and that discussed in ch. 3, n. 18. I hope that by this point in the book I do not need to explain why I do not follow such a universalist path.

22. Chesterfield 1890, 177–79 (letter of 12 December 1765, to his godson), reprinted in D. Roberts 1992, 342–43: “The vulgar often laugh but never smile; whereas, well-bred people often smile, but seldom laugh.” Similar sentiments are expressed in Chesterfield 1774, vol. 1, 328 (letter of 9 March 1748), reprinted in D. Roberts 1992, 72.

23. “Kissing,” Jones’s (as yet) unpublished paper given at Columbia University in 2002, also points to the ancients’ careful calibration of different styles of kissing.

24. Le Goff 1997, 48 (“I wonder whether smiling is not one of the creations of the Middle Ages”); see also Trumble 2004, 89.

25. Plutarch, Caes. 4; Edwards 1993, 63.

26. The survival of so much Roman writing on oratory—some of which is concerned with how or whether to make the listener laugh (on which see pp. 107–20, 123–26)—may exaggerate the apparent preponderance of joking terms over laughter terms, but there is no reason to imagine that the whole imbalance should be ascribed to this.

27. A piece of popular wisdom rejected by Quintilian: “Potius amicum quam dictum perdendi” (6.3.28). It is possibly echoed by Horace, Sat. 1.4.34–35 (but different versions of the text and its punctuation give a significantly different sense; see Gowers 2012, 161), and by Seneca, Controv. 2.4.13. There are some echoes in modern sloganizing too, but the point is always reversed: “It’s better to lose a jest than a friend.”

28. Cicero, De or. 2.222 (= Ennius, frag. 167 Jocelyn; ROL1, Ennius, unassigned fragments 405–6).

29. “Cato,” Disticha., prol.: “Miserum noli irridere” (likewise “Neminem riseris”).

30. Sonnabend 2002, 214–21, offers a brisk summary of scholarship on these lives; A. Cameron 2011, 743–82, is a fuller and more recent discussion (though underplaying, as most critics do, some of the work’s importance, whatever its fictionality: “trivial . . . product,” 781). The collection was probably produced in the late fourth century CE.

31. SHA, Heliog. 32.7, 29.3 (“ut de his omnibus risus citaret”), 25.2.

32. Sat. 2.1.15–2.2.16.

33. 2.3.1–2.5.9; on the style of these jokes and Macrobius’ possible sources, see pp. 104–5, 130–31, 202.

34. 2.2.16 (antiqua festivitas); 2.4.21 (Augustus’ “Fescennines”); see pp. 68–69.

35. 2.2.10, 2.2.12–13. On Evangelus and Servius, see Kaster 1980, 222–29.

36. Sat. 2.6.6–2.7.19 (avoidance of lascivia, 2.7.1); for mime’s bawdy character in general, see pp. 168–69, 170.

37. AP 7.155; PLM3, 245–46; see further above, p. 169.

38. 2.7.16 (on the blurring of mime and pantomime here, see pp. 168, 170).

39. For an overview, see Bonner 1949; Bloomer 2007; Gunderson 2003, 1–25 (a more theorized account). Spawforth 2012, 73–81, considers the interface between Greek and Roman traditions.

40. Controv. 9.2.

41. Principally, Livy 39.42–43; Valerius Maximus 2.9.3; Cicero, Sen. 42. Briscoe 2008, 358–59, reviews the variants.

42. On the law in this case, see Bonner 1949, 108–9.

43. 9.2.9, 9.2.11.

44. Drunkenness: 9.2.3; slippers: 9.2.25; ioci: 9.2.1; iocari: 9.2.9–10; laughter: 9.2.6.

45. For the erotics of laughter, see pp. 3, 157–59. Halliwell 2008, 491, collects a wide range of instances (in Greek) of sexualized laughter, from the classical to the early Christian period.

46. Another example of (sexualized) laughter as a transgressive irruption into the public official sphere is found in the trial of Maximus, the (likely fictionalized) Roman prefect of Egypt (P.Oxy. 471). The “transcription” of the prosecution speech focuses on Maximus’ relationship with a young boy, whom he included in his official business. One specific accusation is that the boy used to laugh in the midst of Maximus’ clients. See Vout 2007, 140–50 (but note that the text does not claim that the boy was laughing “in the face of his clients,” 148; the point is that he was laughing in the sphere of serious, official business).

47. Controv. 9.2.7.

48. Ars am. 3.279–90 (discussed on pp. 157–59).

49. Aeneid 4.128; discussed by Konstan 1986, careful to acknowledge the problem of reading this as a smile (“the smile, or perhaps it is a laugh,” 18). Though intended for high school students, Gildenhard 2012, 138–39, offers a concise paragraph summing up the main interpretative problems of Venus’ laugh.

50. Ars P. 1–5 (“Humano capiti cervicem pictor equinam iungere si velit . . . risum teneatis?”). The passage is more puzzling than it seems, for the laughable incongruities are in fact standard themes in Roman painting; see Frischer 1991, 74–85; Oliensis 1998, 199–202.

51. Coleiro 1979, 222–29, reviews the main suggestions; more briefly, Coleman 1977, 150–52.

52. Du Quesnay 1977, 37, is unusual in arguing that the singular “parent” here is the father.

53. “Enigmatic” is the euphemism of Nisbet 1978, 70, for the final four lines of the Eclogue.

54. The text has been a matter of dispute since the Renaissance at least, with both Politian and Scaliger advocating what is now the standard reading against the manuscripts, largely on the basis of the parallel passage in Quintilian (Inst. 9.3.8). Just to add to the complexity, the manuscript versions of Quintilian do in fact include the same version of these lines as the Virgilian manuscripts, but Quintilian’s use of this passage as an example of a plural relative (qui) attached to a singular referent (hunc) makes it clear that he had in mind a different text, more or less as modern editors have it. The issues are reviewed by Coleman 1977, 148–49; Clausen 1994, 144 (from which I take the word natural). Note, however, some remaining support for the manuscript reading: for example, F. della Corte 1985, 80.

55. The quotation is from Clausen 1994, 144 (my emphasis); similarly R. D. Williams 1976, 119; Norden 1958, 63 (“Ridere c. acc. heisst überall sonst ‘jemanden auslachen’, nicht ‘ihm zulachen’”). Both Perret 1970, 55, and Nisbet 1978, 77n135, see that this is far too sweeping and cite many counterexamples, including Ovid, Ars am. 1.87.

56. Pliny NH, 7.2, 7.72 (see p. 25), with Norden 1958, 65–67; Nisbet 1978, 70. This modern tradition of seeing the baby’s risus as similar to that of Zoroaster goes back principally to Crusius 1896, 551–53.

57. See, for example, Perret 1970, 55 (“Il ne peut s’agir du sourire de la mère à l’enfant”); the different versions are briefly reviewed by R. D. Williams 1976, 120, and Coleman 1977, 148.

58. Nisbet 1978, 70; words such as tenderness and intimacy (Putnam, 1970, 162; Alpers 1979, 173) recur in these discussions.

59. Whatever the sentimentality, Nisbet is one of the very few translators to stick firmly to the word laugh rather than smile (translations in 2007 reprint of Nisbet 1978).

60. Catullus 61.209–13 (“Torquatus volo parvulus / matris e gremio suae / porrigens teneras manus / dulce rideat ad patrem / semihiante labello”). Modern critics are divided on whether this is merely a close epithalamic parallel (a vague back-reference for Virgil) or a direct source (e.g., Putnam 1970, 163: “borrowed”). Hardie 2012, 216–18, reviews the more general links between this Eclogue and Catullus 61 and 64. We should note that there is no hint of divinity in the laughter of Catullus 61 and that the divinity implied in Theocritus, Id. 17.121–34, a possible inspiration for the final line of the Eclogue, has nothing to do with any laughing baby.

61. Bataille 1997, 60. He continues, “All of a sudden, what controlled the child falls into its field. This isn’t an authorization but a fusion. It’s not a question of welcoming the triumph of man over deteriorated forms, but of intimacy communicated throughout. Essentially the laugh comes from communication” (italics in the original).

62. Parvulescu 2010, 161–62, rightly detects echoes of Virgil in Kristeva’s treatment of the laughter exchanged between mother and child (esp. Kristeva 1980, 271–94).

63. Warner 1998, 348.

64. It is striking that hardly any classical treatment of this text references its role in modern theory—nor, it must be admitted, vice versa. In fact, there is some sorry mangling of the Latin in the nonclassical discussions; for example, “Incipe, puer parvo” in the first printing of Warner 1992 (348; later corrected), introducing yet another ungrammatical scribal error into a complex text.

65. The bibliography on constructing identity and on cultural change in the Greco-Roman world is now immense. In addition to other works cited in the following notes, significant contributions include Millett 1990; Woolf 1994; Goldhill 2001; Dench 2005; Mattingly 2011.

66. Epist. 2.1.156 (“Graecia capta ferum victorem cepit”). As Wallace-Hadrill 2008, 24–25, points out, modern scholars rarely quote the very different view of Ovid, Fast. 3.101–2, whose language alludes to Horace.

67. For examples, see Van Dommelen 1997; Hill 2001, 14, (constellation); Webster 2001, 217–23 (hybridity and creolization); Wallace-Hadrill 2008, 27–28 (bilingualism); Le Roux 2004, 301 (crossbreeding, métissage). The influence (and terminology) of such theoretical and comparative studies as Bhabha 1994, esp. 112–16 (for “hybridity”), and Hannerz 1987 is clear.

68. Wallace-Hadrill 2008. The clearest summary of the arguments is at 17–27, which also offers a punchy critique of some of the currently favorite metaphors while opting instead for the model of bilingualism (and also for a model of Greco-Roman cultural interaction based on the diastolic and systolic phases in the operation of the human heart). Wallace-Hadrill 1998 offers a brisk earlier version of his linguistic (code-switching) analogy.

69. Some sensible reflections on the shared traditions of laughter between elite and nonelite are found in Horsfall 1996, 110–11 (though Horsfall is overall more confident than many about our ability to access Roman “popular culture”).

70. Again, there is a vast bibliography. Significant contributions among the new wave of studies of Greek literature and culture in the empire include Swain 1996 (reflecting on “how the Greek elite used language to constitute themselves as a culturally and politically superior group,” 409); Whitmarsh 2001 (the question is “how ‘the literary’ is employed to construct Greek identity in relationship to the Greek past and the Roman present,” 1–2); Spawforth 2012 (“Where Greek culture was concerned, an ‘imperial style of signalled incorporation’ made clear the ‘pure’ brand of Hellenism that the ruling power sought to uphold as morally acceptable to the Romans,” 271). Konstan and Saïd 2006 includes a particularly useful range of essays.

71. Goldhill 2001; Woolf 1994 (the phrase is also used as the title of Woolf 1998, which focuses on Gaul).

72. Fraenkel 1922 (the English translation, Fraenkel 2007, reviews the impact of the book, on xi–xxii). From a more strictly historical perspective, the work of Erich Gruen has been particularly influential here; see, for example, Gruen 1990, 124–57.

73. Christenson 2000, 45–55; Beard 2007, 253–56.

74. Terence, Eun. 1–45; with Barsby 1999, 13–19; Brothers 2000, 20–26. Terence’s Thraso derives from Menander’s Bias. But the matter is complicated by the fact that there is a character named Gnatho in Menander’s The Toady and another, Strouthias, who seems to be (from the fragments that remain) the inspiration for part of the portrayal of Terence’s Gnatho. Perhaps Terence conflated the two, keeping Gnatho’s name, or perhaps the same character went under two different names in Menander’s play. See further Brown 1992, 98–99; Pernerstorfer 2006, 45–50 (for the arguments that a single character was called by two different names). Pernerstorfer 2009 attempts a major reconstruction of the play, reprising the conclusions of the earlier article; for another, succinct, attempt to summarize the plot, see Gomme and Sandbach 1973, 420–22.

75. Menander, Kolax frag. 3 (= Plutarch, Mor. 57a = Quomodo adulator 13): γελῶ τὸ πρὸς τὸν Κύπριον ἐννοούμενος. Plutarch does not mention the title of the play but does name two of its characters. See Gomme and Sandbach 1973, 432; Pernerstorfer 2009, 112–13. Lefèvre 2003, 97–98, is almost alone (and unconvincing) in believing that these words “have nothing to do with Terence.”

76. Gomme and Sandbach 1973, 432; Brown 1992, 94; Pernerstorfer 2009, 113.

77. Wallace-Hadrill 2008 (lamps: 390–91); Spawforth 2012 (cultural comportment: 36–58).

78. Halliwell 2008, 343–46, 351–71 (with 332–34, clearly summarizes the evidence and impact—including Beckett 1938, 168). McGrath (1997, vol. 1, 101–6; vol. 2, 52–57, 58–61) offers useful discussions of several of Rubens’s versions of Democritus. For Heraclitus, see Halliwell 2008, 346–51.

79. De or. 2.235. He assumes Democritus’ expertise in laughter, not necessarily that Democritus is known as a laugher.

80. “Laughing Mouth” (Γελασῖνος) is Aelian’s term (VH 4.20); Halliwell 2008, 351, 369 (for “patron saint”); Juvenal 10.33–34; see also Horace, Epist. 2.1.194–96.

81. Hippocrates, [Ep.] 10–23 (with text and translation in W. D. Smith 1990). Hankinson 2000 and Halliwell 2008, 360–63, offer clear introductions.

82. [Ep.] 10.1 (ὁ δὲ πάντα γελᾷ).

83. [Ep.] 17.5 (ἐγὼ δὲ ἕνα γελῶ τὸν ἄνθρωπον).

84. The only reference to laughter in a (possibly) authentic surviving fragment of Democritus is 68B107a DK, which states that one should not laugh at the misfortune of others. The earliest explicit reference to Democritus being a renowned laugher himself (rather than an expert) is Horace, Epist. 2.1.194–96.

85. Plutarch, Lyc. 25 (statue); Agis and Cleom. 30 (shrine); Halliwell 2008, 44–49, offers a brief survey of the evidence for Spartan laughter.

86. Plutarch, Lyc. 12, 14.

87. Plutarch, Mor. 217c = Apophthegmata Lac., Androcleidas.

88. A temptation not resisted by David 1989.

89. The Roman-period reconstruction of (and investment in) primitive Sparta is a theme in Spawforth 2012 (e.g., on the traditions of the sussitia, 86–100). In part, this tradition was no doubt the Spartans’ own way of claiming a distinctive identity (happy to provide theme-park reenactments of primitive rituals); in part, it was a literary/discursive phenomenon, as writers of Roman date created a distinctive vision of the Spartan past.

90. Cordero 2000, 228, reviews the possibilities. They suggest that the tradition may go back to the third century, but “rien ne le prouve.”

91. Plutarch, Lyc. 25, cites the Hellenistic historian Sosibios (Jacoby, FGrHist 595F19).

92. Chesterfield 1774, vol. 1, 262–63 (letter of 3 April 1747).

93. Cicero, De or. 2.217, sums it up; Plautus, Pers. 392–95, is a comic version of the hierarchy.

94. Plutarch, Mor. 854c = Comp. Ar. & Men. 4. The cultural complexity is nicely signaled by the fact that Plutarch here not only Hellenizes a Roman term to talk about the Greek dramatist Menander but goes on to compare Menander’s “salt” to the salt of the sea from which Aphrodite was born. The reference at Plato, Symp. 177b, is almost certainly literally to salt rather to than wit.


1. Quintilian, Inst. 6.3.47–49. The force of the pun relies on the particular similarity between quoque and the vocative case of coquus (coque), so “I will vote for you too” is heard as “I will vote for you, cook,” jokingly rubbing in the man’s humble origins. The second pun was at the expense of a man who had been flogged in his youth by his father: the father was constantissimus (completely steadfast), the son varius (“vacillating” or “multicolored,” i.e., black and blue).

2. Quintilian, Inst. 6.3.49. The background and outcome of the trial are discussed by Mitchell 1991, 198–201; Riggsby 1999, 112–19; Steel 2005, 116–31. In pondering this pun, I have canvassed other possible linguistic resonances (with sericus, meaning “silk,” sero, “to bolt or bar,” and sero, “to join or contrive”) but without finding any plausible or pointed result.

3. Rawson 1975, xv; Simon Goldhill, interviewed by an Australian newspaper (The Australian, 24 September 2008) about his ideal ancient dinner party companions, chose Sappho, Hypatia, Aristophanes, Alcibiades, and Phryne, as “that would be more fun than Augustus, Caesar, Jesus, St Paul and Cicero.” I am not so sure.

4. Brugnola 1896 is a nice monument to Cicero “the jokester,” very much in the ancient tradition.

5. Plutarch, Cic. 1 (chickpea), 24 (self-importance), 27 (jokes—ἕνεκα τοῦ γελοίου). Against the man with ugly daughters, he quoted a line of some tragic drama (“It was against the will of Phoebus Apollo that he sired children”). The joke against Faustus Sulla (son of the dictator) rested on a double entendre. He had fallen into debt and issued notices (προέγραψε) advertising his property for sale; Cicero quipped that he preferred the son’s notices to the father’s (Sulla senior had issued notices with lists of those to be put to death—the word προγράφω, or proscribo in Latin, refers to both kinds of notice).

6. Plutarch, Cic. 38.

7. Though written in the form of a speech, this was never actually delivered and most likely was always intended for written circulation only; Ramsey 2003, 155–59.

8. Cicero, Phil. 2.39–40.

9. The possibility (or difficulty) of laughter in times of trouble is a common theme in Cicero’s letters: Att. 7.5.5 (SB 128); Fam. 2.4.1 (SB 48), 2.12.1 (SB 95), 2.16.7 (SB 154), 15.18.1 (SB 213).

10. Comp. Dem. & Cic. 1.

11. Comp. Dem. & Cic. 1 (also quoted at Plutarch, Cat. Min. 21); on possible senses of λαμπρός, see Krostenko 2001, 67–68.

12. “Funny”: Rabbie 2007, 207; “comedian”: Krostenko 2001, 224. Dugan 2005, 108, offers “amusing.” The Loeb Classical Library version of Cat. Min. 21 runs “What a droll fellow our consul is,” and of Comp. Dem. & Cic. 1, “What a funny man we have for a consul.”

13. Inst. 6.3.1–5.

14. Macrobius, Sat. 2.3.10, 7.3.8; Seneca, Controv. 7.3.9. The repartee starts with a gibe by Cicero against Laberius, who had just been given equestrian rank by Caesar and was trying to take his seat in the designated equestrian area—when everyone sat close together so as not to let him in. Cicero quips, “I would have let you in except that I am cramped in my seat” (the implication being that elite rows had become full of any riffraff promoted by Caesar). Laberius retorts, “How strange, given that you usually sit on two seats” (a dig at Cicero’s vacillations of support between Caesar and Pompey). Seneca makes the parallel absolutely explicit: “Both men speak very wittily, but neither man has any sense of boundary in this area.”

15. Sat. 2.1.12 (a phrase here ascribed to Vatinius); with Cicero, Fam. 9.20.1 (SB 193), implying that his friend Paetus had called Cicero scurra veles (a “light-armed scurra,” “the scurra of the troop”), presumably in friendly banter.

16. Other, in my view less likely, suggestions for Cato’s original words include facetus or lepidus (Leeman 1963, 61, 398n100; Krostenko 2001, 225); the quip would then point to the “overaestheticized” implications of those terms, incompatible with the masculine traditions of public speaking and office holding.

17. Inst. 6.3.5. Macrobius, Sat. 2.1.12 notes that some people suspected that Tiro himself had made up some of the jokes.

18. Fam. 15.21.2 (SB 207).

19. Macrobius, Sat. 2.3.3.

20. Macrobius, Sat. 2.3.7. This is a subtler pun than it at first seems, as Ingo Gildenhard has helped me appreciate, playing on the conflict between military preparations and those for a dinner party (see Brugnola 1896, 33–34). As I have translated it, the joke consists in Cicero displacing the life and death issues of civil war by turning to the trivial business of when you should arrive at a dinner party, but the military reading surely remains latent, with nihil . . . paratum also referring to the general lack of preparation of the Pompeians (“Look who’s talking: the state of preparation in this camp is pathetic”). Corbeill’s reading (1996, 186) produces a more frigidus point: “You’ve arrived late in the day” . . . “But not too late, as you have nothing ready.”

21. It was Petrarch in the fourteenth century who established Cicero as a jester for the humanists (Rerum memorandarum Lib. 2.37, 2.39, 2.68), with further discussion in Bowen 1998.

22. Fam. 7.32.1–2 (SB 113). The name (or perhaps it is the nickname) of the correspondent points to the artful wit of this letter, which is as much a joke itself as a comment on the treatment of Cicero’s bona dicta; see further Hutchinson 1998, 173–74; Fitzgerald 2000, 97; Krostenko 2001, 223 (which gives the passage a rather different stress—that Cicero is happy to be credited with the jokes of others, provided they are good ones). Note Cicero’s claim elsewhere that Caesar would be able to recognize which quips were bona fide Ciceronian: Fam. 9.16.3–4 (SB 190).

23. Quintilian, Inst. 6.3.77; Macrobius, Sat. 2.4.16 (Vatinius, in order to show that he had recovered from his gout, boasted that he was now walking two miles a day [in Macrobius, only one]. The retort is: “Yes, I’m not surprised; the days are getting a bit longer”). The slippage and migration of jokes is discussed by Laurence and Paterson 1999, 191–94.

24. Among studies earlier than those I discuss here, note Haury 1955 (focused particularly on irony); Geffcken 1973 (on comic aspects of Pro Caelio), now with Leigh 2004; Saint-Denis 1965, 111–61 (focusing especially on Pro Caelio, In Verrem, and De oratore).

25. Att. 1.18.1 (SB 18)—he can neither joke nor sigh. Hutchinson 1998, 172–99 (quotes on 177); see also Griffin 1995.

26. Richlin 1992a. For the rhetoric of invective and the main coordinates of sexual humor, see 57–104.

27. Freud 1960 [1905], 132–62 (quote on 147); Richlin 1992a, 59–60.

28. Corbeill 1996 (quotes on 5, 6, 53); on the persuasive or reassuring function of jokes and laughter, see also Richlin 1992a, 60 (again drawing on, and developing, a Freudian perspective).

29. Reflected in, for example, Connolly 2007, 61–62; Vasaly 2013, 148–49. Another important strand of work, with a strongly linguistic emphasis, is found in Krostenko 2001 (though his focus on “social performance” offers in many ways a complementary approach to the construction of identity through wit, laughter, and their terminologies). It is important to stress that what sets this “new orthodoxy” apart from some apparently similar earlier approaches (focusing on derision and humorous invective) is the constructive social function (one sense of controlling in Corbeill’s title) it ascribes to laughter.

30. Inst. 6.3.7.

31. The expression of Fantham 2004, 186.

32. In particular, shorter sections in Orat. 87–90 and Off. 1.103–4.

33. Guérin 2011, 151, rightly refers to the provocation of laughter as “une zone de risque”; for Richlin 1992a, 13, it is the use of obscenity rather than the ambivalence of laughter that makes courtroom joking a risky proposition.

34. The first certain reference to De Oratore is in a letter to Atticus of November 55 (when the work is finished enough to suggest that Atticus copy it), Att. 4.13.2 (SB 87).

35. All recent work on this text is underpinned by the five-volume commentary of Harm Pinkster and others, which appeared between 1981 and 2008 (the relevant volume for the discussion of laughter in book 2 is Leeman, Pinkster, and Rabbie 1989), and it can be assumed to be a major reference point in what follows. This edition has largely replaced the earlier commentary of A. S. Wilkins, published between 1879 and 1892 (the relevant volume being Wilkins 1890). The best up-to-date translation, with introduction, is May and Wisse 2001; Fantham 2004 is an illuminating guide to the text and its literary, cultural, and historical significance.

36. At De or. 1.28, the participants agree to “imitate Socrates as he appears in the Phaedrus of Plato” and to sit down under a plane tree for their discussion; see Fantham 2004, 49–77. Although they are, in our terms, oratorical experts, they are keen to distinguish themselves from professional Greek experts (e.g., De or. 1.104).

37. May and Wisse, 2001, 14–15, succinctly introduces the characters; Fantham 2004, 26–48, discusses Crassus and Antonius in detail. Cicero adopts the Platonic device of setting his dialogue just before the death of the lead character (Socrates in Plato’s Phaedo and Crito); here all the characters but one (Cotta) were dead by the end of 87 BCE. The year 91 BCE might be seen a loaded choice: only the year before, Crassus, as censor, had expelled the Latini magistri (Latin teachers of rhetoric) from Rome (De or. 3.93; Suetonius, Rhet. 1).

38. Fam. 1.9.23 (SB 20). Aristotle’s dialogues are almost entirely lost, but they certainly featured much less cut and thrust, and longer expository speeches by the participants. Cicero may also have had in mind Aristotelian content as well as form.

39. See, e.g., R. E. Jones 1939, 319–20; Dugan 2005, 76.

40. In addition to the works already cited, notable recent interventions, often with a particular focus on the section on laughter, include Gunderson 2000, 187–222 (“Love”); Krostenko 2001, 202–32; Dugan 2005, 75–171; Guérin 2011.

41. De or. 2.216–90. In addition to the commentaries noted above, Monaco 1974 offers a text, an Italian translation, and extensive notes on this section of the work alone; Graf 1997, 29–32, offers a succinct discussion.

42. De or. 2.234. This image is taken up again at the end of the section (2.290).

43. De or. 2.217, 2.231, 2.239.

44. De or. 2.216.

45. De or. 2.235.

46. Leeman, Pinkster, and Rabbie 1989, 188–204; Rabbie 2007, 212–15 (a revised, less “speculative,” English version). The earlier tradition is represented by Grant 1924, 71–87, 100–131 (drawing on Arndt 1904). To be fair, it did admit a few Ciceronian additions to or deviations from Greek precedents (“Sed iam abscedere videtur Cicero a fontibus Graecis ac suum tenere cursum,” Arndt 1904, 36, in relation to De or. 2.268.), but the default position was that everything went back to a lost Greek source unless there was overwhelming evidence to the contrary. The old view is still assumed in some popular writing on the subject (such as Morreall 1983, 16) and is more or less revived wholesale by Watson 2012, 215–23, in yet another attempt to pin the Tractatus Coislinianus (see above, p. 31) to Aristotle.

47. De or. 2.217; see also 2.288. These Greek books do not survive; see p. 34.

48. De or, 2.216 (suavis), 2.236 (locus . . . et regio)—though Corbeill 1996, 21–22, nuances the parallels between Aristotelian and Ciceronian terminology.

49. There is an unresolved controversy (conveniently summarized by Fantham 2004, 163–64) around the availability in antiquity of some of the works of Aristotle, and so to which ones Cicero could have had direct access.

50. Leeman, Pinkster, and Rabbie 1989, 188–89.

51. As argued, for example, by Monaco 1974, 29, in relation to the Memmius story of De or. 2.283.

52. De or. 2.2.

53. See p. 54 for the textual confusion between locus and iocus.

54. These veteres could be in theory either Greek or Roman (as Pinkster, Leeman, and Rabbie 1989, 214, makes clear). But the strongly Latin character of the terms makes the latter much more likely, although no doubt versed in Greek theory.

55. It is an even smarter exchange than it might appear. As A. S. Wilkins 1890, 113, and Leeman, Pinkster, and Rabbie 1989, 216, clearly document, “bark” (latrare) was a word used for shrill speakers. Krostenko 2001, 214–15, points to Cicero’s use of the word venustus for “spur-of-the-moment” humor of this kind.

56. Guérin 2011, 271–303, discusses these two antitypes in detail, though suggesting an oversystematic, rigid distinction between the two (the scurra is the antitype of oratorical dicacitas, the mimus of oratorical cavillatio). Grant 1924, 88–96, offers a convenient collection of sources. See further above, pp. 152–55, 167–70.

57. The Latin is hard to pin down: “In re est item ridiculum, quod ex quadam depravata imitatione sumi solet; ut idem Crassus: ‘Per tuam nobilitatem, per vestram familiam.’ Quid aliud fuit, in quo contio rideret, nisi illa vultus et vocis imitatio? ‘Per tuas statuas,’ vero cum dixit, et extento bracchio paulum etiam de gestu addidit, vehementius risimus.” I follow Monaco 1974, 124, here in seeing this as laughter generated by the mimicry (depravata imitatione), with the imitation of the statue (extento bracchio) prompting the most raucous chuckles. Leeman, Pinkster, and Rabbie 1989, 248, argues that the joke rests on the unexpected addition (aprosdokēton) of “per tuas statuas” after “per tuam nobilitatem, per vestram familiam” and that the extended arm is a reference to the position of a man taking an oath. But this interpretation hardly delivers on the mimicry that Cicero emphasizes. See further, p. 119.

58. De or. 2.216; Off. 1.108. Dugan 2005, 105, puts the strongest recent case for seeing Cicero’s choice of Strabo (“whose public persona and oratorical style provoked suspicions that were similar to those which he himself incited”) as significant.

59. Zinn 1960, 43.

60. Fam. 7.32.2 (SB 113).

61. Ingo Gildenhard has suggested to me that the name is significant: at the very least there is something a bit joking in having the disquisition on joking delivered by a man whose name means “squinter.” And just suppose we were to imagine that “Strabo” was a stock comic character; then we might also imagine a running metaliterary joke in the criticism of mime.

62. De or. 2.218 (“leve nomen habet utraque res”).

63. Or. 87.

64. Leeman, Pinkster, and Rabbie 1989, 189, followed by Fantham 2004, 189.

65. Inevitably, the influence of earlier Greek terminology has been sought here. Kroll 1913, 87, for example, sees the Peripatetic terms charis/gelōs behind facetiae/dicacitas (though in this case even Grant [1924, 103–18] is unconvinced and finds no exact Greek equivalent for the pairing).

66. De or. 2.251 (ridicula/faceta), 2.260 (frigida/salsa), 2.222 (bona dicta / salsa).

67. Grant 1924, 100–131, while acknowledging the difficulties, attempts a series of systematic definitions; likewise Leeman, Pinkster, and Rabbie 1989, 183–88 (“Einige Differenzierung zwischen dem Gebrauch der verschiedenen Termini ist . . . möglich, wobei aber Grant . . . manchmal zu weit gegangen ist,” 183), and Guérin 2011, 145–303. Krostenko 2001 offers a highly technical sociolinguistic study of many of these key terms, emphasizing their mutability. Ramage 1973 attempts to track ideas of urbanitas throughout Roman history. Fitzgerald 1995, 87–113, is the clearest introduction to the issues.

68. Krostenko 2001, 207–14.

69. Inst. 6.3.18–19: “Salsum in consuetudine pro ridiculo tantum accipimus: natura non utique hoc est, quamquam et ridicula esse oporteat salsa. Nam et Cicero omne quod salsum sit ait esse Atticorum non quia sunt maxime ad risum compositi, et Catullus, cum dicit, ‘Nulla est in corpore mica salis,’ non hoc dicit, nihil in corpore eius esse ridiculum. Salsum igitur erit quod non erit insulsum.” This passage reveals some of the acute difficulties in translating, let alone in making precise sense of, Roman discussions of wit and its terminology. In the first sentence, is Quintilian saying that salsa ought also to be ridicula, or that ridicula ought also to be salsa? The position of the etstrongly suggests the former, but the explanations that follow (after nam) make the latter almost certain. And what is the sense of ridiculum? Modern translators render Quintilian’s comment on Catullus as “He does not mean there is nothing ridiculous in her body” (D. Russell in the Loeb Classical Library) or “Non c’è niente di ridicolo” (Monaco 1967). It makes perfect sense in English (or Italian), but it ignores the other, active Latin sense of ridiculum—to make you laugh. Catullus could well be saying (as some modern commentators agree; see, e.g., Quinn 1970, 424) “there is not a spark of wit” in her. Throughout the passage there is an instability between the active and the passive sense of these words (as in ad risum compositi). Matters are further confused by the fact that Cicero (De or. 2.251) attempts (tendentiously maybe) to distinguish the salsum of the orator and the mime actor.

70. De or. 2.235. For the reading of venas or genas, see Leeman, Pinkster, and Rabbie 1989, 238.

71. De or. 2.236.

72. De or. 2.279.

73. De or. 2.248.

74. De or. 2.248.

75. De or. 2.254.

76. De or. 2.255, 2.260; see also p. 28.

77. De or. 2.255 (for the financial sense, see Plautus, Rud. 1327).

78. De or. 2.245.

79. De or. 2.252.

80. De or. 2.90–92; though there are dangers even in this kind of imitation, as Antonius points out (you have to make sure that you copy the most important features of the model, not merely those that are easy to imitate).

81. De or. 2.242.

82. See, e.g., Edwards 1993, 98–136 (see 117–19 for the comparison of actors and orators). Dupont 2000 is a subtle discussion of the interrelationships between Roman oratory and theater, as is, more briefly, Fantham 2002 (drawing particularly on Quintilian, Inst. 11.3). See further above, p. 167.

83. De or. 2.251.

84. De or. 2.247, 2.256.

85. Corbeill 1996, 26.

86. De or. 2.262.

87. One classic statement of this “brain-balkanisation” is Feeney 1998, esp. 14–21.

88. Krostenko 2001, 223–25; Dugan 2005, 105–6.

89. Seneca, Constant. 17. Vatinius is here dubbed (like Cicero) a scurra—but also venustus and dicax. “He used to joke about his own feet and scarred neck; in this way he escaped the wit [urbanitas] of his enemies—who outnumbered his deformities—and particularly Cicero’s.”

90. Macrobius, Sat. 2.3.5. Relations between Cicero and Vatinius were more complicated than the terms of simple enmity in which they are often painted. Cicero defended Vatinius in 54 BCE. Even if this was largely under pressure from Caesar and Pompey (see his lengthy explanation in Fam. 1.9 [SB 20]), there are later clear signs of cordiality, in, e.g., Att. 11.5.4 (SB 216); Fam. 5.9–11 (SB 255–59).

91. “Interactive” (as Ingo Gildenhard encourages me to say) is key here, and a feature lost from the necessarily nondialogic character of the speeches as circulated in written form. One might be tempted to say that the aggressive humor is a feature more of the written versions than of the original oratorical scene; that, in writing, invective has replaced the dialogic banter that is so central to the picture of joking in De Oratore.

92. Inst. 6.3 (with Monaco 1967, including Italian translation and notes); Fernández López 2007 is a brief introduction to the work as a whole.

93. Cicero’s account is explicitly referenced at, for example, Inst. 6.3.8 (De or. 2.236), 6.3.42 (Orat. 87).

94. Inst. 6.3.23 (verbo/re), 6.3.26 and 29 (funny faces), 6.3.34 (classes of people).

95. Inst. 6.3.50.

96. De or. 2.267; Inst. 6.3.67.

97. Inst. 6.3.102–12.

98. De or. 2.271 (see also 2.227); Inst. 6.3.19.

99. Inst. 6.3.28.

100. As suggested in another context (see pp. 131, 252n11) by Sherwin-White 1966, 305.

101. Inst. 6.3.82. See above, n. 89, for a scurra, Vatinius, who apparently told jokes on himself to his advantage.

102. Inst. 6.3.112, 6.3.54 (“est enim dictum per se urbanum ‘satagere’”). Martial, Epigram. 4.55.27–29, suggests that foreign place-names could be funny too.

103. Inst. 6.3.8, 6.3.32.

104. Inst. 6.1.48.

105. De or. 2.240–41.

106. Inst. 6.3.6, 6.3.70 (“ridiculum est autem omne quod aperte fingitur”).

107. Phaedrus, Fabulae 5.5; see also John Henderson 2001, 119–28. Here, as Henderson observes (224n70), the phrase urbanus sal signals Roman “show biz.”


1. SHA, Heliog. 26.6, 25.1.

2. Variations on this theme are found in other ancient reflections on the autocrat’s relationship to laughter and joking—in, for example, the story of the young Julius Caesar’s encounter with the pirates. In captivity, Caesar joked with the pirates that when he was free, he would crucify them, which is what he did. Suetonius (Iul. 4; see also 74) underlines the point: he really carried out “what he had often threatened them as a joke” (“quod saepe illis minatus inter iocum fuerat”). The message is that in different ways, the jokes of the powerful could turn out to have a greater truth-value than you might want.

3. Laurence and Paterson 1999 is an important introductory study on the whole theme of emperors and jokes.

4. Nicolaus’ Historiae does not survive complete; this passage is quoted by Athenaeus, Deipnosophistae 6.261c = Jacoby, FGrHist 90F75. Nicolaus was writing in Greek, hence the stress on “native language.”

5. Plutarch, Sull. 2, 36.

6. Succinctly characterized by Le Goff 1993, 26; in a slightly later period, Bowen 1984.

7. See further Laurence and Paterson 1999, 191–94; SHA, Avid. Cass. 2.5–6, a late antique reflection on such migration. In what follows, I hope it goes without saying that “Augustus quipped” is shorthand for “Augustus is said to have quipped.”

8. Dio 65(66).11.

9. Sat. 2.4.3; quoted by Quintilian, Inst. 6.3.59, as an example of raising a laugh by similitudo, or comparison. Other examples of friendly imperial jocularity include Suetonius, Tit. 3.2 (“cum amanuensibus suis per ludum iocumque certantem”); SHA, Hadr. 20.8.

10. Sat. 2.4.19–20. Roughly the same quip is told by Valerius Maximus (9.14 ext. 3), made to a republican governor of Sicily.

11. Ep. 4.25 (picking up a story from Ep. 3.20). The overall sense of the anecdote is clear, but there are some difficulties in the details. One crucial (and awkward) sentence is “Quid hunc putamus domi facere, qui in tanta re tam serio tempore tam scurriliter ludat, qui denique omnino in senatu dicax et urbanus et bellus est?” I have translated this, in common with others, as “What do we imagine that the kind of man who plays around just like a scurra in such a weighty matter and at such a serious moment does at home—when he is so sarcastic, facetious, such a sharp talker even in the Senate?” This would imply that Pliny sees the Senate as no place for the dicacitas, etc., that Cicero admired (and for Sherwin-White 1966, 305, is an illustration of a shift in the culture of wit). But I have wondered if it might rather mean “What do you imagine the man does at home who plays around just like a scurra in such a weighty matter and at such a serious moment yet in the Senate is a wonderfully witty, elegant, and smart speaker?”—implying approval of dicacitas, etc.

12. SHA, Comm. 15.6. See also Suetonius, Cal. 27.4 (a writer of “Atellan farces” burned alive in the amphitheater by Caligula for a dodgy pun, “ob ambigui ioci versiculum”).

13. Claud. 21.5.

14. Suetonius, Cal. 32.3. Suetonius, Cal. 33, repeats a similar quip (“among his various jokes,” when he was smooching around the neck of his wife or mistress, he would say, “What a lovely neck—off it could come just as soon as I give the word”).

15. SHA, Comm. 10.4.

16. Suetonius, Iul. 45.2; Suetonius, Dom. 18.2; Juvenal 4.38 (calvus Nero). Emperors also joked about the baldness of others; Caligula famously, and nastily, ordered a line of prisoners to be executed “from bald head to bald head” (Suetonius, Cal. 27.1; Dio 59.22.3); see also SHA, Heliog. 29.3 (see p. 77).

17. Sat. 2.5.7.

18. Suetonius, Claud. 41.1 (“ne sedato quidem tumultu temperare potuit, quin ex intervallo subinde facti reminisceretur cachinnosque revocaret”).

19. Vesp. 22–23 (compare, for example, Cicero, De or. 2.236, 2.257). The specter of inappropriate wit also hovered over the emperor Augustus. We might, for example, wonder how far the adverse side of the mime was to be seen in his last words as reported by Suetonius (Aug. 99.1): Had he played his part well, he asked, in the mime of life?

20. Suetonius, Cal. 24.2; the classic account of Xerxes at the Hellespont is Herodotus 7.33–35.

21. Suetonius, Cal. 27.4; Seneca, De ira 2.33.3–5 (without specific reference to laughter).

22. Aug. 98. Wallace-Hadrill 2008, 38–41, discusses other aspects of this passage.

23. Dio 59.26.8–9. A story told of Hadrian, as of other rulers, focuses on his encounter with an ordinary woman he passed on a journey and points in a similar direction. In Dio’s account (69.6.3). she tries to waylay him with a request, but he brushes her off, saying that he does not have time. Her retort, however, turns him in his tracks: “Don’t be emperor, then.” The simple idea was that the emperor ought to have time for the humble and that the humble could answer back. This is discussed (with the parallels) by Millar 1977, 3–4.

24. SHA, Hadr. 17.6–7.

25. Met. 2.676–707. Barchiesi 2005, 295, compares this with the encounter between Athena and Odysseus at Homer, Od. 13.287, where Athena is said to “smile” (μειδιᾶν). He admits that it “develops very differently” (“lo sviluppo sarà ben diverso”)—so differently, I would suggest, that it points to the very different significance of ridere and μειδιᾶν.

26. Met. 9.306–23.

27. Met. 5.662–78. As Stephen Halliwell has suggested to me, there is a similarity between the sound of some of these creatures and human laughter, or it is easy enough to imagine one; for hearing the sound of crows (in the same family as magpies) as laughter, see Halliwell 2008, 3.

28. Unsurprisingly, Ovid’s work is a treasure chest of clever comments and reflections on and around laughter both human and divine. We shall focus on some more of these in the next chapter (see pp. 157–59). For more on divine laughter (as well as the misfit between the Greek μειδιᾶν and the Latin vocabulary of laughter), see Ovid, Fast.4.5–6, with the parallels in Ennius and Lucretius noted by Fantham (1998, 91), though she treats ridere here as unproblematically “smile.”.

29. The “clever slave” of comedy is usefully discussed by Fitzgerald 2000, 10–11, 24–26, 44–47, and McCarthy 2000, esp. 211–13.

30. The most convenient edition of this text is Perry 1952, 35–208 (from which I take my references, with G and W indicating the different manuscript versions). For a translation, see Lloyd Daly in Hansen 1998, 111–62; Jouanno 2006. The complexities of the manuscript and papyrological tradition and the questions of cultural background are summarized succinctly by Hopkins 1993 (esp. 11) and in greater detail by Kurke 2011, 1–49 (including an excellent review of the secondary literature). In general, Kurke is more inclined than I am to identify earlier Greek traditions in the Life rather than to stress the Roman surface detail (such as monetary denominations; see Vita Aesopi W 24, 27); Pelliccia 2012 also resists Kurke’s intention to “frog-march the evidence backward” (40).

31. Note the carefully agnostic comments of Kurke 2011, 13 (citing further references to the ongoing debate on the “real existence” of Aesop).

32. Hopkins 1993, 13; Vita Aesopi G 1; Vita Aesopi W 1.

33. Vita Aesopi G 7 (in W 7, the goddess concerned is Tyche).

34. Vita Aesopi G 2–3; W 2–3, with Kurke 2011, 191–92. Kurke also points to other cultural roles of mutism in this text: for example, as a signal of social exclusion (162–63) or an analogue of fabular speech (201). Figs are also prominent in various laughter stories discussed above, p. 177.

35. Vita Aesopi G 24; W 24 (with reference to the “turnip” not in G).

36. Vita Aesopi G 25–27; W 25–27.

37. Freedom: Vita Aesopi G 90; W 90; death at Delphi: G 140–42; W 140–42. Kurke 2011, 53–94, fully discusses the critique of Delphic authority that the story implies.

38. Vita Aesopi G 36; W 36.

39. Inst. 6.3.71. The original Latin does not quite say “stupid” at the end, as in the English idiom of such gags, but it very nearly does: “Stulte interrogaverat exeuntem de theatro Campatium Titius Maximus an spectasset. Fecit Campatius dubitationem eius stultiorem dicendo: ‘<non> sed in orchestra pila lusi.’”

40. Baths: Vita Aesopi G 38; lentil(s): G 39–41; W 39–41.

41. Philo, Leg. 349–67.

42. Smallwood 1970, 3–50, discusses the historical background and the literary tradition of the Legatio. Conybeare 2013, 28–39, discusses the stress on laughter in Philo’s philosophical and theological works.

43. Stackelberg 2009, 135–40, explores the physical context of the meeting between the emperor and the envoys.

44. Leg. 349–59; mime: 359 (καὶ γὰρ τὸ πρᾶγμα μιμεία τις ἦν ). Smallwood 1970, 321–22, collects other references, in Philo and elsewhere, to the mocking of Jews being compared to mime, though she is carried away by the idea that some ancient figurines that may represent mime actors possibly have a distinctively Jewish physiognomy. The vocabulary at 351 and 368 also signals this episode as “theatrical” in a more general sense.

45. Leg. 361: πάλιν πρὸς τὴν πεῦσιν γέλως ἐκ τῶν ἀντιδίκων κατερράγη τοσοῦτος, τῇ μὲν ἡδομένων τῇ δὲ καὶ ἐπιτηδευόντων ἕνεκα κολακείας ὑπὲρ τοῦ τὸ λεχθὲν δοκεῖν σὺν εὐτραπελίᾳ καὶ χάριτι εἰρῆσθαι, ὥς τινα τῶν ἑπομένων αὐτῷ θεραπόντων ἀγανακτεῖν ἐπὶ τῷ καταφρονητικῶς ἔχειν αὐτοκράτορος.

46. Leg. 361. As Smallwood 1970, 322, puts it, if this was the rule, “Dio and Suetonius know nothing of this.”

47. Leg. 362–67.

48. Inst. 6.3.58 (the standard modern text simply draws from Horace’s account in Sat. 1.5 to fill the obvious gap in what has survived of Quintilian).

49. Martial, Epigram. 1.101. Plutarch, Mor. 760a (= Amat. 16), recounts a joking encounter between Gabba (called a γελωτοποιός) and Maecenas; see also Quintilian, Inst. 6.3.27, 6.3.80 (6.3.62 may also refer to Gabba).

50. Tacitus, Ann. 15.34: “Vatinius inter foedissima eius aulae ostenta fuit, sutrinae tabernae alumnus, corpore detorto, facetiis scurrilibus.”

51. The sense of copreae might be rather “found on the dung heap” (from κοπρία, “dung heap”), but I have been unable to resist “little shits.”

52. I include in this the “courts” of rivals or enemies; Dio (in a speech of Octavian) refers to the table companions of Antony and Cleopatra being called κοπρίαι (Dio 50.28.5).

53. Dio 74(73).6.

54. Tib. 61.6.

55. Suetonius, Claud. 8.

56. Pliny, HN 37.17; Seneca, Ben. 2.12.1. Caligula was said to wear them—see Suetonius, Cal. 52: “socco muliebri.”

57. Soccus could, in fact, be used as a metonym for comedy, as cothurnus (buskin) was for tragedy; see Horace, Epist. 2.1.174; Ovid, Rem. am. 376. For a parasite’s soccus, see Plautus, Persa 124.

58. I am aware that there may seem to be something risky about assuming that Suetonius’ account is much closer than that of the SHA to the reality of court life. But it’s not too risky. Suetonius had inside experience of the Roman palace (Wallace-Hadrill 1983, 73–96), and the use of the term copreae in different contexts and writers implies a recognizable referent. It is, as I have been suggesting, another case where these late imperial biographies hit the spirit if not the fact of Roman imperial life.

59. CIL 6.4886 (= ILS 5225): “. . .] Caesaris lusor / mutus argutus imitator / Ti. Caesaris Augusti qui / primum invenit causidicos imitari.” The fullest and most acute recent discussion is Purcell 1999 (who, however, prints the text as “mutus et argutus”).

60. Wallis 1853, 79–80.

61. Argutus on its own is a term that is more generally associated with the repartee of the Roman wit or jokester; see, for example, Plautus, Truculentus 491–96.

62. Garelli 2007, 251; a late antique glossary defines a female pantomime actress as “omnium artium lusor” (CGL 5.380.42); Petronius, Sat. 68, has perhaps a similar household “imitator.”

63. Laes 2011, 470, evades the problem by punctuating differently, to read “Mute and bright imitator. Of the household of Tiberius.” But the isolated phrase “Of the household of Tiberius” is very awkward, even by the standards of this awkward Latin.

64. Purcell 1999, 182–83, reviews various possible settings (including public performance), but the repeated stress on the emperor in this text strongly suggests that we are dealing principally with a court entertainer.

65. See, for example, Pliny, Ep. 3.1.9, 9.17; with further references and discussion in C. P. Jones 1991 and Dunbabin 2008.

66. Ep. 50 (esp. 2). Pliny, Ep. 5.19, also concerns a resident household comedian; similarly, Petronius, Sat. 68 (n62).

67. Barton 1993, esp. 107–8 (“What did the Romans see in the mirror of deformity?”) and 141 (Seneca’s Harpaste as a “freakish avatar” of the elite philosopher). This is an extremely powerful discussion (also linking the mimes I will be treating in the next chapter); in general, however, Barton stresses the roles of derision and monstrosity more strongly than I think plausible.

68. Vesp. 19.2.

69. Suetonius, Iul. 51. See also Suetonius, Iul. 49.4; Dio 43.20; and the discussion in Beard 2007, 247–49.

70. The clearest ancient example of laughter presented in these terms is found in the Greek story of Baubo, who exposes her genitals and makes the mourning Demeter laugh; it is explicitly called apotropaic by, for example, Zeitlin 1982 (145). For further references and brief discussion, see above, p. 174.

71. The “evil eye” is far too catchall a solution to be useful; see further Beard 2007, 248.

72. Barton 1993, 140, briefly discusses Vespasian’s funeral (though not the triumph)—seeing the joker along these lines, as “the monstrous double” of the emperor.

73. For example, Juvenal 5; Martial, Epigram. 2.43, 3.60, 4.85; Pliny, Ep. 2.6. Gowers 1993, 211–19, discusses the ideology and the practice of such inequalities.

74. SHA, Heliog. 25.9.

75. Petronius, Sat. 49, raises all kinds of questions about food and deception. Apicius’ “patina of anchovy without anchovy” is a more mundane case (4.2.12).

76. D’Arms 1990 is a useful overview of the general paradoxes of equality and inequality of the convivium; further aspects are discussed by Barton 1993, 109–12; Roller 2001, 135–46; Roller 2006 (for the hierarchies implied by posture), esp. 19–22, 85–88, 130–36.

77. The most acute discussions of this particular area include Roller 2001, 146–54 (focusing on verbal exchanges witty and otherwise at the dinner party), and Damon 1997, an important study that lies in the background of much of my exploration in the pages that follow.

78. I am borrowing here Lévi-Strauss’s famous phrase, for which see Lévi-Strauss 1997 [1965].

79. Schlee 1893, 98.18–21.

80. Damon 1997, 1–19, is a good introduction, with further bibliography, to some of the main debates about parasites; 23–36 sketches the main characteristics of the figure; 252–55 summarizes her key conclusions on the “sites of discomfort” (255) in the institution of patronage. Other useful recent discussions of different aspects of the parasite, and his cultural origins, include Nesselrath 1985, 88–121; J. C. B. Lowe 1989; Brown 1992; J. Wilkins 2000, 71–86; Tylawsky 2002; König 2012, 242–65.

81. Xenophon, Symp. 1.11–16, and, for example, 2.14, 2.20–23, 4.50. Halliwell 2008, 139–54, is a sharp discussion of different modes of laughter throughout this work, rightly stressing the role of mimicry and questioning quite how uninvited we should imagine Philip to be (143–55). Huss 1999, 104–6, lists numerous close—or not so close—ancient parallels.

82. Damon 1997, 37–101, reviews these plays. Maltby 1999 discusses four particular characters (from Plautus’ Menaechmi, Captivi, Persa, and Stichus). How far we are meant to identify significantly different types in this repertoire of characters—to distinguish, say, the “parasite” from the “flatterer”—is anything but certain; I have not here attempted to delineate any precise calibration of these hungry, flattering jokesters.

83. Arnott 1972 remains one of the best, most sympathetic introductions to the play—and to the role of its parasite.

84. Stich. 221–24: “logos ridiculos vendo. age licemini. / qui cena poscit? ecqui poscit prandio? / . . . ehem, adnuistin? nemo meliores dabit.” Logi is a loan word whose Greek associations may have remained strong (see also ll. 383, 393), but later in the play (l. 400) the Latin dicta is used as an exact equivalent for these jokes.

85. Stich. 454–55: “Libros inspexi; tam confido quam potis, me meum optenturum regem ridiculis logis.” For the role of jokebooks, see above, pp. 201–5.

86. Ridiculus: Stich. 171–77 (whose precise order is uncertain), 389. Catagelasimus: Stich. 630 (the slightly awkward translation brings out the point). Ritschl 1868, 411, asserts that ridiculus never holds a passive sense in this period (“non sit is qui risum movet invitus, sed qui iocis et facetiis risum dedita opera captat”), a view widely followed (by, e.g., Maltby 1999). This seems to me highly implausible and—by missing the subtlety signaled in l. 630—reduces the Stichus to the uninteresting play it has been taken to be. (See the damning comments on it summarized by Arnott 1972, 54.) Bettini 2000 reaches similar conclusions on Gelasimus to my own, by a different route (see esp. 474); Sommerstein 2009, despite an apparent zeal to oversystematize laughter in Aristophanes, also points to some of these ambivalences.

87. It provides, for example, the main subject of a long essay by Plutarch: Mor. 48e–74e (= Quomodo adulator).

88. Seneca, Ep. 27.5–7: “Habebat ad pedes hos [servos], a quibus subinde cum peteret versus quos referret, saepe in medio verbo excidebat. Suasit illi Satellius Quadratus, stultorum divitum arrosor et, quod sequitur, arrisor et, quod duobus his adiunctum est, derisor, ut grammaticos haberet analectas.” Satellius’ quip (that he should have “scholars to gather up the bits”) appears to work by pushing further the idea of the commodification of knowledge and its relation to the slave economy: analecta was the title of the slave whose job it was to pick up crumbs around the dinner table, here imagined as scholars picking up the dropped crumbs of the host’s quotations. Roller 2001, 148–49, briefly discusses the passage, linking the three terms rather differently. Similar connections underlie a clever (but usually overlooked) pun in Juvenal 5. This poem sends up a dysfunctional dinner party where a client puts up with the humiliation of his status, to the scorn of the satirist. Toward the end, we learn what scraps the client is to be served, in contrast to the lavish food of his host. They include semesum leporem—or “half-eaten hare,” as the commentaries explain (from lepus, leporis). But, of course, that leporem could also come from a word we noted in the last chapter among the vocabulary of joking: that is lepos, leporis (wit or joking). So on the client’s menu may be half-eaten hare, but it could also be a half-eaten joke. A nice illustration of the overlap between laughter and hierarchical banquets!

89. Athenaeus, Deipnosophistae 6.234c–262a; sympathetically discussed by Whitmarsh 2000, with reference to the wider Greek (prose) tradition of parasites and flatters.

90. 6.248d–f.

91. 6.252d.

92. 6.249e.

93. Green 2006, 1–47, is a clear introduction to the work (though Green’s interests focus on Diodorus’ account of the fifth century BCE); Stylianou 1998, 1–139 (specifically on the early fourth century BCE), has greater detail.

94. Diodorus Siculus 34/5.2.8–9. Sources for the Sicilian slave revolts and brief discussion can be found conveniently in Shaw 2001.

95. Suetonius, Tib. 57.2.

96. Nat. D. 1.93: “Latino verbo utens scurram Atticum fuisse dicebat.” The passage has caused critics considerable trouble (see, for example, Dyck 2003, 177), but the basic point (often missed) is that it almost certainly exposes an untranslatable difference between the Greek and the Roman idiom of laughter (while paradoxically seeing Socrates in distinctively Roman terms). I say almost certainlybecause (as Stephen Halliwell reminds me) if Zeno was addressing an audience including Romans (such as Cicero), he may have adjusted his vocabulary accordingly.

97. Fraenkel 1922 (pinpointed in Fraenkel 2007, xiii).

98. Corbett 1986 collects many of the wide-ranging citations, but he struggles (probably fruitlessly, as I shall suggest) to impose any clear explanatory structure on the sometimes bafflingly varied usages of the word scurra (and his efforts certainly did not impress Don Fowler: “It is almost a model of how not to go about an investigation of this kind” [1987, 90]). By far the sharpest discussions I know are Barton 1993, for which the scurra is part of the repertoire of elite antitypes in Rome; Habinek 2005, 182–85, stressing the scurra as a category of anxiety.

99. See, for example, Plautus, Trin. 199–211; Curc. 296–97 (assuming the servi of the scurrae are like their masters); Most. 15–16.

100. SHA, Heliog. 33.7; Corbett 1986, 73.

101. On this view, the wide range of usages of the term reflects the range of boundaries that could be laid, in different places, between the proper and improper practice of laughter at Rome—hardly now recoverable.

102. Palmer 1989 and M. Roberts 1993 give useful overviews of these poems.

103. Conybeare 2002, 197–98, explains how critics have tried to get rid of the word iocantur, which has an impeccable manuscript tradition.

104. Conybeare 2002.


1. Macrobius, Sat. 2.5.

2. Sat. 2.5.9.

3. Julia’s jokes are the subject of Long 2000 (especially the Macrobian context) and Richlin 1992b (with a discussion of her life). The text signals, without explicitly mentioning, Julia’s fate: the account is tied to her “thirty-eighth year” (2.5.2), that is 2 BCE, the year of her exile to Pandateria. The different phases of her exile, in conditions of varying severity, are reviewed by Fantham 2006, 89–91.

4. Carter 1992, 190.

5. I am referring here not just to moments when a woman laughs (or women laugh) at a man (or men) but when she laughs, in a gendered role, as a woman, at a man (which is what, in its powerful and positive valuation, the giggle signifies). Halliwell’s prostitutes (2008, 491) and most uses of κιχλίζειν do not quite match this, though Theocritus, Id. 11.77–78 (girls giggling at the unfortunate Polyphemus), comes close; in Latin, Horace, Carm. 1.9.22, is rather further away.

6. Carter 1992, 189 (she continues on 190: “To reproduce this giggle, a man must identify with a woman rather than with another man and perceive some aspects of male desire as foolish”).

7. Ars am. 3.279–90 (“Quis credat? Discunt etiam ridere puellae,” 281). Martial, Epigram. 2.41, explicitly looks over his shoulder at Ovid (the Paelignian poet) in ridiculing Maximina, a girl with three black teeth: “Ride si sapis, o puella, ride / Paelignus, puto, dixerat poeta.” The quotation “Ride . . .” is probably a loose allusion to this passage of Ars Amatoria rather than taken from a lost Ovidian poem; see Cristante 1990; C. Williams 2004, 150–51.

8. Gibson 2003, 211, lists various passages in Latin where lacuna is used for other types of “bodily hollows.” Martial, Epigram. 7.25.6, uses gelasinus (a transliteration from the Greek) for “dimple.” But in general, dimples are not major players in Roman literary culture.

9. Gibson 2003, 212.

10. I follow the reading and punctuation of Gibson 2003, 60 (with 212–13)—“est quae perverso distorqueat ora cachinno; / risu concussa est altera, flere putes; / illa sonat raucum quiddam atque inamabile: ridet / ut rudet a scabra turpis asella mola” (ll. 287–90)—though none of the uncertainties affect the main point of my argument here.

11. Critchley 2002, 29. Critchley’s observations in this section (25–38) have influenced some of the main themes of this chapter, in particular his stress on the role of humor at and across the boundaries between the human and the animal (“Humour explores what it means to be human by moving back and forth across the frontier that separates humanity from animality, thereby making it unstable,” 29). As I hope to show, Roman writing strikingly foreshadows this major point.

12. Ars am. 3.283.

13. Lucretius 6.1195; Suetonius, Claud. 30.

14. Met. 1.640 (where rictus is a convincing emendation for the manuscript ripas), 1.741. This is a repeated image in the poem: see, for example, 2.481 (the beautiful face of Callisto deformed by a lato rictu on her transformation into a bear), 13.568 (Hecuba on the cusp of transformation into a dog “rictuque in verba parato latravit”). The thirteenth-century pseudo-Ovidian De Vetula picks up the animality of the rictus: “Rictus ei, non risus inest, et sacrificari / Deberet certe potius quam sacrificare” (2.148–49); a rictus belongs to the sacrificial animal, not the human sacrificer. See also Miller 2010, 15, 150.

15. The poem is discussed as a literary play on the traditions of flagitatio by Fraenkel 1961; Selden 2007, 524–27. Goldberg 2000; 2005, 108–13, stress its comic legacy.

16. Translators and critics differ on the precise point of comparison between the dog and the woman. Most take it, as I have, to refer to the facial distortion; a few stress instead the sound of yelping, taking os as “mouth” rather than “face”: “with the noisome yap of a Gallic hound,” as Selden renders it (2007, 525). For the rictus of dogs and possible points of comparison with human laughter, see Lucretius 5.1063–66; Plautus, Capt. 485–86; Apuleius, Apol. 6 (discussed by Tilg 2008, 113–15).

17. For this popular usage—eliding the different species and subspecies, the tailed and the tailless, the chimps, baboons, gorillas, and other simians —I must apologize to primatologists. Scientists (modern and ancient) identify a wide variety of different characteristics and crucial distinctions. In particular, monkeys and apes belong to different scientific families (apes being hominoids; monkeys being either Cercopithecidae, Cebidae, or Callitrichidae). But these technical distinctions do not significantly impact on day-to-day debates and representations.

18. The title of this section is borrowed from Connors 2004; it was too good to miss (and is not wholly unparalleled in antiquity: see n. 24).

19. Athenaeus, Deipnosophistae 14.613d.

20. See pp. 46–47 for “laughing” primates (and “laughing” rats).

21. Connors 2004 is the most up-to-date and sophisticated study of Roman ideas of apes (summing up, on 179, their perennial fascination: “Our human shape is replicated in them but also [from one point of view] distorted: wild, hairy, they meet our gaze across an unbridgeable divide between human and animal, nature and culture”). McDermott 1935; 1936; 1938 are still useful points of reference. All these provide an important background to the rest of this section. For “ape lore” in later periods and the cultural construction of modern primatology, see Janson 1952; Haraway 1989; De Waal 2001. Although chimpanzees’ tea parties may be a thing of the past, the use of primates higher up the cultural food chain is alive and well: see, e.g., Self 1997, a satiric novel in which human beings have been changed into chimpanzees.

22. Pindar, Pyth. 2.72–75. I am skating over some of the difficulties of this “critic-bedevilled sentence,” on which see C. Carey 1981, 49–55 (quote on 49).

23. In addition to McDermott 1935 and 1938, Demont 1997 and Lissarrague 1997 assemble and discuss a wide range of classical Greek references to the habits of monkeys; for those in comedy in particular, see Lilja 1980. As these studies show, the stereotype of the monkey in classical Greece is not restricted to imitation and deception but also includes, for example, ugliness, low birth, and ferocity.

24. Aristophanes, Eq. 887–90. The context is some political banter in which two rivals are trying to bribe Dēmos, the personification of the Athenian people, with a cloak. The repartee shows that the reference to the monkey signals both mimicry (“No, I’m only copying your ways, as a man at a drinking party might when he borrows another man’s slippers to go and have a crap”) and flattery or bribery (“You’re not going to out-toady me”). Sommerstein 1981, 93, 191, misses some of the point, which is seen by Neil 1901, 127, and Demont 1997, 466. Suda, s.v. πιθηκισμοῖς περιελαύνεις, points explicitly to the various possible significances of “monkey business” here: trickery, flattery, and imitation.

25. Phrynichus, frag. 21 (Kassel and Austin). The best guess is that the final “monkey” would have been a sycophant (see also Demosthenes, De cor. 242; Aristophanes, Ach. 904–7).

26. Summed up briskly at Connors 2004, 183–84, 189. Isidore, Etym. 12.2.30, refers to the etymology but insists that it is false. The Greek pairing of πίθηκος (monkey) and πιθανός (persuasive) could open up other related possibilities, puns, and associations.

27. Cicero, Nat D. 1.97 (Ennius, Satir. frag 69 [Vahlen] = ROL2 Ennius, Satir. 23). The pun works despite (or because of) the fact that the first i in similis is short, in simia long. Other examples of such wordplay include Ovid, Met. 14.91–98; Martial, Epigram. 7.87.4; Phaedrus, Fabulae 4.13.

28. Connors 2004, 189–99, 202; briefly, hitting the nail on the head, John Henderson 1999, 34.

29. Lissarrague 1997, 469.

30. Sat. 1.10.18; with Gowers 2012, 316–17.

31. Aelian, NA 5.26 (see also 6.10); for the snares, see 17.25 (with Diodorus Siculus 17.90.1–3—though in a nice inversion of teaching and learning, Diodorus claims that the monkeys taught the hunters this trick). It is noteworthy that Aristotle’s main discussion of apes and monkeys (HA 2.8–9, 502a16–b26) does not stress their capacity for mimicry.

32. A. King 2002, 433–34, reviews the representations of monkeys, etc., at Pompeii and includes a brief discussion of those I refer to here; McDermott 1938, 159–324, is a comprehensive catalogue of images of simians in all media from the classical and preclassical Mediterranean world.

33. M. Della Corte 1954, 210n498 (it is now lost).

34. From the House of the Dioscuri (6.9.6–7); see PPM 4.976, no. 225. It is not impossible that there were such performing monkeys in Pompeii, as the discovery there of a simian skeleton hints (Bailey et al. 1999).

35. Often the image of the escape of Aeneas is discussed alone, but de Vos 1991, 113–17, makes clear the link between it and the image of Romulus; followed by J. R. Clarke 2007, 151–52. For dog-headed baboons (cynocephali), see McDermott 1938, 4–13, 35–46.

36. Brendel 1953.

37. McDermott 1938, 278–80; J. R. Clarke 2007, 153–54 (“comic resistance”). Cèbe 1966, 369–70, lists further explanations.

38. Plutarch, Mor. 64e (= Quomodo adulator 23). Plutarch elsewhere—Mor. 60c (= Quomodo adulator 18)—casts the mythical simian Cercopes as flatterers, again eliding monkey, laughter, and flattery. Hercules carried off this mischievous pair of creatures, upside down, hanging over his shoulder, after they tried to steal his weapons. In the longest, late version of the story (ps.-Nonnus, Comm. in IV Orationes Gregorii Naz. 4.39, of the sixth century CE; with Nimmo Smith 2001, 29–30), they start to discuss his “black arse”—and Hercules bursts out laughing and lets them off. For the complex tradition of the Cercopes (who in some versions gave the name to Pithecusae, modern Ischia), see Marconi 2007, 150–59; note also Woodford 1992; Kirkpatrick and Dunn 2002, 35–37; Connors 2004, 185–88.

39. Phaedrus, Fabulae 4.14; acutely discussed by John Henderson 2001, 180–86. The text survives largely in a medieval paraphrase.

40. Athenaeus, Deipnosophistae 14.613d.

41. Lucian, Piscator 36; the anecdote is included as a fable in Perry’s collection (1952, 504, no. 463).

42. Strabo, Geographica 17.3.4 (= Posidonius, frag. 245. [Kidd]).

43. De usu part. 1.22 (Helmreich) = 1, pp. 80–81 (Kuhn).

44. I am half tempted to see this phrase also proleptically; that is, “the ape imitates for the worse.”

45. De usu part. 3.16 (Helmreich) = 3, pp. 264–65 (Kuhn).

46. Horace, Ars P. 1–5, might (almost) count as another.

47. For further discussion, see pp. 119–20.

48. Fantham 1988. The influence of mime on particular authors and genres is discussed by, for example, McKeown 1979; Wiseman 1985, 28–30, 192–94; Panayotakis 1995, xii–xxv (summarizing the main theme of the book).

49. The modern literature on Roman mime is now very large. Panayotakis 2010, 1–32, is a useful résumé with copious bibliography; Bonaria 1955–56 collects fragments and testimonia; some of Webb 2008, 95–138, is relevant to earlier periods of the Roman Empire. On women, see Webb 2002; Panayotakis 2006.

50. The essays in E. Hall and Wyles 2008 give a good coverage of the debates about ancient pantomime. A standard list of the features supposed to distinguish ancient mime from pantomime is summarized by Hall 2008, 24. But Wiseman 2008 draws attention to the overlap between the two. As Panayotakis crisply sums it up, “The boundaries demarcating mime from pantomime were not always as clear as some scholars, seeking to impose order on inherently diverse and contradictory source materials, have liked to imagine” (2008, 185).

51. De or. 2.251 (“. . . non ut eius modi oratorem esse velim, sed ut mimum”).

52. Marshall 2006, 7, and Manuwald 2011, 183, offer the standard view; Panayotakis 2010, 5–6, is more cautious. Hunter 2002, 204–5, discusses the character of the sannio.

53. Tertullian, Apol. 15.3. Plautus, Truculentus 594, suggests that masks did not necessarily preclude the idea of facial expression; however, Athenaeus, Deipnosophistae 10.452f, is rather better evidence for an unmasked tradition in mime. Richter 1913 chooses (overconfidently) to identify grotesque figurines as mime actors because they have no masks.

54. Note that according to Servius (see below, n. 57), even Cicero—whatever his expressed disdain—went to watch the mime actress Cytheris.

55. Macrobius, Sat. 2.7.1–5; with Barton 1993, 143–44, who sees the story of Laberius as part of Rome’s “physics of envy.”

56. The most extreme case is the so-called Charition mime (P.Oxy 413; Cunningham 1987, app. no. 6; the date is uncertain but sometime before the 200s CE, which is the date of the papyrus).

57. Aulus Gellius 16.7.10 refers to the vulgar vocabulary of Anna Peranna; Panayotakis 2008, 190–97, discusses Virgilian renderings in mime, e.g., Servius ad Ecl. 6.11—the particular performer is elsewhere (Cicero, Phil 2.20) called a mima. Panayotakis imagines the performances were relatively straight. I wonder . . . I am likewise more skeptical than most about how far we can hope to identify precise roles for those known as “first mime,” “second mime,” etc.

58. Walton 2007, 292.

59. Panayotakis 2010, 1; Fantham 1988, 154 (“Best defined negatively. Whatever did not fit the generic categories of tragedy or comedy, Atellane or the Italian togate comedy, was mime”).

60. Philistion: AP 7.155 (there are numerous scattered references to “Philistion” in the context of mime—e.g., Martial, Epigram. 2.41.15; Ammianus Marcellinus 30.4.21; Cassiodorus, Var. 4.51; it may have been a common stage or pen name); Vitalis: PLM 3.245–46.

61. For example, Choricius, Apologia mim. 31–32 (at the mimes, Dionysos takes pity on human beings and is “so generous . . . as to prompt laughter of every kind”), 93 (“Humanity shares two things with the divine: reason [or speech] and laughter”). For a clear recent review of this text (with earlier bibliography), see Malineau 2005; Choricius is important to Webb’s (2008, 95–138) discussion; Bowersock 2006, 61–62, notes similar gelastic themes in contemporary Syriac defenses of mimes.

62. The close link between mimicry and Roman laughter is emphasized by Dupont 1985, 298–99 (in the context of a wider discussion of mime, 296–306), which likewise distinguishes these aggressive forms of imitation from mimesis more generally.

63. Csapo 2002 reviews some of the main issues and includes a good discussion of Aristotle’s anecdote about the fifth-century actor Callippides (Poet. 26, 1461b34–35), attacked for being a “monkey.” As Csapo rightly insists, the criticism did not rest on the fact that he acted with “exaggerated gestures”; his crime was not overacting in our sense but rather “imitating actions that are best not imitated at all” (128), including, in Aristotle’s words, those of “the inferior” and of “lower-class women” (Poet. 26, 1462a9–10). Csapo draws a clear and useful distinction between this mimicry and more general issues of tragic mimesis.

64. Note also the mimicry implied by Suetonius, Cal. 57.4—discussed in terms of the (imitative) roles of the different actors in the mime company by Kirichenko 2010, 57; our lawyer imitator (see p. 144) might fit under this general heading too.

65. GLK 1.491.13–19; Evanthius, Excerpta de comoedia (Wessner) 4.1.

66. Lee 1990, 43; Godwin 1999, 67; Whigham 1966, 100; Quinn 1970, 217 (“The mimae were the cinema stars of the ancient world. . . . Her pout looks like a dog showing its teeth”).

67. On the overall articulation of the full plot of the novel (of which only a small section survives), see Schmeling 2011, xxii–xxv; Sullivan 1968, 45–53, discusses the (irresolvable) problems of the ordering of this particular section.

68. Plaza 2000, 73–83.

69. Sat. 18.7–19.1 (“Complosis deinde manibus in tantum risum effusa est ut timeremus. . . . Omnia mimico risu exsonuerant”).

70. Branham and Kinney 1996, 17 (“stagy”); Walsh 1996, 14 (“low stage”); “farcical” is M. Heseltine’s version in the Loeb Classical Library (27); “théâtral” is A. Ernout’s in the Budé (15).

71. Panayotakis 1994 stresses the resonances of the figure of Quartilla with mime acting (“like an archimima in her own production of a mimic play,” 326), though sometimes pushes the exact parallels too far (even rewriting the episode as a mimic script on 329–30); largely reprised in Panayotakis 1995, 38–51. Other studies also point to the general influence of mimes, here and elsewhere, in the novel. See, e.g., Schmeling 2011, 55 (with earlier bibliography).

72. I am here developing some of the implications of Plaza’s discussion of the episode (2000, esp. 77–79), including her interest in the “inversion of social and literary norms.”

73. It is a text with a complicated history: Festus was drawing on the work of the Augustan scholar Verrius Flaccus, but part of Festus’ dictionary is now known only through a summary by an eighth-century scholar, Paul the Deacon. And that is only part of the text’s vicissitudes—which are a major theme of the essays in Glinister and Woods 2007.

74. Festus, s.v. “Pictor Zeuxis,” p. 228L. My translation glosses over some of the predictable textual confusions.

75. Golahny 2003, 199–205, clearly justifying the identification of the scene.

76. For a brief collection of misogynistic themes on old women in Roman culture, see Parkin 2003, 86–87.

77. Pliny HN 35.65–66 (the second part of the passage tells the story of Zeuxis’ dissatisfaction with his own lifelike rendering of a child). Discussions include Elsner 1995, 16–17; Morales 1996, 184–88; S. Carey 2003, 109–11.

78. Warner 1994, 149–50.

79. The scattered ancient evidence to Baubo (and her relation to the similar figure of Iambe) is collected and discussed from a classical perspective in, for example, H. King 1986; Olender 1990; O’Higgins 2001, 132–42. For modern feminist explorations, see Cixous and Clément 1986, 32–34; Warner 1994, 150–52. See also ch. 6, n. 70.

80. Athenaeus, Deipnsophistae 14.614a–b.

81. Jacoby, FGrHist, no. 396 (the story in question is F10). No surviving quotations from Semus are found in authors earlier than the late second century CE; how long before that he wrote is frankly impossible to be certain.

82. In addition to this story, see Pausanias 9.39.13 and, more explicitly, Suda, s.v. εἰς Τροφωνίου μεμάντευται.

83. My translation tries to capture the verbal echoes of the oracle’s response: promising soothing laughter for the “unsoothed” Parmeniscus.

84. “Mothers” in literary oracular responses were never what they seemed: in another famous example, “kissing your mother” turned out to mean kissing the earth (Livy 1.56).

85. It is often assumed (by, e.g., Rutherford 2000, 138–39) that this Parmeniscus was identical with the Pythagorean philosopher “Parmiscus” of Metapontum listed in a third-century CE treatise by Iamblichus (De vita Pythag. 267, p. 185 (Nauck), emended to “Parmeniscus”) and perhaps also with the Parmiscus whose dedication at the sanctuary of Leto is recorded on an inscribed temple inventory of 156/5 BCE (IDelos 1417A, col. 1, 109–11). Maybe, or maybe not. The passing reference to a Pythagorean Parmeniscus in Diogenes Laertius (Vitae 9.20) does not clinch it either; as LGPN makes very clear, Parmeniscus and its cognates are commonly attested Greek personal names.

86. Kindt 2012, 36–54, based on Kindt 2010.

87. Kindt 2012, 49: “Parmeniscus’ laughter, we may assume, changes in quality as it becomes self-reflective. It starts off as a naïve and unreflected response to the apparent crudeness of divine form and turns into an astonished appreciation of the complexities of divine representation as Parmeniscus grasps the meaning of the oracle.” Kindt 2010, 259, is more tentative (“we may suspect” rather than “assume”).

88. παραδόξως ἐγέλασεν gives absolutely no hint of any change.

89. Halliwell 2008, 38–40, provides a useful collection of Greek agelasts (though some of this laughter avoidance is not attested before the Roman period; see, e.g., Plutarch. Per. 5).

90. Cicero, Fin. 5.92; Jerome, Ep. 7.5; Pliny, HN 7.79. Other references include Fronto, Ad M. Antoninum de eloquentia (van den Hout) 2.20; Ammianus Marcellinus 26.9.11.

91. In Jerome’s letter (Ep. 7), the focus is not so much on Crassus himself but on the proverb: “. . . secundum illud quoque, de quo semel in vita Crassum ait risisse Lucilius: ‘similem habent labra lactucam asino carduos comedente.’” The idea of the donkey eating thistles as a visual spectacle, lying behind the popular saying, is clearly suggested in one of Babrius’ collection of fables (133): a fox spots a donkey eating thistles and asks him how he can eat such spiky food with his soft tongue.

92. N. J. Hall 1983, 1035–39 (a less lurid version of the Trollope story than is often told). There is always the temptation to track down some medical cause, as in the case of the Kings Lynn bricklayer: see www.bbc.co.uk/news/ukengland-18542377.

93. Valerius Maximus, 9.12, ext. 6.

94. Diogenes Laertius, Vitae 7.185.

95. For the obscene associations of figs, see Jeffrey Henderson 1991, 23, 118, 135. Is it relevant that it was figs that Aesop made his thieving fellow slaves vomit up (see above, p. 138)?

96. Diogenes Laertius, Vitae 7.184.

97. Tertullian, De anim. 52.3.

98. The curious text known as the Testamentum Porcelli (The piglet’s last will and testament) provides another example here. Jerome stresses that it was well known to get people cracking up, cachinnare, rather than ridere (Contra Rufinum 1.17).

99. The subtitle of Schlam 1992 has inspired this section’s title.

100. My terminology on donkeys is not quite so loose as that on monkeys, but I recommend M. Griffith 2006 to anyone wanting precise information on the varieties of ancient (especially Greek) equids and their cultural resonances.

101. The essays collected in Harrison 1999 offer a good conspectus of recent Anglophone approaches to the Metamorphoses, from what is now a vast bibliography. Fick-Michel 1991, 395–430, assembles references to laughter in the novel; Schlam 1992, 40–44, is a briefer critical résumé.

102. It is generally agreed that this cannot be the second-century CE satirist Lucian; Mason 1999a, 104–5, sums up the arguments. In what follows, I will usually call the work Lucianic.

103. Photios, Bib. Cod. 129. The problems in getting to the bottom of what Photios is saying are laid out as clearly and sharply as anywhere in Winkler 1985, 252–56; see also Mason 1999a, 103–4.

104. The usual modern assumption is that the lost work of Lucius of Patrai is the earliest, but there has been endless learned conjecture (and plenty of false certainty) about the precise relationships of the various versions (summed up well by Mason 1999b), in particular which sections of Apuleius’ novel were his own invention and which derived from Lucius of Patrai. The wildly different conclusions on the extent of Apuleian originality reached (on the basis of minute philological dissection of the text) by Bianco 1971 and van Thiel 1971 are instructive (as well as dispiriting); Walsh 1974 clearly summarizes their differences.

105. Apuleius, Met. 10.13–17; ps.-Lucian, Onos 46–48.

106. Met. 10.13. I wonder if we should detect here a nod toward the saying about the donkey and the thistles.

107. “Ne humanum quidem”: Met. 10.14. As Zimmerman 2000, 214, observes, “The ironical play with humanum becomes more complex when one considers that it is his very sensus humanus . . . that makes the ass steal human food.”

108. Met. 10.16.

109. J. R. Heath 1982 discusses the role of human nutrition in Apuleius (though not focusing on this passage in particular); for the presentation of the ass as a (human) friend, see Met. 10.16, 10.17.

110. R. May 1998; 2006, 300–302.

111. Met. 10.16; Onos 47 (τοσοῦτον γελῶσιν, πολὺν γέλωτα, etc.).

112. Apuleius could not possibly have read the work of the third-century Diogenes Laertius, though Valerius Maximus was writing at least a century earlier. But my claim does not depend on whether Apuleius was familiar with these precise texts (and indeed there are no verbal echoes between the Latin versions of Valerius and Apuleius, and Apuleius in any case offers a different account of Philemon’s death, in Florida 16). The implication of what I have shown so far is that the “dining donkey” story was a well-known popular joke in the Roman world—and that common knowledge underpins my discussion of Apuleius’ use of it here.

113. I do not see other significant differences between the two accounts that are relevant to my arguments on the culture of laughter. Zimmerman 2000, 229–30, contrasts the donkey’s reaction in each text to being laughed at when first caught eating: pleasure in Apuleius (10.16), shame and embarrassment in ps.-Lucian (47). But pleasure very soon returns in the Lucianic account, as Zimmerman allows.

114. Onos 47.

115. Met. 10.16.

116. Bakhtin (1981 [1937–38]) underlined the polyphonic aspects of the novel in an essay first published half a century earlier.

117. Onos 10 (before transformation), 15 (braying), 55 (for the implication of laughter after his return to human shape).

118. Met. 2.31–3.13.

119. Met 2.31.

120. Met 3.2 (“nemo prorsum qui non risu dirumperetur aderat”).

121. The story of the “murder” and the revelation of what “really” happened is, of course, more complicated than I am making it seem; for its literary precedents and the confrontation between reality and illusion staged here, see Milanezi 1992; Bajoni 1998; R. May 2006, 195–98.

122. Met. 3.13.

123. He is in fact called “victim” (victimam) at Met. 3.2.

124. D. S. Robertson 1919 casts around for real-life ancient ritual parallels (involving the leading of a scapegoat around town); partly followed by James 1987, 87–90. Habinek 1990, 53–55, stresses the (structural) role of Lucius as scapegoat. Kirichenko 2010, 36–39, 45–58, identifies mimic elements (comparing the risus mimicus of Petronius). R. May 2006, 182–207, the best introduction to the episode and previous scholarship on it, points to its theatricality and metaliterary aspects.

125. R. May 2006, 190–92; Zimmerman 2000, 25–26, 225–26 (for verbal echoes in the description of laughter between the two episodes).

126. Cachinnus: Met. 3.7 (with Van der Paardt 1971, 67; Krabbe 1989, 162–63).

127. Met. 3.11: “Iste deus auctorem et actorem suum propitius ubique comitabitur amanter, nec umquam patietur ut ex animo doleas, sed frontem tuam serena venustate laetabit assidue.” Or so it reads if we accept an early twentieth-century emendation of the manuscript tradition. Auctorem et actorem is Vollgraff’s conjecture (1904, 253) for the unsatisfactory or incomprehensible manuscript reading: whether auctorem with the meaningless et torem written into the interlinear space above, or the alternative and feeble auctorem et tutorem. It is generally now accepted that auctorem et actorem is correct, but given the phrase’s celebrity status, it is worth remembering that this is (only) a conjecture. Tatum 2006 discusses Vollgraff’s conjecture, plus the background of the phrase in earlier Latin, at length, leading to (in my view) difficult conclusions on Apuleius’ links with Cicero, though La Bua 2013 takes a similarly Ciceronian direction in the discussion of Lucius’ mock trial.

128. Winkler 1985, 13.

129. Kirichenko 2010, 58, also stressing the contrast between the actor as “passive” (“Lucius improvises in accordance with a pre-ordained storyline”) and the auctor as auctorial/authorial (he “creatively co-authors the entire performance”); see above, pp. 119–20, 167, on the role of actors as “only mouthpieces of the scripts of others.”

130. Schlam 1992 picks up the ambiguity here, with a slightly different emphasis from mine: “In an ironic sense the promise offered by the magistrates turns out to be true. Laughter does accompany the Ass, but he is the wretched object at which others laugh, often maliciously” (43).


1. Σχολαστικὸς καὶ φαλακρὸς καὶ κουρεὺς συνοδεύοντες καὶ ἔν τινι ἐρημίᾳ μείναντες συνέθεντο πρὸς τέσσαρας ὥρας ἀγρυπνῆσαι καὶ τὰ σκεύη ἕκαστος τηρῆσαι. ὡς δὲ ἔλαχε τῷ κουρεῖ πρώτῳ φυλάξαι, μετεωρισθῆναι θέλων τὸν σχολαστικὸν καθεύδοντα ἔξυρεν καὶ τῶν ὡρῶν πληρωθεισῶν διύπνισεν. ὁ δὲ σχολαστικὸς ψήχων ὡς ἀπὸ ὕπνου τὴν κεφαλὴν καὶ εὑρὼν ἑαυτὸν ψιλόν· Μέγα κάθαρμα, φησίν, ὁ κουρεύς· πλανηθεὶς γὰρ ἀντ’ ἐμοῦ τὸν φαλακρὸν ἐξύπνισεν. Different manuscripts of the text (see pp. 186–87) include a shorter and slightly differently worded version of this joke, with the same point.

2. I cite the jokes from the edition of A. Thierfelder (1968), which is in general to be preferred to the more recent Teubner edition of R. D. Dawe (2000), on which see the important and wide-ranging review Jennings 2001. The Philogelos has been the subject of several recent studies (on both its textual tradition and—rather less often—its cultural significance). Note especially Thierfelder 1968; Baldwin 1983 (though the translations are sometimes misleading); Andreassi 2004 (the best modern introduction)—all these underlie what follows and are cited only to draw attention to particularly significant discussion or to indicate disagreement. Brief cultural explorations include Winkler 1985, 160–65; Bremmer 1997, 16–18; Hansen 1998, 272–75; Schulten 2002. In addition, there are several more or less popular modern translations, along the lines of “the world’s oldest jokebook”: for example, Cataudella 1971, 89–154 (with a useful scholarly introduction); Löwe 1981; Zucker 2008; Crompton 2010.

3. These three examples are based on 104, 231, and 173 (I confess that my paraphrases here have adjusted the ancient jokes to familiar modern comic idioms).

4. This is not, in other words, a case of creative translation from the Greek into modern comic clichés. Note, however, this is the only joke in the collection to start in this way; the trio of characters was not in general the cliché of ancient joking that it is of modern.

5. Johnson 1741, 479. His translation runs: “The Sage fell to scratching his Head, and finding no Hair, abused the Barber for not calling the Philosopher in his Turn, for do not you know, says he, that I, who am the bald Man, was to have been called up last.” It is a useful example of the varied responses that jokes get as they travel through time.

6. Wilson 1996, 212.

7. Thierfelder 1968, 129–46, is the clearest account of the whole manuscript tradition; note also Perry 1943. Rochefort 1950 discusses the full contents of the main manuscript (A = Par. Sup. Gr. 690). The first joke (now 265) in the earliest manuscript (G = Cryptoferratensis A 33) has a point similar to that of two others in the full collection but is significantly different in language and detail. “Ascholastikos was asked how many pints the jar held and answered: ‘Do you mean of wine or water?’” Compare number 92, which has a scholastikos ask his father how much a three-pint (πεντακότυλος) vessel holds, and 136, which has a teacher from Sidon ask a pupil (though the text is uncertain) how much a three-pint vessel holds—“Do you mean wine or oil?” he replies.

8. Tzetzes, Chil. 8.969–73 (Leone).

9. It may be significant that Tzetzes elsewhere tells a very similar joke, which he ascribes to a “story” or “fable”; see Epistulae 50 (Leone).

10. These possibilities and more are explored by Baldwin 1986; Andreassi 2004, 63–65. We should bear in mind that book titles and their authors can, and do, blur; Mrs. Beeton refers to both book and author, as in many cases does Livy (and there was likewise confusion in the medieval world over whether Suda was the title of an encyclopedia or the name of its compiler).

11. On the Alexandrian Hierokles and other homonyms, see Andreassi 2004, 28–29. The dual authorship between Hierokles and Philagrios given by the longer manuscript selections, in contrast to the shorter selections ascribed to Hierokles alone, has predictably launched theories about originally separate works of Hierokles and Philagrios that were at some point combined—a combination that might (or might not) explain some of the complexity of the manuscript tradition (intricately discussed by Thierfelder 1968, 129–202, with diagram on 202).

12. Suda Φ 364 (Adler); the text as printed there runs οὗτός [Philistion] ἐστιν ὁ γράψας τὸν Φιλόγελων, ἤγουν τὸ βιβλίον τὸ φερόμενον εἰς τὸν Κουρέα (but a minor textual emendation, or even just the substitution of a lowercase for an uppercase Κ, would produce very different senses). For further possible links with Philistion, see Cataudella 1971, xxv; Reich 1903, 454–75 (which trusts theSuda’s attribution).

13. New Pauly, s.v. “Philogelos”; Bremmer 1997, 16, with 25n32. On the culture of barbershops, see S. Lewis 1995 (a survey of Greek material); Polybius 3.20; Plutarch, Mor. 508f–509c (= De garr. 13).

14. Abdera: 110–27; Kyme: 154–82; Sidon: 128–39; Rome: 62; Rhine: 83; Sicily: 192.

15. Drakontides: 170; Demeas: 102; Scribonia: 73; Lollianus: 162.

16. Denarii: 86, 124, 198, 213, 224, 225; anniversary: 62. Other Latinizing forms in the Greek (in, e.g., 135, 138) may also point to the cultural background, as well as reflect early Byzantine Greek usage.

17. 62. Other hints of a possibly third-century CE context have been squeezed from the text: the use of myriads as a unit of currency in 80 and 97, and in 76 the possible reference to the temple of Serapis in Alexandria (the destruction of that Serapeum in 391 would give a terminus ante quem for the origin of the joke—but Alexandria is not actually mentioned!); see Thierfelder 1968, 224 (noting that the joke implies “going up” [ἀνελθόντι] to the temple—for the Alexandrian Serapeum was on a hill).

18. “It is generally thought” sidesteps many divergent views. Robert 1968, 289, is unusual in using the reference to the millennial celebrations to pinpoint (more or less) the principal date of composition; Rapp 1950–51, 318, by contrast, considers many of the jokes to be at least in the dress of the ninth or tenth century.

19. 148; Plutarch, Mor. 177a (= Regum et Imperatorum Apophth., Archelaus, 2); Mor. 509a (= De garr. 13).

20. 150; Plutarch, Mor. 534b (= De Vitioso Pudore 14). For other parallels, see 206 (with Athenaeus, Deipnosophistae 8.350b; Diogenes Laertius, Vitae 1.104); 264 (with Plutarch, Mor. 178f [= Regum et Imperatorum Apophth., Philip, 24]); Valerius Maximus, 6.2 ext. 1; Stobaeus, Anthologium 3.13.49 (attributing the story to “Serenus”).

21. 73. On the possible identification, see Thierfelder 1968, 224. The funny idea of people objecting to the unhealthy siting of tombs (which could not harm those already dead) is also the theme of 26.

22. 78: Σχολαστικὸς εἰκόνας ἀρχαῖα ζωγραφήματα ἐχούσας ἀπὸ Κορίνθου λαβὼν καὶ εἰς ναῦς ἐμβαλὼν τοῖς ναυκλήροις εἶπεν· Ἐὰν ταύτας ἀπολέσητε, καινὰς ὑμᾶς ἀπαιτήσω. Andreassi 2004, 71–80, is a good discussion of the processes of anonymization of these jokes: “Dallo ‘storico’ al ‘tipico’ (e viceversa . . . )” (71).

23. Velleius Paterculus 1.13.4 (ending with the punch line “. . . iuberet praedici conducentibus, si eas perdidissent, novas eos reddituros”). We should add to this pair 193, which reprises a joke told by Cicero (De or. 2.276) about the poet Ennius and Scipio Nasica (discussed on p. 200).

24. This is why some of the final few jokes return to the theme of the scholastikos, otherwise found in the first half of the book, and the first joke in the earliest manuscript, unattested elsewhere, is relegated to the final entry in the modern text, number 265.

25. The best discussion of the scholastikos, stressing the connections with comic performance, is Winkler 1985, 160–65, with Andreassi 2004, 43–51 (including a review of modern translations), and Kirichenko 2010, 11–16. The character is a leitmotiv of Conte 1997 (though not specifically as he appears in the Philogelos). I have borrowedegghead from Baldwin 1983.

26. 3: Σχολαστικῷ τις ἰατρῷ προσελθὼν εἶπεν· Ἰατρέ, ὅταν ἀναστῶ ἐκ τοῦ ὕπνου, ἡμιώριον ἐσκότωμαι καὶ εἶθ’ οὕτως ἀποκαθίσταμαι. καὶ ὁ ἰατρός· Μετὰ τὸ ἡμιώριον ἐγείρου.

27. 40; some manuscript versions do not include the detail of the father’s status, so making only the humorous contrast between the small boy and the large crowd.

28. 6; with 253, a briefer version. 174 (“A man from Kyme”) is on a similar theme, and 27 inverts the point.

29. 164: Κυμαῖος ἐν τῷ κολυμβᾶν βροχῆς γενομένης διὰ τὸ μὴ βραχῆναι εἰς τὸ βάθος κατέδυ; 115: Ἀβδηρίτης εὐνοῦχον ἰδὼν γυναικὶ ὁμιλοῦντα ἠρώτα ἄλλον, εἰ ἄρα γυνὴ αὐτοῦ ἐστι. τοῦ δὲ εἰπόντος εὐνοῦχον γυναῖκα ἔχειν μὴ δύνασθαι ἔφη· Οὐκοῦν θυγάτηρ αὐτοῦ ἐστιν. The first of these jokes points nicely to the different status of water when in a pool or when falling from the sky: we don’t, after all, think of swimming as “getting wet.”

30. The origins of these modern traditions of joking in national(ist) geopolitics set them clearly apart from the ancient traditions, despite superficial similarities often noted (by, for example, Toner 2009, 98). The cities of the Philogelos are more internal than foreign objects of jocularity. And the jokes are probably closer to the English “disgusted-of-Tunbridge Wells” style of quip, where Tunbridge Wells stands for a town whose inhabitants are caricatured as elderly, conservative, and out of touch with modernity (and always writing to newspapers to express their “disgust”).

31. Strabo, Geographica 13.3.6; briefly discussed by Purcell 2005, 207–8 (which appears to find the passage, in detail, as puzzling as I do). This similar anecdote definitely referring to the city in Asia Minor makes it virtually certain that the jokes on Kyme in the Philogelos are not referring to either of the other ancient towns that could be spelled in the same way (in Euboea or our Cumae, in South Italy).

32. Martial, Epigram. 10.25 (see also Juvenal 10.50); Cicero, Att. 4.17.3 (SB 91), with 7.7.4 (SB 130). See also Machon, frag. 11, 119–33 (Gow); Lucian, Hist. conscr. 1.

33. 35; 158.

34. See above, n. 7. Occasionally too the type characters might be combined, as in 131, which concerns a Sidonian scholastikos.

35. 137 (essentially the same joke as 99).

36. 162: Κυμαίων <τὴν> πόλιν τειχιζόντων εἷς τῶν πολιτῶν Λολλιανὸς καλούμενος δύο κορτίνας ἰδίοις ἐτείχισεν ἀναλώμασι. πολεμίων δὲ ἐπιστάντων ὀργισθέντες οἱ Κυμαῖοι συνεφώνησαν, ἵνα τὸ Λολλιανοῦ τεῖχος μηδεὶς φυλάξῃ ἀλλ’ ἐκεῖνος μόνος.

37. Parts of the Philogelos show signs of an internal logic or significant ordering within the division into type characters: 25, 26, and 27, for example, are a trio concerning death; 52 is a neat inversion of the preceding joke. It is, of course, impossible to be sure whether such patterns are to be put down to the compilers or to whatever source text they might have been using.

38. There is a trace of another standard joke line in the scholastikos group: on three occasions (15, 43, 52), just before the punch line (and as if to signal it), the egghead says words to the effect of “What an idiot I am,” “No wonder they call us idiots” (μωροὶ καλούμεθα, μωροὶ νομιζόμεθα, μωρός εἰμι).

39. West (1992, 268) comes close to suggesting an academic function for the book when she writes, “But it seems worth raising the question whether it was really intended as a joke book, or whether it embodies an attempt at a motif-index, compiled, perhaps, to assist an analysis of various forms of wit and humour.”

40. Andreassi 2004, 37–43, reviews the various connections with other genres. Jokes with a probable link to fable include 142 and 180 (see also Andreassi 2006, on the “greedy man” [λιμόξηρος] in the Philogelos and the Life of Aesop). Kirichenko 2010, 11–16, discusses mimic themes in the scholastikos jokes. Floridi 2012 discusses links between the Philogelos and scoptic epigram; for specific points of comparison, see, e.g., 97 and AP 11.170; 235 and AP 11.241. The few sexual jokes in the collection include 45 (see above, p. 198), 244, 245, 251. Whether this reflects the character of the Philogelos from its earliest phases or is the result of medieval bowdlerization, we do not know.

41. E.g., 4, 135, 184, 189.

42. 14; the vocabulary and metaphors suggest the influence of comic performance and/or mime (see Aristophanes, Thesm. 797; Herodas, Mimiambi 2.15).

43. 239: “Οἴμοι, τί δράσω; δυσὶ κακοῖς μερίζομαι” (“Alas, what shall I do? I am torn between two evils”).

44. 121: οὐκέτι τρέχει, ἀλλὰ πέτεται. There are links with AP 11.208; see Floridi 2012, 652–53. But the epigram is simpler, resting only on a play between running (to dinner) and flying.

45. 8: Σχολαστικὸς θέλων πιάσαι μῦν συνεχῶς τὰ βιβλία αὐτοῦ τρώγοντα κρέας δακὼν ἐν τῇ σκοτίᾳ ἐκάθισεν . . . “Sententiam non completam esse monuit Dawe” is the comment in his Teubner edition. Others have not been so despondent. Perhaps the joke is that the scholastikos was pretending to be a cat (so Thierfelder 1968, 205).

46. Σχολαστικὸς ἀργυροκόπῳ ἐπέταξε λύχνον ποιῆσαι. τοῦ δὲ ἐξετάσαντος, πηλίκον ποιήσει, ἀπεκρίνατο· Ὡς πρὸς ὀκτὼ ἀνθρώπους.

47. Felice 2013.

48. To be honest, I find this interpretation slightly puzzling. For—to think it through in finer detail than the joke probably deserves—the scholastikos can hardly have mistaken the silversmith for a fish seller in the first place. Are we to imagine that he is the clever exploiter of the pun by replying to the silversmith’s question with an answer that exposes the double meaning? Nor am I convinced that the “very occasional” usage is enough to give ποιέω a clear resonance of food preparation; so far as I can see, we are dealing with one passage from the Septuagint (Genesis 18:7).

49. Different kinds of ingenuity are on display in Thiel 1972 (emendation of the text of 237, accepted and elaborated in Dawe’s Teubner text); Morgan 1981 (attempting to restore sense to 216 by translating κυβερνήτης as “governor” rather than “steersman”); Rougé 1987 (elucidating some of the nautical and navigational terminology); Lucaszewicz 1989 (emending the text of 76 to produce a joke about the scholastikos’ slave relations).

50. 200: Ἀφυὴς μαθητὴς ὑπὸ τοῦ ἐπιστάτου κελευσθεὶς ὀνυχίσαι οἰκοδεσπότην ἐδάκρυσε. τοῦ δὲ τὴν αἰτίαν ἐρωτήσαντος ἔφη· Φοβοῦμαι καὶ κλαίω· μέλλω γὰρ τραυματίσαι σε, καὶ παρωνυχίδας ποιήσεις, καὶ τύψει με ὁ ἐπιστάτης. Thierfelder (1968, 261–62) does his best, pointing to links with the previous joke and to the logical confusion of the simpleton’s complaint; Baldwin’s mistranslation (1983, 38) does not help. The joke does, however, serve to remind us of the difficulties (and pain) of nail trimming in antiquity.

51. 214: Φθονερὸς εἰς γναφεῖον εἰσελθὼν καὶ μὴ θέλων οὐρῆσαι ἀπέθανεν.

52. For an up-to-date (and rather less lurid than usual) view of the use of urine in the fulling industry, see Flohr and Wilson 2011, 150–54; Flohr 2013, 103–4, 170–71 (though without reference to this joke). I have been helped with this joke by some careful comments by Istvan Bodnar on an earlier podcast version of these ideas. Even so, some problems remain—including my translation meanie,implying niggardly (so not wanting to give his urine away for free). That is not the most obvious sense of the Greek φθονερὸς, which more usually (as in the other jokes in this category) suggests spitefulness.

53. These include jokes on parricides: 13, 152; the death of a slave: 18; misunderstanding about or disputed death: 22, 29, 70; inheritance: 24, 104, 139, 229; tombs: 26, 73; funerals: 38, 40, 123, 154, 247; coffins: 50, 97; infanticide, 57; the death of a son: 69, 77; suicide: 112, 231; crucifixion: 121; condemnation to death: 168; sudden death: 214; the death of a wife: 227.

54. Critchley 2002, 65–66, partly drawing on Mary Douglas’s famous essay on jokes (1968) and crisply encapsulating approaches that underlie other, more specific contributions to joke studies (see, for example, Kerman 1980, discussing “light-bulb” gags in broadly similar terms). It is striking that some jokes in the Philogelos explicitly make issues of relativism (or the failure to understand the nature of a different perspective) the topic of joking: see, for example, 49, in which a scholastikos, looking at the moon, asks his father if other cities have moons like theirs.

55. 5: Σχολαστικῷ τις ἀπαντήσας ἔφη· Κύριε σχολαστικέ, καθ’ ὕπνους σε εἶδον. ὁ δέ· Μὰ τοὺς θεούς, εἶπεν, ἀσχολῶν οὐ προσέσχον; the alternative version is 102.

56. 15: Σχολαστικὸς καθ’ ὕπνους ἧλον πεπατηκέναι δόξας τὸν πόδα περιέδησεν. ἑταῖρος δὲ αὐτοῦ πυθόμενος τὴν αἰτίαν καὶ γνούς· Δικαίως, ἔφη, μωροὶ καλούμεθα. διὰ τί γὰρ ἀνυπόδητος κοιμᾶσαι; 207 (see also 124, 243). The theme of dreaming versus reality is also found in scoptic epigrams; see Floridi 2012, 643.

57. W. V. Harris 2009 is an important survey; Harris-McCoy 2012, 1–41, is a useful introduction to the dream interpretations of Artemidorus.

58. 45: Σχολαστικὸς νυκτὸς ἐπανέστη τῇ μάμμῃ αὐτοῦ. πληγὰς δὲ διὰ τοῦτο ὑπὸ τοῦ πατρὸς λαβών· Σύ, εἶπεν, τοσοῦτος χρόνος ἐστὶν ἐξ οὗ τὴν μητέρα μου ὀχεύεις, μηδὲν ὑπ’ ἐμοῦ παθών, καὶ νῦν ὀργίζῃ ἐπὶ τῇ μητρί σου ἅπαξ με εὑρών. Baldwin 1983, 65, detects the influence of mime.

59. 57: ∏ρῶτον, ἔφη, σὺ τὰ τέκνα σου κατόρυξον, καὶ οὕτως ἐμοὶ συμβούλευε τὸν ἐμὸν ἀνελεῖν.

60. 80: Σχολαστικοῦ πλέοντος ἐκινδύνευεν ὑπὸ χειμῶνος τὸ πλοῖον. τῶν δὲ συμπλεόντων ἀπορριπτούντων ἐκ τῶν σκευῶν, ἵνα κουφισθῇ τὸ πλοῖον, κἀκείνῳ τὸ αὐτὸ ποιεῖν παραινούντων, ὁ δὲ ἔχων χειρόγραφον ἑκατὸν πεντήκοντα μυριάδων, τὰς πεντήκοντα ἀπαλείψας· Ἲδε, φησίν, ὅσοις χρήμασιν ἐπεκούφισα τὴν ναῦν. Rougé 1987, 10–11, sees the point of this most clearly.

61. 12: Σχολαστικῷ ἀποδημοῦντι φίλος αὐτοῦ ἔλεγεν· Ἀξιῶ σε δύο παῖδας ἀγοράσαι μοι, ἑκ<άτερον> πεντεκαίδεκα ἐτῶν. ὁ δὲ εἶπεν· Ἐὰν τοιούτους μὴ εὕρω, ἀγοράσω σοι ἕνα τριάκοντα ἐτῶν. I include the sexual reading in deference to my graduate class at Berkeley, who had no doubt at all that that was the sense.

62. IOU: 161; country estate: 131 (a doublet of 60; Baldwin 1983’s translation is misleading); ladder rungs: 93; tertian fever: 175a. Others jokes on related themes include 3, 62, 71, 84, 196, and the gags about wine and water in n. 7.

63. 22: Σχολαστικὸς ἀπαντήσας τινὶ φίλῳ αὐτοῦ εἶπεν· Ἤκουσα, ὅτι ἀπέθανες. ὁ δὲ ἀπεκρίνατο· Ἀλλ’ ὁρᾷς με ζῶντα. καὶ ὁ σχολαστικός· Καὶ μὴν ὁ εἰπών μοι κατὰ πολὺ σοῦ ἀξιοπιστότερος ἦν.

64. 193: Δύσκολόν τις ἐζήτει. ὁ δὲ ἀπεκρίνατο· Οὐκ εἰμὶ ὧδε. τοῦ δὲ γελάσαντος καὶ εἰπόντος· Ψεύδῃ· τῆς γὰρ φωνῆς σου ἀκούω—εἶπεν· Ὦ κάθαρμα, εἰ μὲν ὁ δοῦλός μου εἶπεν, εἶχες ἂν αὐτῷ πιστεῦσαι· ἐγὼ δέ σοι οὐ φαίνομαι ἀξιοπιστότερος ἐκείνου εἶναι;

65. De or. 2.276.

66. Twin: 29; baby’s name: 95; corpse: 171.

67. Proofs of identity and status, including birth certificates, were not nonexistent in the Roman world; they were presumably commoner where issues of status and privilege were at stake and in some parts of the empire more than others (though how far the pattern of their survival reflects the original distribution is unclear). Schulz 1942; 1943 remain useful surveys of the evidence. Wallace-Hadrill 2011, 144–45, briefly discusses a case in Herculaneum where an individual’s birth details remained obscure (as I strongly suspect must have been the norm).

68. Lucian, Demon. 12–62. Schlapbach 2010, esp. 258–60, offers a sophisticated reading of the relationship between these witty sayings and Lucian’s construction of the written life of Demonax. Modern usage of the ancient terms apophthegmata and chreiai tends to imply too clear a division between the two categories: the former being “clever sayings,” the latter more specifically “moral maxims” or witty parodies of such. In practice, the categories merge, as they also do with proverbs and riddles. On the interchangeability, see McClure 2003, esp. 274.

69. Suetonius, Aug. 56. It was apparently a juvenile collection by the dictator and presumably consisted of his own dicta (though that is not explicitly stated), or why would Augustus have wanted to keep it under wraps?

70. Such compilations as Plutarch’s Sayings of the Spartans are also classified according to speaker, even if the book as a whole adds up to a portrait of cultural or ethnic character.

71. Suetonius, Gram. 21. The title probably militates against a compilation of sayings, but it does clinch the issue; the collection of Aristodemus (see p. 204) seems to have been much more biographical than its title would suggest.

72. Stich. 454–55, 221–24.

73. Persa 392–94: “Librorum eccillum habeo plenum soracum / . . . / dabuntur dotis tibi inde sescenti logi.”

74. Maltby 1999, referring to Persa 395 (“atque Attici omnes; nullum Siculum acceperis”). We might compare Gow’s confidence that Machon’s Chreiai could well have been “a valuable vademecum” for an ancient jester, similar to “a modern book of jokes from which a public speaker or raconteur . . . can refresh his memory or replenish his repertoire” (1965, 24). As Kurke 2002 makes clear, whatever this puzzling text was, it certainly was not that.

75. It is hard to get much sense of the work (and its date is, in any case, a matter of guesswork: second century BCE or later, but how much later?). The quotations in Athenaeus are all jokes attached to named individuals—kings, gluttons, parasites, and prostitutes; see, e.g., 6.244f (giving the terminus post quem), 6.246d–e, 8.345b–c, 13.585a. They may have some resemblance to the types of thePhilogelos, but how close a resemblance I am not sure.

76. The text is found most conveniently in Siegmann 1956, 27–37, which discusses in detail the different readings and interpretations up to that date. Almost everything about this text is contested. It is unclear, for example, whether Pyrrhos—if thought of as a proper name—is a “real” name or a nickname (such as Ginger). The only other more or less comprehensible heading, though much restored, appears to be εἰ[ς] φα[λ]ακρόν (few letters are entirely clear, and again there has been debate on whether it refers to a bald man or is some form of proper name).

77. Kassel 1956. This view is accepted by, e.g., Maltby 1999; Andreassi 2004, 22–23 (“ha convincentemente sostenuto che il papiro costituisse una sorta di Witzbuch”).

78. See Siegmann 1956; more briefly Andreassi 2004, 23.

79. Aristophanes, Vesp. 1427–31.

80. Links with fable: Aristophanes, Vesp. 1259. The most recent survey of Sybarite stories is Bowie 2013, 252–55. I am not as confident as Bowie (252) that the genre of these stories must in some form go back to before the destruction of Sybaris (as the place had such proverbial renown), but I am struck that he concludes that the collection of these stories only just predated Ovid (255); Aelian, VH14.20 (writing in the late second or early third century CE) implies that he had read a collection. See also, on the tradition of Sybaris, Gorman and Gorman 2007 (usefully showing how much Athenaeus “contributes” to the fragments he cites).

81. The context and characters of the Deipnosophistae are well discussed in various essays in Braund and Wilkins 2000, especially Milanezi 2000, on the section on joking, and Braund 2000, on the Roman background (including the identity of Ulpian: see esp. 17).

82. Athenaeus, Deipnosophistae 14.614d–e: τοσαύτη δ’ αὐτῶν δόξα τῆς ῥᾳθυμίας ἐγένετο ὡς καὶ Φίλιππον ἀκούσαντα τὸν Μακεδόνα πέμψαι αὐτοῖς τάλαντον, ἵν’ ἐκγραφόμενοι τὰ γελοῖα πέμπωσιν αὐτῷ. A similar but shorter version of the story is found at 6.260a–b, citing the second-century CE Hegesander of Delphi as the immediate source.

83. Quotation from Hansen 1998, 273; see also Andreassi 2004, 18–19.

84. As Hansen 1998, 273, carefully concedes.

85. Deipnosophistae 14.614c. Athenaeus’ gloss: Ἀναξανδρίδης δ’ ἐν Γεροντομανίᾳ καὶ εὑρετὰς τῶν γελοίων φησὶ γενέσθαι Ῥαδάμανθυν καὶ ∏αλαμήδην, λέγων οὕτως; followed by the quotation itself: καίτοι πολλοί γε πονοῦμεν. / τὸ δ’ ἀσύμβολον εὗρε γελοῖα λέγειν Ῥαδάμανθυς / καὶ ∏αλαμήδης.

86. For example, Milanezi 2000, 402, though the chapter is, in general, a useful study of this section of Athenaeus’ work. On Palamedes as a mythical inventor and culture hero, see Gera 2003, 122–27; for another appearance of Rhadamanthys, see p. 161; on this pairing, see Ceccarelli 2013, 69 (which is slightly more careful than most on what exactly it attributes to Anaxandrides).

87. Of course, there may have been earlier, now lost, claims about the role of Palamedes and Rhadamanthys as inventors of the joke, but the fact is that this is the only testimony we have—and whatever parallels might once have existed (or not), the slippage in Athenaeus and the effective reinterpretation of Anaxandrides’ claim are striking.


1. Marzolph 1987 explores the similarities with Arabic traditions. Andreassi 2004, 81–124, collects further parallels in different joking cultures.

2. The work of Barbara Bowen has opened up the world of Renaissance jokebooks. See, for example, Bowen 1984; 1986a; 1986b; 1998. For an earlier period (Cicero’s jokes in the culture of the twelfth-century English court), see J. M. Martin 1990.

3. This story goes back to the nineteenth century at least. It makes its point nicely (about both Porson and Joe Miller)—though it may not be strictly true; see Baldwin 1983, xii.

4. The joke is less apocryphal than it might seem. It is told in the diary of Powell’s fellow politician Woodrow Wyatt (Wyatt 1998, 282–83, entry for 31 January 1987): “There is a very chatty barber in the [House of] Commons who never stops telling MPs whose hair he cuts about politics and what his views are on the world. Enoch Powell went to have his hair cut by him one day, sat down and the barber said ‘How would you like your hair cut, sir?’ ‘In silence,’ Enoch replied.” Wyatt makes clear that the barber was well known for his chattiness, so Powell would have had ample time to prepare his classical joke. Even better, since first investigating this story, thanks to Gloria Tyler of the House of Commons Library, I have been able to access an interview with the barber himself, Stephen Silverne (British Library, Sound and Moving Image Collection, C1135/14); in this, he gives a very similar account of the story. For another modern version of this gag, see Andreassi 2004, 75–76.

5. Murdoch 1999 [1978], 182; my italics. Her claim that it was Freud’s favorite joke is partly intended to resonate with the complex sexual intrigues and anxieties of the novel.

6. Freud 1960 [1905], 107.

If you find an error please notify us in the comments. Thank you!