1. Nicolet (1980), 57–60, provides a description of the census which makes this principle clear.

2. Flower (2004), 6: ‘The characteristically Roman concept of an ever expanding citizen body not defined by ethnicity, religious belief, or social origins proved to be a powerful tool of control but also of assimilation.’ This nicely gets at the double-edge of the principle of diversity in civic life; my claim is that this extends to the diversity within the household as well.

3. This is not to say that slave reproduction was unimportant in Roman households, but that it is important for bolstering the particular kind of resources that slaves offer, not for contributing directly to familial reproduction. On slave reproduction, see Bradley (1994), 33–53.

4. E.g. Amerasinghe (1950); Spranger (1984); Cupaiuolo (1991), 37–50; Kruschwitz (2004), 196f.

5. The best discussion of Terence in relation to the practices of other Roman playwrights is still Wright (1974), 127–51, who recognises the possibility of a radical spirit motivating Terence’s ‘retreat’ to Greek originals.

6. As an example of some of the ways these principles could be pursued, it would be important to consider the other side of this coin as well: the ambiguities of status and identity that free people experience in relation to tbefamilia. For example, in the Heauton Timoroumenos, the discussion of the choice to expose the daughter (623–67) with attendant issues of the expenses of raising a child, esp. a daughter (cf. Phorm. 645–47); and for the Phormio, pushing further Konstan’s observation that the citizenship of the girl makes sense dramatically in the first part of the play but not in the latter part (Konstan [1983], 122f.), to get a better sense of the way that illegitimacy in this play hovers on the boundary between a private and a public matter; finally, in addition to reversing the connection between slaves and privileged knowledge (by segregating Parmeno from the action), the Hecyra also inverts the usual logic of reproduction by focusing on the fear of introducing suppositious children into the line of descent rather than on the desire to produce legitimate heirs.

7. For recent historicising analyses that take up the social contexts of comedy in the two societies, see Lape (2004) on Menander in Athens and Leigh (2004) on Plautus and Terence in Rome.

8. See n.4 above. The two quotations that follow are from Amerasinghe (1950), 62.

9. To be accurate, it is not the case that all free characters in Plautus are governed by the naturalistic aesthetic and all slaves by the farcical. Rather, Plautus’ world is generally mixed from both of these sources; but we might note that while we frequently see free characters absorbed into farce, we rarely see slave characters (barring those who will be recognised as free) represented prominently within naturalistic norms. The plays of Plautus that perhaps most clearly employ the slave/outsider as joker (i.e. operating under different rules from everyone else) would be Rudens, in which Gripus inhabits a world at odds with that of the freeborn family, or Captiui, in which the parasite Ergasilus is left alone in his farcical realm.

10. The Amphitruo demonstrates how close the extraordinary powers of the clever slave are to those of divinity, but also how the power associated with the lowly is the one that will triumph in comedy; notice that Jupiter gets his way not by appearing in all his divine glory, but by changing himself into a man. Cf. below on the Eunuch, esp. n.48.

11. For a reading of Terence based on such an (in my mind anachronistic and only loosely argued) assumption, see Cupaiuolo (1991).

12. See now Lape (2004) which offers a rich investigation into the many ways that Menan-drian comedy intersects with democratic ideology.

13. Perhaps the most emphatic example is the intrigue of the Phormio, in which the false law case which ‘forces’ the young man to marry the girl he loves turns out to have been predicated on the true facts.

14. Obviously this language could be described neutrally as just the lexicon of comedy, but it seems to appear in clusters; see e.g. And. 650f., where Pamphilus, rebuking Davos, calls him meu’ carnufex and jingles consiliis with conflauit. See also Cupaiuolo (1991), 69–71, where he analyses the uses that Terence makes of some specifically Plautine phrases.

15. See esp. the prologue to the Eunuch (35–43). Also, thinking of the tradition as an archive allows Terence to see Plautus as no closer or more distant than Menander, while thinking of the tradition as performance would accent the audience and production context he shares with Plautus.

16. Frangoulidis (1993, 1994, 1995, 1996) has made the most concerted effort to describe a Terentian practice of metatheatre, but, white he offers many good observations, I believe he over-counts specifically ‘theatrical’ language and remain unconvinced that there is a strong performative or improvisatory element. Saylor’s (1975) analysis of ‘planlessness’ in the Eunuch, for example, ends up accounting for these moments better than Frangoulidis’ (1994) ‘improvisation’.

17. See esp. 471–80, 490–94, 582–84. Zeitlin (1998), 46, offers an ethical reading of Simo’s ‘knowledge’ of comedy.

18. See esp. 470–75, 529–40.

19. See Davos’ monologue at 206ff. He starts off exactly as a Plautine clever slave would do (rousing his energies by self-address and noting the difficulty of the mission), but then (209–14) reveals that he is genuinely worried for the young master and acknowledges in a realistic, not fantastic, way the possibility that Simo will punish him by sending him to the mill. Cf. also 675–80, though there perhaps is some sarcasm in this declaration, especially at the end, where Davos says, basically, ‘If you don’t like the way I do things, you can always let me go.’

20. Another possible metaliterary joke here: the old men are about the age of people in the audience who would have seen Plautine plays at Rome in their youth (Andria is produced in 166, roughly twenty years after Plautus’ last play) and so maybe they are being tagged as ‘old-fashioned’ in not getting that slaves aren’t like that any more. This shows how even a strict translation from Menander (assuming that the father there cooked up the trick about the false wedding) takes on new meaning in a different historical context.

21. Compare the very similar scene at Phorm. 465–70.

22. Shipp (1960), ad loc, connects futtili with fundo, i.e. ‘of persons given to too much talking’; cf. OLD s.v.futtilis 3b, which advocates the more neutral meaning of ‘unreliable’, ‘worthless’. For the record, Donatus sides with the OLD here, but I agree with Shipp that the richer metaphorical meaning seems well suited to the context.

23. This first scene has an important role to play in Terence’s biography: when he arrived at Caecilius’ house as an unknown poet to get approval for this, his first play, he was first kept standing but after only a few lines was invited to the table (Suet. Vita Terenti 2). Given the interest that the Vila shows in Terence’s life as a slave and his manumission based on his beauty and talent, it is interesting that it does not offer, as later authors have done, the potential parallels between the content of this scene and Terence’s own life. See esp. Jacoby (1909); see also Hunter (1985), 153 n.16, who says about the ancient biographical tradition that Terence had been freed by his master, ‘I cannot suppress the observation that this detail reminds me of the exchange between Simo and Sosia at the opening of Terence’s first play, the Andria (cf. especially vv. 35–9).’

24. The only other clear example I know of is the chorus of freedmen in Plautus’ Poenulus. For a reading of both the Plautine example and the current scene in terms of historical evidence for the legal position of freedmen, see Rawson (1993), who advances the speculation that freedmen might have played an important part in the togata, in contrast to their poor showing in the palliata.

25. For recent assessments, see Anderson (2003/04); Kruschwitz (2004), 45–47. For a reading that sees this in the light of ‘Sklaverei und Humanität’, see Buchner (1974), 31. Hunter (1985), 35, finds Simo iustus and clemens.

26. Sosia’s obsequiousness is instantiated in his willingness to offer a moralising bromide on the value of obsequiousness in the current climate: namque hoc temporelobsequium amicos, ueritas odium parit (‘these days, obsequiousness brings you friends, telling the truth makes you unpopular’, 67f.). Rawson (1993), 223, sees this as evidence that obsequium could not have had its technical, legal meaning by this time.

27. Nurses are obviously particularly prominent in this respect; see, e.g., Phormio 728–65, Menander Samia 236–61 (this isn’t a recognition scene in the usual sense, but does perform the function of revealing the identity of the child’s father).

28. Of special interest here is the collocation seruibas liberaliter, a compliment that superficially erases but more deeply affirms the difference between the free and the unfree. Donatus is absolutely right when he glosses liberaliter as bene, explaining ‘Because all that is good is associated with the free, all bad with the slave’ (quia omne bonum libero aptum est, malum semo), but I think it’s fair to say that his literal-mindedness here dampens the effect of this striking phrase.

29. In his note on line 49, Donatus is specific that Apollodorus gave the site of the initiation as Samothrace, thus confirming that even this level of detail in the speech was present in the Greek original. This deduction has been challenged, most strenuously by Lefèvre (1969 and 1978), but more recently and more convincingly upheld by Sommaruga (1998).

30. On this scene, see Spranger (1984), 68f.; Cupaiuolo (1991), 85f., esp. n.12. I know of no other evidence for this kind of gift-giving from slaves to masters’ children within the Roman household, though the birthday of the paterfamilias was certainly an occasion for gift-giving; for evidence from comedy, see Plaut. Pseud. 165–229.

31. Contra Bohm (1976) who connects the theme of money and friendship sounded here to the rest of the play. This is true, but it is such a common theme that without further specifications of the way it gets discussed here (esp. its transference to the slaves) it seems hard to agree that this is a very pointed foreshadowing. See also Gilula (1991), 634f, and Sommaruga (1998), 389, the latter offering the pertinent observation that what we have here and in the protatic scene to the Hecyra are the facts of the plot filtered through the consciousness of someone who is both unrelated to the household and displaced in social status from the citizen family.

32. It is interesting that Donatus finds Davos’ speech overblown and comic: garrulos seruos et sententiosos atrial comoedia, tristes etparce loquentes tragoedia (‘comedy prefers talkative and sententious slaves, as tragedy prefers sober and taciturn ones’, ad 41).

33. Cf. Plaut. Aul. 724a-5 (egomet me defrudauilanimumque meum geniumque meum, ‘I’ve cheated myself, denied my own inclinations and desires’), spoken by the paterfamilias; and a more general application in True. 183 (istos qui cum geniis suis belligerant parcepromi, ‘those tight-fisted guys who are at war with their own desires’).

34. OCD ³, s.v. genius; cf. ingenium, another word which expresses social value in the language of biological birth.

35. The scene in the Eunuch (308–11) in which Chaerea seeks Parmeno’s help is quite similar but written specifically with the younger master in mind. The favour he asks to be remembered for is not manumission, as in the Andria, but secretly bringing the contents of the larder to Parmeno in his room. Chaerea, very like the patron in the Andria, reminds Parmeno of their past together, emphasising his generosity in the context of asking for a favour. Note too, that this is about the material resources of the household (penum), the son acting not to preserve these (as he might in his role as future paterfamilias) but perhaps doing more to favour the long-term succession strategy of the household in using them to (try to) bind a slave to the household and its interests. This scene also plays on the idea of the alliance between sons of the household and slaves: in Plautus they are united by (ineffectual) rebellion against authority (ineffectual for slave, unnecessary for son whose real interests are the same as the father’s), but here this ‘playful’ (i.e. infantilising of slaves) alliance is shown to be part of household politics, binding slave to the future authority of household.

36. Cupaiuolo(1991),86n.l2.

37. One play in which this is especially foregrounded but does not fall within the scope of the discussion here is the Adelphoe. See e.g. Fantham (1971); Leigh (2004), 158–91. One particular way this pattern might be related to the argument offered here is the importance of Micio’s odd household, formed not by natura but consiliis (126), and the fact that he winds up at the end of the play with an explicitly non-reproductive (931) household, and losing a pair of slaves; cf. also his teasing representation of a mixed-up use of the household personnel at 745–53. For a reading of the play that foregrounds the issues of adoption, see Anderson (1995).

38. an quoiquamst usus homini se ut cruciet? (‘is it necessary for any man to torture himself?’, 81). Note that this line sums up the thought of the Greek title, but rather than using a word related to ulciscor employs the term that is specific to slave punishment.

39. Except for those (few?) who can ‘patch together’ (exsercirent) their living expenses by working on the farm. This implies that the majority of Menedemus’ slaves were used for the display and luxury associated with urban life, a somewhat different picture from that we saw in 124–27.

40. Konstan (1995), 120–30. See esp. 129: ‘Menedemus relents in the case of Clinia and Antiphila, and because the contrast between the courtesan [i.e. greedy Bacchis] and the concubine [i.e. modest Antiphila] is drawn so sharply in this comedy, Menedemus seems right to have done so. Antiphila’s character and way of life, as opposed to Bacchis’, compel sympathy irrespective of her status, though in the end her status is, as it must be, converted to that of a citizen.’

41. Even if we agree with the argument of Knorr (1995) that Bacchis is less bad than she is at first set up to be, we can note that the way this setting up happens (and so an index to the moral frame of the characters) is through the contrast between her entourage and that of Antiphila.

42. Kruschwitz (2004), 50, interprets this as making a plug for improvisation over scripted performance, which is possible but still it exhibits a sense of theatre in which truth overtakes fiction every time.

43. Although the plays of Plautus are full of disguise tricks, rarely even there is the freeborn young lover the one who dons a costume. One exception is Pleusicles in the MG, who appears briefly in the final act as a sea captain and generally makes a hash of it; and even he, who has little honour to worry about, laments the degradation of dressing up (1284–89). (Gorier [2000], 283, remarks on the parallel between Pleusicles and Chaerea; Smith [1994], 22, notes the rarity of freeborn characters in disguise.) Another Plautine example which comes closer to what we have here is Philocrates in the Cap., who trades identities with his slave in order to make his escape; there, as here, the overarching frame of naturalism both offers us the spectacle of a young man assimilated to a slave and keeps the meaning of it within bounds. Another parallel to consider is Sostratos in Menander’s Dyskolos, who dresses as a farmer and works in the fields in order to give the appearance of proper seriousness and to gainsay the implications of his urban upbringing. But in Terence’s other plays no one wears costumes; there are plenty of tricks and lies and pretences, but they are told through language, not produced visually on stage.

44. Cf. Plautus’ Persa which produces a similar hybrid, though from the opposite direction (a slave ‘disguised’ as an adulescens amans); the difference in effect between these two plays is due in large part to the fact that the Persa is as situated in metatheatrical farce as the Eunuch is in naturalism, so the way that Toxilus ‘becomes’ a freeborn young man there is through metatheatrical allusion, not through physical disguise.

45. He tries to discard it (610) and fails (840–47); Frangoulidis (1993), 149f., thinks he has discarded it when he emerges from Thais’ house for the last time, since at 1030 Thraso doesn’t recognise him, but there is no unambiguous mark in the text of his returning to his adulescens guise.

46. Amerasinghe (1950), 67; Zeitlin (1998), 151.

47. E.g. Smith (1994), 27; James (1998), 39; Zeitlin (1998), 149–51; Fantham (2000), 297. Each of these scholars makes something slightly different of Chaerea’s speech, but all note how inaccurate it is as a description of Thais’ business practices as we see them in this play.

48. See 366–75. Considering the success that Jupiter has in Plautus’ Amph. once he takes on the disguise of a lowly mortal, perhaps it’s not surprising that Chaerea takes the god’s disguise tricks as his inspiration (584–91). See Barsby (1999), ad 588, on the illogicality of expressing Jupiter’s change as both a shower of gold and sese in hominem conuortisse (‘changing himself into a man’).

49. Dessen (1995) has many good observations on the ways that the associations and metaphorical meanings of the eunuch make this figure the ‘controlling metaphor’ of the play. She notes at 128 that ‘…Dorus and Chaerea express the double social construct of the eunuch. Old and young, ugly and attractive, impotent yet oversexed, physically powerless yet mentally powerful—the eunuch holds all these contradictions within himself, but for this very reason could never be enacted as one character.’ Cf. also 133: ‘The eunuch thus becomes more than a tawdry symbol of deficit; as impersonated by Chaerea, he becomes a space of possibility for change.’ I would emphasise here ‘as impersonated by Chaerea’; i.e., it is not the eunuch himself, but the conflation of the eunuch with the young man (a conflation made possible by the fluidity Dessen sees in this figure) which produces this space of possibility.

50. E.g. Brothers (2000), 30f.; Barsby (1999), 20; Lape (2004), 93 n.79.

51. Cf. Lape (1994), 102, on the ideological importance of pregnancy resulting from rape.