Widely different views have been held concerning the structure of Plautus’ Menaechmi. On the one hand, the sequence of misunderstandings arising from the presence in the same city of a pair of identical twins with the same name has been likened to clockwork and attributed in essentials to an unknown Greek dramatist.Footnote ¹ On the other hand, E. Stärk has stressed features of the play which are typical of improvised comedy and put forward the bold theory that it was constructed by Plautus himself, following traditions of pre-literary Italic drama but using stock motifs of Greek New Comedy.Footnote ² I wish to suggest that the truth lies between these extreme positions.

That the broad comedy of Plautus, in contrast to the subtler comedy of Terence, owes a debt to Italic traditions of improvised drama has long been communis opinio.Footnote ³ Stärk argues that aspects of the Menaechmi are unlike what we know of Greek New Comedy but can be paralleled by the unscripted performances of the Commedia dell'arte, for which considerable evidence is available. Many of the features, however, which Stärk sees as inspired by Italic sub-literary drama, such as loose structure, inconsistent characterization and amorality, can be attributed to superficial rewriting by Plautus and need not imply that the basic structure of the play was not inherited from a Greek model. Inconsistencies of characterization can result from small additions by Plautus, momentary jokes or inflated cantica. Inconsistencies in the movements of characters can result from additions, omissions or other changes by Plautus, like the one now documented in the Bacchides.Footnote ⁴ The amorality of the conclusion of the play, whereby the ill-used Matrona is to be auctioned off, can be attributed to an ad hoc joke in a passage otherwise rewritten by Plautus. That the quest for a lost twin brother advertised in the prologue tends to be forgotten in the play may be due to Plautine omissions. Often Stärk fails to distinguish between the surface structure of the play, in which Plautine contributions are only to be expected, and the main lines of the plot. As to the basic improbability of the plot, whereby Menaechmus of Syracuse (S), who is searching for news of a lost twin brother, is so slow to guess that he is being mistaken for this twin brother, this is acceptable in a play, which is not real life and requires a willing suspension of disbelief.Footnote ⁵ It is noteworthy that the fundamental implausibility of Menaechmus’ failure to have any suspicion that the cause of his puzzling reception might be connected with the known fact of his having an identical twin brother is somewhat mitigated by Messenio's urgent warnings of the trickery to be expected by a visitor in a strange city (258–72, 338–45, 375–7, 415–16, 438, 441–2), as well as by repeated hints that his interlocutors might be mad (282, 289–92, 309, 312–15, 325, 336, 373, 390, 394), a recurring motif throughout the play, culminating in the pretence of Menaechmus S to be mad himself and its hilarious consequences. Elsewhere it is suggested that a speaker is joking (317–18, 381, 396).

Stärk's theory has had a generally sceptical reception.Footnote ⁶ A strong argument against it has been found in the fact that the action of the Latin play seems to conform to the five-act structure which we now know to be the norm for Menander and can reasonably suppose was the convention for other dramatists of New Comedy. Long ago Legrand postulated a Greek play with act-divisions at 225, 445, 700 and 881.Footnote ⁷ Up to 701 the action is broken three times by an empty stage during which a lapse of dramatic time has to be supposed. This is fully in accordance with Menander's use of choral interludes to separate acts. Given that palliatae were at least normally based on a particular Greek model,Footnote ⁸ it is far more likely that this structure is inherited from a Greek play than that Plautus created it independently. Nor is it likely to be a coincidence that the action involving a corona in 555–66 conforms to a Greek convention concerning dramatic space, namely that the wing-exit supposed to lead to the harbour and that leading to the forum are on opposite sides of the stage. The evidence is confused but at least points to such an opposition being the norm, even if there is doubt over some aspects of the convention.Footnote ⁹ It has also been noted that the Menaechmi conforms to the observed practice of Menander in being actable by only three actors;Footnote ¹⁰ but that could be accidental and in itself is a weaker argument, especially as the nature of the three-actor rule is disputed.Footnote ¹¹

The comparison with clockwork is not inappropriate to this whole section of the play. The encounters of the twin brothers with other characters in different combinations lead inevitably to comic misunderstandings. Of course, Plautus contributed much of the detail,Footnote ¹² but there is no good reason to doubt that the basic structure derives from a Greek play. A structural change is probable only in one scene. While Peniculus is in Matrona's house, a maid comes out of Erotium's house carrying a gold spinter with a request from Erotium that Menaechmus S should take it to be refashioned. As has not gone unnoticed, the spinter is a dramaturgical doublet of the palla. Footnote ¹³ It has no consequences later in the play separate from those associated with the palla. The little episode of the maid and spinter would not be missed if it were omitted. Moreover, it is unrealistic that the maid should apparently expect to find Menaechmus S still nearby long after he had left her mistress’ house at 466–8. The most likely hypothesis is that the spinter is a Plautine doublet of the palla.Footnote ¹⁴ Such duplication by Plautus has probable parallels in the second letter of the Bacchides,Footnote ¹⁵ in the second trick of the Poenulus,Footnote ¹⁶ in the Sceledrus episode of the Miles Gloriosus,Footnote ¹⁷ in the secondary scheme to extract twenty minae from Simo in the Pseudolus,Footnote ¹⁸ and in the use of the Fides temple as a hiding place in the Aulularia.Footnote ¹⁹

It has generally been supposed that the fourth Greek act-division was at 881, where the stage is empty and a lapse of time is required before the return of Senex, father-in-law of Menaechmus of Epidamnus (E), with a doctor (cf. 882–8). There are serious grounds, however, for seeing extensive Plautine innovation in the following section of the play. Menaechmus S enters at 701 and meets Matrona, who at first is pleased to see that he is carrying the palla (705), but becomes increasingly angry during their exchanges and finally at 736 sends her slave to fetch her father. Stärk notes various incongruities in the dialogue and is no doubt right to see a substantial contribution by Plautus,Footnote ²⁰ but there is no good reason to deny the basic situation to a Greek model. Senex arrives quickly in 723; we must suppose that he lives in the vicinity.Footnote ²¹ As he enters, he delivers a bacchiac canticum on his inability to move fast. For over fifty lines Menaechmus S is idle during Senex's entrance-canticum and following dialogue with Matrona before Senex accosts him at 809;Footnote ²² this unnaturalness is hardly mitigated by hints in the text that he has angrily separated himself from Matrona (777, 810). Plautine expansion, however, is almost certain in the canticum,Footnote ²³ and very likely also in the following dialogue, in which Senex shows surprisingly little sympathy for his daughter's complaints. Senex is characterized as a long-winded comic figure, who sometimes preaches the typical sentiments of a Roman paterfamilias (787–9),Footnote ²⁴ sometimes expresses conventional attitudes of the palliata, which is hostile to the uxor dotata (766–7) and condones the sexual escapades of men (790–7).Footnote ²⁵ There is no reason, however, not to attribute the basic dramatic situation to a Greek dramatist. Finally, Senex is convinced by the apparently absurd statements of Menaechmus S that he must be mad, a conclusion which the latter encourages with a paratragic pretence of madness.

Now it is essential to the following action that Senex should believe Menaechmus S to be mad. This motivates his calling for the help of slaves (844–6) and ultimately the assault by his slaves on Menaechmus E instead (990–1000). The pretence of madness, on the other hand, is not essential, although very amusing. E. Fantham has made a good case for attributing Menaechmus’ pretence of madness to Plautus and her arguments can be strengthened.Footnote ²⁶ Divine possession is attested for New Comedy by Menander's Theophorumenē, but, as Stärk observed, a Greek dramatist is unlikely to have imagined possession by two gods, Dionysus as well as Apollo, whereas the cult of Bacchus was topical in Plautus’ Rome.Footnote ²⁷ The appeals to Apollo are repetitive (840 imperat, 850 iubes, 855 imperas, 858 iubes, 862 imperas, 868 iubes),Footnote ²⁸ and the colourful insults which Menaechmus S incorporates in his pretended ravings (837 rabiosa canes, 864 leonem uetulum) bear the stamp of Plautus.Footnote ²⁹ Finally, there is an unmistakable similarity between the chariot motif of 862–71 and Merc. 931–46, which bears an even more striking resemblance, both in the imagined journey through a series of Greek cities and in the imagined chariot, to the madness of Heracles described in Eur. HF 922–1015. The usual view is that in the Mercator Plautus was following the Emporos of Philemon, who was parodying Euripides;Footnote ³⁰ and FanthamFootnote ³¹ suggests that in the Menaechmi Plautus borrowed the idea from his Mercator. In fact, however, there are strong grounds for supposing the passage of the Mercator drastically rewritten by Plautus, and it is more likely that in both plays Plautus was directly inspired by the famous Euripidean scene, either in the original Greek or in a Roman adaptation.Footnote ³² Whether or not Plautus was directly inspired by a Latin version of Euripides’ Hercules Furens of which we have no record, there is a general consideration which points to Plautus rather than to a Greek dramatist as the creator of the Menaechmi scene. Besides the madness of Heracles, the other known cases of madness in Greek tragedy—Orestes pursued by the Furies in Eur. Or. 34–45 and IT 238–339, Ajax slaughtering cattle in Soph. Aj. 214–327—are narrated rather than enacted on stage. There is evidence, on the other hand, that Roman audiences of Plautus’ time were familiar with, and no doubt enjoyed, scenes of madness enacted on stage, certain at least in Ennius’ Alcmeo (fr. XV Jocelyn).Footnote ³³

Again there are strong grounds for regarding the whole doctor-episode as a Plautine addition. This was first proposed by T. LadewigFootnote ³⁴ and argued in detail by W. Steidle.Footnote ³⁵ This episode differs from the rest of the play in that its comedy is not based on mistaken identity but on satire of the medical profession, and the satire is more appropriate to the circumstances of Plautus’ Rome, when Greek doctors were a novelty and regarded with suspicion, than to the Athens of the era of New Comedy.Footnote ³⁶ The whole episode, however funny, is dramatically dispensable and does not advance the action at all.Footnote ³⁷ Moreover, there are certain incongruities which can best be explained by the hypothesis of a Plautine insertion. At 645 Senex prepares to fetch slaves to bind the supposedly mad Menaechmus S, but fails to carry out this plan and at 875 instead goes to fetch a doctor; at 952–6, however, Senex is instructed by the doctor to fetch slaves. It has been argued that the change of plan adopted at 875 is justified by the increased violence of Menaechmus’ threats in his pretended madness;Footnote ³⁸ this argument, however, fails to explain the other incongruities noted by Steidle and, if the pretence of madness is a Plautine addition, it in fact supports Steidle. The violent threats of 831–71 do indeed motivate the change of plan, according to which Senex decides to fetch a doctor, but can be seen as designed by Plautus, amongst other things, to prepare for his insertion.

Steidle took Plautus’ insertion to be a case of ‘contamination’,Footnote ³⁹ but it is questionable whether this term is appropriate, even in the broad sense in which it has come to be used. Doctors are occasionally criticized in New Comedy, in Philemon's Sikelikos for hypocrisy (fr. 78 K.–A.), in Men. Mon. 659 for ineffectiveness; and Athenaeus attributes plays entitled Iatros to Antiphanes, Aristophon, Philemon and Theophilus. There is no definite evidence, however, for scenes depicting a doctor as a comic type. In contrast, the comic mageiros, another professional, is well documented as a stock type, and the difference is hardly due just to the culinary interests of Athenaeus. The doctor-scene of Menander's Aspis is not really like that of the Menaechmi. The essential dramatic function of Menander's bogus-doctor is to deceive Smicrines and to that end he needs to be credible. That he purports to speak Doric, especially associated with the medical profession, is part of his disguise, and his diagnoses use genuine technical terms.Footnote ⁴⁰ As far as one can tell from a damaged text, he is no braggart but, at most, defensive in defence of professional dignity.Footnote ⁴¹ The comedy of the scene lies in the situation, not in ridicule of a quack. In any case, FanthamFootnote ⁴² convincingly demonstrates the ‘umorismo latino’ of the Latin scene in its present form, even though she holds a nucleus of it derived from the Greek original. She shows that the diseases and the remedies referred to are attested in other genres of Latin literature and would be familiar to the Roman man-in-the-street. The questions put by the doctor to the patient about his drinking habits, eyes and digestion are random and prompt comic retorts from Menaechmus (915–19, 923–6, 928–9). The references in 936–43 to the rabiosa femina … canes of 837 and to the threatened chariot-attack of 868–9 must be Plautine, if Menaechmus’ pretence of madness is. The latter prompts a retort referring to the sacra corona of Capitoline Jupiter, the absurdity and Plautine origin of which is established by the occurrence of the same motif in a fuller form at Trin. 85; and for an impossible crime of sacrilege Senex is said to have been imprisoned and beaten under a (Roman) furca (942–3). In short, some scene of a Greek play could have provided inspiration for Plautus but, if so, hardly more than the basic idea. Here Stärk was surely right to see free composition.Footnote ⁴³

There is some further awkwardness in the stage-action at the end of the doctor-scene which can also be ascribed to Plautus’ alterations. The doctor makes his exit at 957 having given instructions for slaves to carry Menaechmus E to his house. Menaechmus E remains alone on stage and delivers a short monologue, 957–65, in which he reflects on how he has been treated as if he were insane and that as a result he is now excluded both from his own house and from that of Erotium, repeating what he had said in his earlier monologue (698–9). He therefore sits down at the side of the stage in the forlorn hope that at least he will be admitted in the evening; it has to be supposed, unrealistically, that he did not hear the doctor's order to summon slaves in 952–6, which should have warned him to make his escape.Footnote ⁴⁴ The dramaturgical reason for this piece of stage-business is clear: he must remain on stage in readiness for the arrival of the lorarii in 990, but must be kept out of the way during the long canticum of Messenio (966–89). It has long been recognized that this canticum, on the favourite Plautine theme of slave-duties, is at least largely the work of Plautus.Footnote ⁴⁵ GratwickFootnote ⁴⁶ supposes that Plautus brought forward Messenio's entrance and that in the Greek play he did not enter until after the assault on Menaechmus E had begun. This may be true, but in any case we can assume that in the Greek play Messenio's entrance-monologue was brief, that there was no long gap between the entrance of Menaechmus E from the town (899–907) and the arrival of the lorarii, and no need for him to sit idly on the stage for an extended period.

If the doctor-episode is removed, together with the preceding pretended madness of Menaechmus S and the following canticum of Messenio, during which Menaechmus E sits idly on stage, the following sequence of action remains. At 701 Menaechmus S returns from the direction of the harbour, still carrying the palla, meets Matrona, who naturally mistakes him for Menaechmus E. Increasingly angered by his attitude, she sends a slave to fetch her father (734–8), who soon appears. Mystified by what appears to be totally irrational talk, in particular violent denial, with extravagant oaths, of ever having entered her house, Matrona and Senex come to the conclusion that her supposed husband is suffering from an attack of madness, and go to fetch slaves to carry him home, either from her house on stage or from the old man's house in the vicinity. Menaechmus S then seizes the opportunity to exit towards the harbour, where he expects to find Messenio (878 quid cesso abire ad nauem dum saluo licet; cf. 552). No sooner has Menaechmus S left in the direction of the harbour than Menaechmus E enters from the forum (899–908). He is immediately seized by the four slaves fetched by Senex, but rescued by Messenio, who appears just in time (1009–17). Messenio of course mistakes Menaechmus E for his master Menaechmus S and claims his freedom as a reward for his help. This suggested scenario is necessarily speculative and subject to modification in detail, but it could well form a Greek fourth act, during which first Menaechmus S is mistaken for Menaechmus E and accused of madness, then Menaechmus E suffers for the actions of Menaechmus S. The assault on Menaechmus E by the four slavesFootnote ⁴⁷ and his rescue by Messenio would provide a spectacular climax to the play, such as occurs in the fourth act of several of Menander's plays.Footnote ⁴⁸ The act would end at 1049, where the stage is empty in the Latin play, with Messenio returning to the inn to fetch the marsuppium (1035–7) and Menaechmus E going to call on Erotium in a final attempt to persuade her to relent (1048–9).Footnote ⁴⁹ An act-break at 1049 would provide more time for Messenio's trip towards the harbour; in the Latin play it takes place during the monologue of Menaechmus E (1039–49) and Menaechmus S is supposed to have met Messenio on the way (1056–7). Plautus’ insertion can be regarded as extending from 830, when Menaechmus S pretends to be mad, to the entry of the lorarii at 990, with the exception of 844–5, when Senex sends for the lorarii, and 876–8 and 898–9, which effect the exit of Menaechmus S and entry of Menaechmus E respectively. The fourth act thus postulated would have a symmetry similar to that of the previous act. This is as far as we can go in reconstructing our hypothetical Greek play. GratwickFootnote ⁵⁰ has amply demonstrated that the final scene of the Latin play, dominated by Messenio as detective, must have been completely rewritten by Plautus. What was in the final act of the Greek play is impossible to say, although the dénouement was presumably brought about, as in the Latin play, by a face-to-face meeting of the twins. At 1050 Menaechmus S re-enters with Messenio from the direction of the harbour, mystified and angered by Messenio's claim to have rescued him and to have been given his freedom; we might perhaps have expected a rather longer dialogue in which the marsuppium played a larger role (cf. 265, 701–2, 1036–7, 1043–5). In Plautus’ play the dénouement comes very suddenly with the entry of Menaechmus E from Erotium's house and the face-to-face meeting of the twins. Gratwick plausibly suggests that what Plautus omitted included more information about the family background of the plot.Footnote ⁵¹ This could have provided a more realistic ending than the cynical joke with which Plautus disposes of Matrona (1160–1).

To sum up, underlying the Latin play it seems possible to see a neatly constructed five-act Greek play which can be summarized according to the involvement of the protagonists as (1) Menaechmus E (2) Menaechmus S (3) Menaechmus S / Menaechmus E (4) Menaechmus S / Menaechmus E (5) Menaechmus S + Menaechmus E. In the Latin play this basic schema has been enriched, not only with the musical element of five polymetric cantica,Footnote ⁵² but with much extra comic material, in particular the ancilla and spinter scene, the pretended madness of Menaechmus S, the quack-doctor's examination of Menaechmus E, and the final recognition scene in which the slave Messenio takes control. For all this, on the hypothesis here offered, we should be grateful to the genius of Plautus.