The scripted audience in Roman comedy

1 Introduction: historical reality and make-believe

As twenty-first-century classicists (unlike our counterparts forty or fifty years ago) we are used to treating dramatic works as performances rather than as literary texts: we scarcely need to be reminded nowadays that Greek and Roman plays were theatrical events that took place in front of audiences.^[1] Nevertheless, despite the prevalence of audience-centred criticism, it is worth reminding ourselves that we know very little for certain about these ancient audiences. (This is true whether we happen to be thinking about Greek or Roman audiences.) We are aware that our plays were designed to be acted out in front of someone, in a theatrical venue of some sort, but usually we can only speculate who was actually in the audience, or exactly what size that audience was, or what they were sitting or standing on, or what shape and size the theatre was, or exactly when the performances took place, or exactly how the spectators responded to the performances. We possess surprisingly few pieces of evidence for theatres and audiences in the periods in which we are most interested, and such scraps of evidence as we do possess are of a miscellaneous and inadequate sort.^[2] Inevitably much of what is written on the topic must take the form of educated guesswork.

This article is concerned with the audiences of the comedies of Plautus and Terence in early Republican Rome.^[3] The wretched state of the evidence needs to be emphasized at the start because theatre historians have sometimes tended to assume that we can know more about these audiences than we actually do. Methodologically this assumption is based on what is sometimes called the ‘hermeneutic circle’: that is, in the absence of sufficient external evidence for the historical context, scholars have had to resort to internal evidence from the plays themselves. Plautus’ and Terence’s comedies tend to be treated as data for their own interpretation: the so-called context is reconstructed largely from the plays’ content, and these comedies end up being treated simultaneously as the texts to be interpreted and as the supporting evidence with which to interpret themselves. I find this sort of circular approach unsatisfactory.

The ‘internal evidence’ from within the plays of Plautus and Terence includes various direct and indirect references to the spectators, or descriptions of specific groups within the audience. It might also include (broadly speaking) the entire contents of the plays, which scholars have used as a way of assessing the audience’s composition and socio-political status, or their average level of sophistication, taste or theatrical competence. The fact that scholars have used this evidence to reach such widely differing conclusions (e. g. on the crucial question of whether or not slaves were present in the theatre, or whether women were widely represented alongside men) is a sign of how precarious this sort of enterprise is.^[4]

This ‘internal evidence’ also includes moments within the comedies at which the characters appear to be directly addressing the spectators. For example:

mirum uideri nemini uostrum uolo, spectatores,

quid ego hinc quae illic habito exeam: faciam uos certiores.

I don’t want any of you, spectators, to start wondering why I’m exiting from this house when I live in that one – so let me put you in the picture.

Utterances of this sort are often encountered in prologues, which may be thought to have a semi-detached or extra-dramatic status, but they may also appear in the mouths of various characters at any point during the main action of a play. (The lines above are spoken by the slave girl Stephanium at Plautus, Stichus 674–675.) Moments when the characters explicitly acknowledge the presence of an audience are inherently metatheatrical. They make us artificially aware of a contrast between two distinct worlds that co-exist in the same moment: the fictional plot-world inhabited by the characters, and the real world of the audience. Such instances of audience acknowledgement are often described as ‘breaks in the illusion’ (or similar phrases).^[5] They do not always take the same form, and they might be exploited for various ends within each dramatic context, but essentially they remind us – just in case we should need reminding – that the world of the play is a fiction and that it is separate from real life. Metatheatrical gestures of this type have been treated as realistic encounters with the Roman audience members, as if they constituted straightforward and largely accurate descriptions of the people whom the actors could see in front of them in the auditorium. Even when due allowance is made for exaggeration or humour, on the whole Plautus’ and Terence’s references to the audience have been interpreted more or less literally, as documentary data.

In this article I want to suggest another approach to the problem, based on a fundamentally different conception of the nature of these playscripts. I do not believe that the plays, however carefully interpreted, can be trusted to contain factual information about their own performance context or about the composition or behaviour of their audiences. Whatever the Roman comedians may have been doing when they set out to write their plays, they were not aiming to furnish future historians with evidence by leaving behind sober, factual descriptions of the world around them. They were trying to entertain and to arouse laughter. Nothing that we encounter in the world of comedy can safely be taken seriously or literally, even (or especially) when it looks serious or literal. On closer inspection it usually turns out to be silly, ironical, nonsensical, unreliable, distorted or exaggerated in some way. This is a principle of fundamental importance to the critic of comedy: when reading any given utterance in the script we need to be constantly alert to the possibility of humour and its consequences for interpretation. The principle applies just as much to audience address as to any other type of utterance in comedy. We should begin by assuming that it is designed to be funny, and by questioning whether the ‘obvious’, straightforwardly literal reading is the only interpretation open to us.^[6]

Another basic principle that I believe we must acknowledge is that every word in the playscript, in the form in which we have it, is part of the fictional plot-world. This applies to everything that is said to the audience or about them. All the words spoken by characters in the play, even when they seem to be ‘breaking the illusion’, are nonetheless part of the illusion. Moments of audience address may look like spontaneous outbursts of ad-libbing or improvisation on the spot, but in fact they are scripted, premeditated and rehearsed. As is now widely recognized, the literary imitation of certain aspects of traditional, improvisatory Italian drama is a frequent habit of Roman comedians (especially Plautus).^[7] These apparent ad-libs to the audience can be seen as just one among many other quasi-improvisatory features and techniques. Although the actors are pretending to step out of character, interrupt the script, and engage in real-life, real-time interactions with the audience in the theatre, this is not what they are doing. They are delivering their lines within a scripted scenario.

In this respect my own approach is significantly different from that of C. W. Marshall, whose highly stimulating book The Stagecraft and Performance of Roman Comedy features a speculative discussion of the process by which our written texts may have come into being.^[8] Marshall prefers to talk of ‘degrees of scriptedness’ or ‘blended performances’ combining scripted and unscripted elements, and he suggests that the words heard during the original performances differed significantly from the words in the written texts that we possess. This is a perfectly plausible scenario, but even if it is correct (and there is no way of knowing), all that we can really talk about, in my own view, is the text that does survive, rather than a hypothetical and fluid proto-version of the text that by its very nature cannot have survived. The scripts that we now possess were either (as I believe) fully scripted in advance before the performance, or revised and given a fixed textual form at some point after the performance (perhaps primarily for a reading audience); but either way, they cannot be treated as accurate snapshots of exactly what was going on in the theatre during the performance.

This article has a simple point to make: that the audience being addressed and described by the Roman comedians is essentially a fictional character. Of course, this character may in some respects resemble the audiences who once sat watching these comedies – but then again, it may not. It cannot (in my view) be treated as a historically accurate portrayal of the real-life audience; rather, it should be seen as one among the other fictional characters in the world of the play. Cumulatively, as I shall demonstrate, the image of the audience that emerges from all the comedies of Plautus and Terence is a stereotypical, far-fetched and anti-realistic one. Since the imaginary audience is constructed in a broadly similar way from play to play, with a distinctive ‘personality’ and set of attitudes, I suggest that it can be seen as a recurrent stock character, alongside the scheming courtesan, the silly young man, the nagging wife, the boastful soldier and all the other stock characters of New Comedy. Like other stock characters, the audience does not appear in exactly the same guise every time: there is a significant amount of variety from play to play, and the effect relies on the balance between convention and variation.

What we ought to be doing, in my view, is not trying to treat the playscripts as historical evidence for their own reception in the theatre, but instead trying to understand the function of the scripted audience within the world of the drama. What is the scripted audience doing here? How does it fit in with the plots and themes of the comedies? How does the character of the scripted audience change from play to play? How and why is it funny? How is the real-life audience supposed to respond? These are all important questions of interpretation and literary appreciation, some of which (unlike questions of historical fact) can be answered with a reasonable degree of confidence.

A useful point of comparison is provided by recent work on Greek lyric. Poets such as Sappho and Alcaeus, like our Roman comedians, constantly refer to their own setting and circumstances of performance. Such references have traditionally been treated as evidence for the poems’ real-life historical context and audience, but they are increasingly being seen as a mixture of real and imaginary elements. G. B. D’Alessio, for instance, criticizes previous readers for treating the texts of lyric poetry as ‘a simple record of the performance situation’, and shows that deictic references to the poems’ settings and addressees are part of a fictional world constructed by the poems; such elements ‘have the linguistic capacity to operate free from contextual boundaries’.^[9] David Gribble employs cognitive approaches (including Text World Theory) to make sense of the effects that lyric poems have upon their audiences: he distinguishes between the poems’ real-life circumstances of performance (the ‘discourse world’) and the imaginary world that they construct (the ‘text world’), and shows that the poems’ impact depends on complex interactions between the two worlds.^[10] Most pertinently for our purposes, Jonathan Culler has coined the phrase ‘triangulated address’ to describe moments at which poets are ‘addressing the audience of readers by addressing or pretending to address someone or something else’. In this form of ‘fictive utterance’ two separate groups (the ‘addressee’ and the ‘audience’) are being addressed simultaneously, and ‘the deictic apparatus of the here-and-now’ can denote more than one object at once.^[11]

Such approaches apply equally to written texts for readers and to dramatic performances for spectators: I suggest that they can be very helpful when applied to Roman comedy. Instances of apparent audience interaction in Plautus and Terence can be seen as a form of ‘fictive utterance’ or ‘triangulated address’ rather than as straightforward moments of communication between actors and audience. In effect, we need to think in terms of two audiences – the real-life one and the scripted, imaginary one – both of which are present in the theatre simultaneously. The dramatic effect of these scenes arises precisely because these two entities are not the same.

2 Characteristics of the imaginary audience

(i) Social composition. Some comedies at first glance seem to be offering us explicit information about the composition of their audience. It is impossible to take this information at face value because it is full of exaggerations, inconsistencies and other oddities. As several others have pointed out, many individual plays seem to use their prologues (in particular) to ‘construct’ their own audience in a slightly different way, depending on the purposes of the specific dramatic context or the humour in each case.^[12]

Take, for example, the following lines spoken by Plautus’ Pseudolus:

nunc, ne quis dictum sibi neget, dico omnibus,

pube praesenti in contione; omni poplo,

omnibus amicis notisque edico meis,

in hunc diem a me ut caueant, ne credant mihi.

Now then, so that nobody can say that he hasn’t been told, I’m telling everybody, in this public assembly, in the presence of the adult male population – I’m proclaiming to the entire people, and to all my own friends and acquaintances, that today they should beware of me and not trust a word I say (Pseud. 125–128).

If we were to read these lines literally – despite the fact that the speaker is a self-confessed liar and fraud – we might conclude that Plautus’ plays were performed to an all-male audience consisting entirely of Roman citizens. This may or may not seem plausible in itself, but our sense that this is a literal description is undermined by Pseudolus’ blatantly exaggerated claim that the entire Roman population (omni poplo) is present. His additional claim that his personal acquaintances are present is nonsensical: since Pseudolus is a fictional character, it is impossible for him to have friends in the real world. But in any case, Pseudolus does not even claim to be describing the theatre audience as such: he affects to believe that the group that he is addressing is a political meeting or assembly (contio). This is an example of ‘code-switching’, a familiar comic technique whereby a speaker uses language or gestures to imply that the scene has suddenly shifted to some different setting.^[13] On the whole, then, we have to conclude that Pseudolus’ words here are designed to be disorientating, paradoxical and funny, not a straightforward description of anyone in the real world.

The image of a male-only audience is contradicted by the prologue of Plautus’ Poenulus (17–35), which features an extended description of the composition of the imaginary audience.^[14] This time the spectators are depicted as being considerably more diverse in terms of gender, age and social class. The prologue conjures up a scenario in which the theatre is still filling up and people are still moving around, even though the play has already started. The speaker itemizes the different constituent groups in the cavea and pretends to issue instructions to each group, one by one: they apparently include male prostitutes, lictors and ushers, slaves, married women, children and even nurses with mewling infants. Were any or all of these groups really present at the first (or any subsequent) performance of Poenulus? Does Plautus mention the male prostitutes first because they were an especially prominent sub-group, or because it was funnier to do so? Did women really bring babies along and breast-feed them during the play? It is impossible to know the answers to these questions in reality.^[15] But it seems to me that the function of the description is rhetorical and humorous rather than literal or factual. The audience is described thus because in this scene Plautus is constructing an image of his comedy as appealing to a very broad public, in contrast to the lines just quoted from Pseudolus, in which he is concerned to create the impression that his drama is a terribly important business aimed at men of affairs.^[16] Plautus wants his real audience – whoever they may actually be – to assume that everyone, from all walks of life, is eager to see Poenulus. By creating the fictional image of a packed theatre, with standing room only, he is emphasizing the brilliance and popularity of his work. Maybe fiction and reality coincided, maybe not. Any mismatch between the two would have given rise to laughter because of the contrast and incongruity involved: the more absurd the contrast, the bigger the laugh. But these lines would still be there in the script, exactly the same, even if the audience was considerably less diverse than the group described, and even if the scene had to be played to an almost empty house.

(ii) Fictional biography. Here and there the scripted audience is particularized in greater depth, with invented ‘biographical’ details that might include their supposed socio-economic characteristics, marital status or personality traits. This sort of detail depends on the unrealistic assumption that every individual within the audience possesses identical characteristics, though the type of individual being evoked is not always the same from play to play. For instance, Mercury in Plautus’ Amphitryo (1–7) addresses the entire audience as if they were all tradesmen involved in the import-export business, and promises to help them increase their profits in all their future dealings and speculations. Elsewhere, the prologue-speaker in Casina (23–25) assumes that the entire audience is worried about debt, but reassures them that he has tricked their creditors into temporarily abandoning their pursuit. For the sake of the joke here Plautus imagines an audience somewhat less prosperous and further down the social scale than the audience of Amphitryo. But this image is clearly a reflection of the characters and themes of Casina’s plot-world, where the focus is on slaves, bailiffs and low-life rather than the more upmarket affairs of gods and heroes in Amphitryo. In other words, the rhetoric seems to be designed to suit the dramatic context in each case, rather than simply describing the composition of the respective audiences. (And of course the playwright could not have known, when writing the play, exactly who would be present in the theatre on the day of the performance.)

The gender composition of Greek and Roman audiences is a debated topic, and (as we have already seen) the comedies themselves give us contradictory information on this point. But sometimes humour is generated from the assumption that the audience is not just male but composed exclusively of misogynists – as in the following lines from Plautus’ Cistellaria (678–681):

mei homines, mei spectatores, facite indicium, si quis uidit,

quis eam apstulerit quisu’ sustulerit et utrum hac an illac iter institerit.

non sum scitior quae hos rogem aut quae fatigem,

qui semper malo muliebri sunt lubentes.

Spectators! My dear fellows! Did anyone see who it was who picked up the casket and went off with it, and whether he was heading this way or the other way? If so, give me a sign!

[Aside] I’m none the wiser for bothering these men with my questions: they always enjoy it when women suffer.

Here the speaker (Halisca) treats the spectators as witnesses who might in theory communicate with her and tell her what has happened on stage during her absence. They cannot do this, of course. Even if anyone in the real-life audience does shout out the answers to these questions, the character/actor affects not to hear them and continues delivering the scripted lines; the fictional audience remains obstinately silent. Halisca’s rueful aside assumes both that there are no women in the theatre and that an all-male audience would naturally never identify or sympathize with any female character. Whether or not these assumptions are valid, it is obvious that in this particular scene, as elsewhere, the rhetoric is designed to fit the context. Halisca’s predicament may seem more desperate if she is all alone with no one on her side, but more simply it is funnier to pretend that the spectators are all deliberately refusing to cooperate. The fact that the part of Halisca is played by a male actor adds a further level of ironic humour to these lines.

Another female character, Pardalisca in Plautus’ Casina (1015–1019), makes similar assumptions about the audience when appealing for applause at the end of the play:

nunc uos aequom est manibus meritis meritam mercedem dare:

qui faxit, clam uxorem ducet semper scortum quod uolet;

uerum qui non manibus clare quantum poterit plauserit,

ei pro scorto supponetur hircus unctus nautea.

Now it’s your duty to put your hands together and give these meritorious men the remuneration they richly merit! Anyone who does this will always take home his first choice of prostitute, without his wife finding out; but I’m telling you, anyone who doesn’t give us a big hand, as noisily as he knows how, will go home with a billygoat scented with bilgewater instead of a prostitute!

Pardalisca treats the audience as if all or most of them were married men. Now this is obviously an unrealistic assumption (even if they really were all male), but it is adopted for the purposes of the joke and the final punchline. It is also assumed that all these married men despise their wives – just as Lysidamus in this comedy loathes his wife Cleostrata – and that they are all prone to unfaithfulness, and that every one of them desires sex with prostitutes more than sex within marriage. These ‘biographical’ details cannot literally be accurate. But they are entertaining because they embody a recurrent fantasy of escapism and sexual licence that is encountered in other Roman comedies,^[17] and they are funny in the specific setting of this play because the whole plot of Casina revolves around the idea that a married man will naturally prefer a sexy younger woman to his horrible ageing wife. There is additional humour here in that the characters seem to believe that they can have an effect on the life of the spectators after the play has finished, or that the audience’s own behaviour can influence what happens next. (Similar beliefs are adopted by other comic characters elsewhere, as we shall see.) Once again the mode of audience address combines fantasy, exaggeration and make-believe in a way that rules out literal interpretations.

(iii) Homogeneity. Although spectators are sometimes divided into subcategories or singled out as individuals, normally they are treated as a single, homogeneous entity. It hardly needs pointing out that any real-life audience would have been a diverse group composed of a large number of individual people.^[18] But the scripted audience is treated by the comedians as an undifferentiated mass. Usually it is imagined, unrealistically, as having a single concerted reaction to what is happening on stage. For instance, the prologue-speaker in Plautus’ Truculentus (4–7) pretends to see the entire audience either nodding or shaking its head in reply to questions that he asks them. Mercury during the prologue of Plautus’ Amphitryo breaks off in mid-flow as he pretends to see the audience reacting to what he is telling them: quid? contraxistis frontem quia trageodiam / dixi futuram hanc? (‘What? Are you making a face because I told you that this play is going to be a tragedy?’ 52–53). Here he does not just describe the audience as having a unified psychological response to his words; he actually describes them as if they had a single body and a single forehead to wrinkle in disbelief. Later in the same speech Mercury again pretends to see the audience reacting en masse to what he says: quid? ammirati estis? (‘What? Are you surprised?’ 89). The unanimously startled disbelief of the scripted audience is not, of course, a reliable index of the cognitive state of every single real-life spectator; it is a fictional conceit that exaggeratedly trumpets the originality and generic inventiveness of Amphitryo.

On another occasion (Plaut. Merc. 160), the entire audience (or a significant contingent within it) is imagined as having fallen asleep in their seats:

dormientis spectatores metuis ne ex somno excites?

Are you worried in case you stir the sleeping spectators from their slumber?

The detail (as usual) arises from the situation unfolding in the plot at this particular point, and (as usual) it is a joke rather than a description of the real-life spectators’ response to the play. The slave Acanthio, impatient to know everything that is going on, wonders why his young master Charinus is hesitating to tell him about his problems, and so he sarcastically suggests that it is because Charinus is loath to disturb the audience with a long speech. This is actually a very old joke. A version of it is first seen in fifth-century comedy, when one of Eupolis’ characters calls to the audience to wake up, as if they had dropped off through the sheer boredom of having to sit through the plays of rival comedians.^[19] Can we seriously imagine that audiences in antiquity would often fall asleep en masse in the middle of a play? It is hard to believe that any contemporary historian would take such passages at face value; but in fact, methodologically speaking, there is no reason not to do so, if we are in the habit of reading the comedians’ descriptions literally. (If we take some of them literally, why not all of them?)

(iv) Chronic boredom and hostility. The scripted audience not only reacts as if it were a single person, but it is also imagined as having more or less the same type of reaction, time and time again, to any comedy that is being performed. Both Plautus and Terence – despite considerable differences in the way in which they incorporate such references – repeatedly present an image of spectators who are easily bored, who are extraordinarily hard to please, who take a very long time to settle down and be quiet, who have a severely limited attention span, and who will never give plays a fair hearing unless they are implored to do so. This cumulative image of the audience has been taken largely at face value and treated as literal fact – some version of it is encountered in every single modern textbook and scholarly work on the Roman theatre – but it seems to make better sense as a fictional stereotype.

The speakers’ repeated attempts to make the audience quiet and receptive are usually given the label captatio benevolentiae – a description which assumes that the audience invariably lacks benevolentia to start with. There would be no need for this rhetorical tactic unless it were generally understood that the audience needed to be flattered, coaxed, persuaded or even begged to take the side of the author and his company of actors. Can we really believe that audiences were so uniformly hostile? Did the supposedly put-upon performers always deliver these earnest appeals with entirely straight faces?

quaeso ut benignis accipiatis auribus.

nunc argumentum accipite atque animum aduortite …

I entreat you to take this in with kindly ears. Now take in the plot summary and pay attention … (Plaut. Men. 4–5)

      nunc <uos> animum aduortite,

dum huius argumentum eloquar comoediae.

Now pay attention while I’m explaining the plot of this comedy … (Plaut. Amph. 95–96)

uos omnis opere magno esse oratos uolo

benigne ut operam detis ad nostrum gregem.

I want to make a very earnest appeal to all of you, that you may graciously give your attention to our company … (Plaut. Cas. 21–22)

quaeso, animum aduortite

I implore you, pay attention! (Ter. And. 8)

adeste aequo animo, date potestatem mihi

statariam agere ut liceat per silentium …

Pay attention, keep an open mind, and give me the chance to put on this slow-moving drama in the way I want, without making any noise … (Ter. HT 35–36)

Passages such as these are extremely common (indeed, suspiciously common).^[20] They are also utterly conventional and unvarying in their wording: they almost never seem to relate to specific comedies or individual circumstances of performance. It seems to me that they constitute a serial comic motif or running joke. This is not simply because the playwrights could not have known, when they were writing a script, whether or not the audience would behave well on the day of performance, or at exactly which points during the performance they would cause a commotion. It is also because it is incredible that every single group of spectators always responded in an identical way to every single play that was put before them. Furthermore, these dogged pleas are so very fulsome: are they not just a little too earnest to be entirely convincing? It is hard not to suspect a certain tongue-in-cheek quality in the self-deprecation of any playwright who describes his own play as ‘slow-moving’ (stataria) or implies that it is too boring to hold anyone’s interest. Sometimes the tone is palpably ironical, as in Palaestrio’s monologue at Plautus, Miles Gloriosus 79–82:

mihi ad enarrandum hoc argumentum est comitas,

si ad auscultandum uostra erit benignitas;

qui autem auscultare nolet exsurgat foras,

ut sit ubi sedeat ille qui auscultare uolt.

It’s my pleasure to explain to you the outline of this play, if you’re going to be kind enough to listen. Anyone who doesn’t wish to listen can get up and leave, so that anyone who wishes to listen can find a free seat.

In this scene Palaestrio pretends that – even after his initial warning – the audience remains inattentive, and he is obliged to plead with them again just a few lines later: date operam, nam nunc argumentum exordiar (‘Pay attention, because I’m going to start summarizing the plot now!’, 98). But this is not literally a complaint about the audience’s misbehaviour. It is a way of drawing attention to Palaestrio’s own dilatoriness and digressiveness, in the course of an extraordinarily long and rambling prologue speech which takes ages to get to the point. This sort of technique is all part of the fun of Plautine and Terentian prologues. They may pretend that their function is simply practical and explanatory, as if all the audience wanted was to hear the plot-summary (argumentum),^[21] but in fact they are designed to tease and tantalize, or more generally to provide entertainment in the form of an elaborate ‘warm-up’ act.^[22]

(v) Rowdiness. A frequent complaint, particularly in the prologues of Terence’s comedies, is that the audience refuses to be silent even when instructed or begged to do so.^[23] This complaint is odder than it might seem at first glance. Would any comedian really want his audience to sit in silence throughout the show? Let us try to imagine the effect, in a crowded theatre, of a comedy that elicited a reaction of complete silence from its audience: it would be deadening and dire. Successful comedy needs ambient noise, laughter and applause to achieve its optimum effect. A respectful or awe-struck silence might well be fitting in the case of other performance genres – notably tragedy, where any intrusive sound or out-of-place laughter would be jarring. But in comedy, by contrast, lively background noise is an integral part of the performance as well as a form of active audience participation.^[24] The actors on stage derive energy from signals that they are entertaining the spectators, while the audience members themselves enjoy the performance more if they can hear noise and be reminded that they are participating in a collective social event. (This principle is well understood by modern radio and television producers who use a ‘canned’ audience track to make comedy seem funnier.) So all these repeated pleas for silence must be taken with a pinch of salt.

The ‘noisy audience’ motif is seen in its most extreme form in Terence’s Hecyra.^[25] In the opening lines we are told that the play we are now watching has already failed – not once but twice – because of the bad behaviour of previous spectators. The text as we have it includes two separate prologues (1–8 and 9–57), which purport to come from the second (abortive) attempt and third (successful) attempt at an uninterrupted performance. In the earlier prologue, we are told, rather vaguely, that the play’s original spectators were distracted by the rival attractions of a tightrope act. The later prologue emphasizes the general uproar that ensued, offering a slightly different (and in part contradictory) account of the same circumstances. Here we learn that the initial attempt at a reperformance also failed because, despite Terence’s hopes, the second audience was just as noisy and inattentive as the first. This longer speech is framed as an appeal to the audience from Terence’s producer L. Ambivius Turpio, who begins by assuring us that the fate of Hecyra is not in fact unusual. Turpio explains that throughout his career he has often been involved with plays that were ‘driven off stage’ (exactas) by unappreciative audiences.

in eis quas primum Caecili didici nouas

partim sum earum exactus, partim uix steti.

quia scibam dubiam fortunam esse scaenicam …

In those plays by Caecilius which I was the first to put on, I was sometimes driven off the stage, while at other times I could barely get through a performance – and it’s because of this that I came to understand that a life on the stage is a risky business … (Ter. Hec. 14–16)

These lines are significant because Turpio (or the author) wants to emphasize the point that audiences are repeatedly hostile and unappreciative, especially in the case of new comedies that are being performed for the first time.^[26] If this is so, and if even comedies by the undisputed classic Caecilius Statius suffered a poor initial reception, then it cannot be (or so he implies) that Terence’s Hecyra is an especially unappealing play. We are thus encouraged to conclude that the problem lies with the audience, not with the playwright or actors.

Although Hecyra may have suffered an unusually bad fate, there is no sense that its audience behaved very differently from the putative audience of any other Roman comedy. Nevertheless, these prologues seem to be setting up two types of notional audience in opposition to one another. In the earlier prologue (1–8) the speaker implicitly flatters the audience of the first reperformance, whom he treats as his supporters, and contrasts them with the less discerning audience of the original performance, whom he disparagingly describes as ‘a stupid crowd’ (populus… stupidus, 4). Similarly, the later prologue differentiates the current spectators from unsatisfactory past audiences, flatters their intelligence (31–32), and declares that by their good behaviour they will add prestige to the occasion (44–45). However, the rhetoric of both speeches is somewhat inconsistent, since the addressees are simultaneously treated as if they are discerning and serious-minded people (unlike any other audience in recorded history) but also rowdy and volatile people who might at any moment turn nasty (like every other audience).

It is difficult to know exactly how to read the opening of Hecyra. Many readers have interpreted these prologues literally, as constituting a simple descriptive account of what happened in the theatre at the play’s first two performances. But they can also be read ironically. Several key aspects of these speeches – their odd mixture of flattery and reproof, their wheedling, over-anxious tone of voice, their ostensible humourlessness, their paradoxical combination of self-confidence and diffidence, and their pose of self-defence in the face of criticism or failure – are so strongly reminiscent of Aristophanic parabases or Callimachean polemics that I believe we are dealing with stylized literary posturing rather than straightforward factual narrative.^[27] It has even been suggested that the prologues represent pure fiction. One scholar who has interpreted them in this way is Erich Gruen, who not only draws attention to the internal inconsistencies between their accounts of the abortive first performance, but also points out that the whole scenario being described is uncannily similar to a real-life historical event in 167 BC (shortly before the composition of Hecyra).^[28] This event was a dramatic performance which was organized by the propraetor L. Anicius Gallus to celebrate his victory over the Illyrians and which descended into chaos in just the same way as Hecyra is said to have done. If Gruen is right, these prologues take on a different character. Terence can be seen as entertaining his real-life audience by constructing a fantasy in which he comes across, amusingly, as the perpetually put-upon poet, dogged by disaster at every turn.^[29] This sort of reading seems all the more compelling when we read Hecyra side-by-side with Terence’s other plays. The very same persona is encountered in all of Terence’s prologues, which repeatedly depict him in the guise of an unsuccessful writer, struggling to make his way in the literary world but continually being attacked and accused of misdemeanours by unnamed enemies and critics. All of this could be an accurate account of literary history, of course, but to me it looks more like a serial fiction constructed for the sake of humour.^[30] If Terence’s prologues are read as autobiographical reminiscences or heartfelt pleas for attention, they do not seem funny at all; the poet just comes across as a rather querulous or pathetic figure. Furthermore (as others have noted) there is something inherently unrealistic about the fact that in the prologue to Andria Terence was already presenting himself as an unpopular author, even before the staging of his first comedy.

There is at least one definite indication that the situation being evoked in the second Hecyra prologue is a fiction. At line 43 Turpio – ostensibly describing the current ‘live’ moment in the theatre – says nunc turba nulla est; otium et silentium est (‘This time there is no commotion; there is silence and repose’). This line, like all the others, was scripted in advance. The author did not know at the time of writing it what the conditions in the theatre would be. What if silence had not been forthcoming? What if the current audience had ignored the speaker’s pleas, just as on previous occasions (real or imaginary)? It would have made no difference, because the line had already been written. As in all other cases, we are dealing with a make-believe text world, not with impromptu observation on the spot. The words in the script conjure up their own imaginary setting, which may or may not coincide with the real-life circumstances of performance.

(vi) Stupidity. Another characteristic attributed to the imaginary spectators is that they are naturally slow on the uptake: it is generally taken for granted that they will find the plot difficult to follow. This can be seen as another running joke, repeated in play after play, no matter how complicated or straightforward the plot actually is. For example, the goddess Luxuria in the prologue to Plautus’ Trinummus (4–5) explains:

nunc, ne quis erret uostrum, paucis in uiam

deducam, si quidem operam dare promittitis.

Now then: just so that none of you will go astray, I’ll set you on the right track with a few words, if you promise to pay attention.

Would any audience of average intelligence be likely to go astray without this sort of help? Surely not; but the real-life audience is likely to find Luxuria’s words amusing because they would recognize them as a familiar comic formula.^[31] In this particular example the formula is given an ironical twist, since the prologue of Trinummus is extraordinarily short, by Plautine standards, and it scarcely explains anything at all. The audience have to pick up what is going on – not that this is really very difficult – on the basis of the more naturalistic scenes of dialogue that follow. In general, the idea that the audience continually needs everything to be painstakingly explained to them is part of the fun of Roman comedy. As is widely recognized, the expository function of Plautine and Terentian prologues (in contrast to their Greek models) is much less important than the generation of laughter.^[32] Despite what these prologue-speakers may pretend, there is not actually any need for an argumentum or for laborious explanations at the start of a comedy: indeed, many plays do not include an argumentum at all. (This is formula-based romantic comedy, after all, not rocket science.) It is amusing to pretend that comedy is intellectually taxing stuff, and it is amusing to play around self-consciously with generic conventions. No doubt many of the real-life spectators will have enjoyed listening to the imaginary audience being patronized and treated like idiots, since by contrast these remarks implicitly flatter their own intelligence.

(vii) Scripted audience ‘responses’. On several occasions characters pretend to see the audience reacting to what is going on in the plot, either in the form of collective responses or in the form of individual spectators’ actions or words.^[33] For instance, the prologue-speaker in Plautus’ Captivi (7–14) breaks off mid-way through a speech in order to address imaginary spectators and listen to their replies:

seni huic fuerunt filii nati duo;

alterum quadrimum puerum seruos surpuit

eumque hinc profugiens uendidit in Alide

patri huiusce <hominis>. iam hoc tenetis? optume est.

negat hercle illic ultumus. accedito.

si non ubi sedeas locus est, est ubi ambules,

quando histrionem cogis mendicarier.

ego me tua causa, ne erres, non rupturus sum.

This old man had two sons, one of whom was abducted by a slave when he was a four-year-old lad; the slave escaped from these parts and sold him in Elis to this chap’s father… Do you get it so far? [Pretends to see an affirmative reaction] That’s excellent! [Pretends to see another response] … ah, that man over there in the back row says no. [Addressing the man] Come to the front, then. If there isn’t a free seat for you, then you can be up on your feet – there’s enough room for that. You’re forcing me to beg when I should be acting, but – make no mistake – I’m not going to bust a gut for your sake!

As in the case of the examples just discussed, these lines are based on the notion that the audience is preternaturally slow to grasp what is going on (the parenthetic ne erres in line 14 is a nice extra touch). In this case the prologue-speaker interrupts his own scene-setting narrative and pretends to check with the audience whether they are following. The performance must incorporate a couple of short pauses (after iam hoc tenetis? and again after optume est) during which the actor pretends to see or hear a collective affirmation followed by a response from a single spectator at the back of the cavea. It is obvious that neither of these responses is real: they belong to the fictional play-world. But (crucially) they are there in the script, and built into the performance, regardless of how anyone in the real world may have reacted at this point. Even if any real-life spectators did shout out audible reactions or comments, they would have to be ignored completely. The same is true of a similar passage in Plautus’ Casina (1–7), evidently written for a reperformance of that play, in which the prologue-speaker asks for a definite indication (signum clarum) whether the audience prefers new comedies or revivals of old ones.^[34] As in the Captivi prologue, the actor must pause and pretend to see or hear some sort of response; he then continues on the basis that the signum clarum has been forthcoming and that the audience have positively indicated their approval for old plays. What would happen if the real audience booed or hissed at this point? It would be too late, because the script has already been written, and Casina is the ‘old’ play they are going to get, like it or not.

(viii) The audience as speaking character. On rare occasions the imaginary audience might even be given a few lines in the script, albeit only in the form of direct speech quoted in the mouth of one of the other characters. Like all the other lines in the script, these utterances are in verse, so it is obvious that they are not supposed to be seen as realistic interjections or interruptions to the ‘illusion’; they merely underline the fact that the audience is a character in the play-world alongside all the other dramatis personae. As in other passages of oratio recta within comic monologues, the actor would almost certainly have changed his voice when directly quoting the words of others.^[35] In other words, the performer is required to ventriloquize or impersonate the audience at these moments by putting on some sort of silly voice or mannerism.

nunc si quis est qui hoc dicet aut sic cogitet –

“vetus si poeta non lacessit prior,

nullum invenire prologum posset nouos

quem diceret, nisi haberet cui male diceret,”

is sibi responsum hoc habet …

Now if there is anyone who is saying or thinking something like this – ‘if the old poet hadn’t provoked him first, the young one wouldn’t have had anything to put in his prologue, without having anyone to insult’ – then here’s a response for him …

In these lines, from the prologue of Phormio (12–15), Terence pretends to be able to guess exactly how the audience is reacting to what they are watching. Their words – or even their unspoken thoughts – are imagined as being formulated in iambic senarii. As elsewhere in Terence’s work, the underlying idea here is that the audience, along with the ‘old poet’ (Luscius Lanuvinus), is predisposed to be hostile to Terence and to pick holes in his work. A comparable moment occurs in another prologue when Terence scripts a single line and puts it in the mouth of Luscius (Eun. 15). It is notable that Terence imagines his rival Luscius and other unnamed enemies as being present among the audience watching his plays, for all that they (supposedly) loathe and vilify these works. We do not know whether or not this was the case, but if Luscius really was sitting in the theatre he would no doubt have found it disconcerting to be singled out and misquoted in this way.^[36]

Plautus too gives his imaginary audience some lines to ‘speak’ at Casina 67–78, where the fiction that the audience is in a position to interact with the characters within the plot-world is taken a stage further:

sunt hic inter se quos nunc credo dicere:

“quaeso hercle, quid istuc est? seruiles nuptiae?

seruin uxorem ducent aut poscent sibi?

nouom attulerunt, quod fit nusquam gentium”.

at ego aio id fieri in Graecia et Carthagini,

et hic in nostra terra et in <terra> Apulia;

maioreque opere ibi seruiles nuptiae

quam liberales etiam curari solent;

id ni fit, mecum pignus si quis uolt dato

in urnam mulsi, Poenus dum iudex siet

uel Graecus adeo, uel mea causa Apulus.

quid nunc? nil agitis? sentio, nemo sitit.

There are some people here, I think, who are now talking among themselves, saying: ‘Hey, what’s that you’re saying? A slave wedding? Are slaves going to start getting hitched or proposing marriage? They’ve presented us with something strange, which doesn’t happen anywhere in the world!’ But I’m telling you: it does happen, in Greece and Carthage, and also here, in our country, and in Apulia, where slave weddings are celebrated with even more pomp than those of free citizens. Is there anyone who wants to bet me a jug of honeyed wine that this isn’t so? Let him place a bet – just so long as the referee is from Carthage or Greece, or Apulia for all I care! Well, then? Are you not taking me up on the bet? Oh, I get it: no one is thirsty.

These lines are a variation on a motif which we have already discussed, viz. the audience’s recurrent incredulity and bafflement. As elsewhere, the motif is introduced by Plautus as a way of drawing attention to the novelty of his play, as if the audience had never seen anything remotely resembling the plot of Casina. The fact that the prologue-speaker is prepared to lay a bet to prove that he is telling the truth is a way of humorously exaggerating the play’s originality, while the possibility that the audience might actually take up the bet is far-fetched because it ignores the boundary between the real world and the plot-world. Once again, it does not make the slightest bit of difference whether anyone in the real audience reacted to the challenge: the script demands that the actor/character should see and hear nothing.

(ix) Audience participation? The notion that people in the real world might genuinely affect the course of events in the plot-world is a literal impossibility. The two worlds are separate; the plot is fixed; the script has already been written and cannot be rewritten mid-performance. Nevertheless, as we have already noted, the comedians often like to pretend that their elaborately crafted literary comedies are being improvised before our eyes, and they occasionally have fun with the tantalizing possibility that the audience could influence the outcome of a play. For example, Mercury in the prologue to Amphitryo directly offers the audience a radical choice between watching a tragedy or a comedy, as if they had the power to alter not only the details of the play’s plot but even its entire genre.^[37] Demipho in Mercator, when plotting to buy a courtesan and haggling over the exact price, pretends to see a man in the audience intervening to drive up the amount by five minas.^[38] In Asinaria the audience (who are assumed, as so often, to consist of unhappily married misogynists) are invited to intervene to save Demaenetus from being punished for his adulterous exploits.^[39] The parasite Gelasimus in Stichus, attempting to raise money for food, proposes to auction off his jokes and starts taking bids from the audience.^[40] The final lines of Rudens incorporate a novel twist on the conventional call for applause, by imagining that the plot-world will continue to exist after the play has finished and that the real world can somehow permeate its boundaries. Daemones here invites the audience to join him for a drink – in sixteen years’ time! – provided that they applaud the performance as loudly as possible.^[41] In all these passages the underlying conceit is that the comedy is being made up on the spot during the performance, and thus can be instantaneously altered or adapted depending on the audience’s reactions in real time. But (like everything else in the script) it is make-believe.

Plautus occasionally toys with the theoretical possibility that the audience might enter the stage and interact fully with the other fictional characters. At Casina 949–951 Lysidamus, expecting to receive a thrashing at the hands of his wife, optimistically calls out to the audience for volunteers to take his place. Elsewhere Pseudolus, when wondering why all the young men do not club together and free the people from the wicked pimp Ballio (Pseud. 203–204), appears to be issuing a challenge to the audience to do exactly this. But in one extraordinary scene from Aulularia (716–721) the audience is described as participating in the main events of the plot. Here the old miser Euclio has mislaid a pot of gold, and is so desperate to retrieve it that he begs the audience to tell him what has happened to it, before accusing them of having stolen it themselves!

oro, optestor, sitis et hominem demonstretis, quis eam apstulerit.

quid ais tu? tibi credere certum est, nam esse bonum ex uoltu cognosco.

quid est? quid ridetis? noui omnis, scio fures esse hic compluris,

qui uestitu et creta occultant sese atque sedent quasi sint frugi.

hem, nemo habet horum? occidisti. dic igitur, quis habet? nescis?

heu me miserum …

[To the audience] I beg you, I implore you: help me find the man who’s stolen it by pointing him out. [Turning to an individual spectator] What do you say? I’m sure you’re trustworthy, for I can see from your face that you’re a good man. [To the whole audience] What is it? Why are you laughing? I know your sort: I can see that there are thieves a-plenty in this place, who conceal themselves beneath respectable raiment and sit there as if they were decent folk. [Pretending to hear an individual spectator saying something] Eh? What was that? None of them has it? I’m ruined! Go on, tell me – who has it? Don’t you know? Oh misery!

This final example marks the reductio ad absurdum of the motif, since it assumes that the audience could literally interact with the characters on stage. The notion that a physical object such as a pot of gold could be transferred from the plot-world to the real world, or that people in the real world could purloin property from fictional characters, is a ludicrous impossibility. What makes Euclio’s words here so remarkable is that he is delivering them as himself (i. e., as the fictional character Euclio, rather than as an actor stepping out of character or ‘breaking the illusion’ to address the audience). The ‘illusion’ is technically maintained, even though at this moment the dividing line between the plot-world and the real world might seem close to dissolving. In the context of the plot of Aulularia, what is happening at this moment underlines the extreme crisis that Euclio is facing: in his grief and confusion at the loss of the money that is his whole raison d’être, his little world seems to be literally collapsing around him. But of course all of this is funny rather than genuinely unsettling, because none of it is real. Euclio is addressing the fictional audience, not the real-life audience, and the whole exchange is taking place within the imaginary world of the play.

3 Conclusion

Scenes in which the characters acknowledge the audience, using deictic language to address them or describe them in detail, are a pervasive and highly distinctive feature of Roman comedy. In this article I have attempted to show why I believe that such passages cannot be treated as factual evidence for the audiences who actually went to watch the plays of Plautus and Terence. The scripted audience is a fictional construct. As I have tried to demonstrate, it is not depicted in a realistic or plausible manner: it functions as a stock character, with a broadly consistent ‘personality’ and a recurrent range of attributes. Of course this fictional audience is not portrayed in a completely consistent manner from play to play. Like any other stock character, it functions in different plays in slightly different ways, depending on the dramatic context, plot themes or humour in each case. It is also important to acknowledge that Plautus and Terence differ from one another in certain key respects (for example, in the fact that Terence does not acknowledge the audience at all except in his prologues, which are normally playing complex literary games of a type not seen in Plautus). Nevertheless, on the whole it seems to me that the audience constructed by each individual play is essentially a variation on a recurrent stereotype.

The crucial point to be stressed is that there are two audiences present in the theatre simultaneously during any performance of a Roman comedy: the real one and the imaginary one. The scripted audience may have some – or many – features in common with the group of real-life spectators watching a play on any given occasion, but it still needs to be treated as a separate entity. The humour is produced through the interaction (or contrast) that that the comedians contrive to create between the two groups. I hope to have shed some light on the way in which these scenes function in terms of their metatheatricality. The so-called ‘fourth wall’ or dividing line between make-believe and reality is not drawn, as one might assume, at the edge of the stage, somewhere between the actors and the spectators. Rather, the entire performance space, including the cavea, is transformed for the duration of the play into a space where real and imaginary elements collide. For the duration of the drama, the real-life spectators (along with the actors) are invited to adopt or reject the role that has been scripted for them by the playwright. They can choose to identify with the fictional audience, or they can laugh at the all-too-obvious differences between the fictional audience and themselves. It is the interplay and tension between the two audiences that gives rise to the theatrical effect – which might be amusing or unsettling, depending on individual tastes and temperaments.^[42]

A further reason why these scenes are so striking is that nothing quite like them is found in the Greek comedies that Plautus and Terence translated and adapted. Greek New Comedy generally maintains the ‘dramatic illusion’ in a more realistic and consistent fashion.^[43] The works of Menander and his contemporaries do feature a limited amount of audience address, but only of a very perfunctory sort.^[44] The way in which Plautus and Terence treat the audience marks a significant new development within the genre. What we see in their plays is much more extensive, detailed and elaborate than anything in Greek New Comedy; all the scenes which I have discussed have obviously been freely invented for their new context. In other words, the greatly enhanced ‘presence’ of the audience is something that is distinctively Roman about Roman comedy. This fact has not generally been recognized, but I suggest that it can be seen as one of the most important and visible ways in which the Roman dramatists have creatively transformed their Greek models.