‘ Greek to Latin and Back: Did Roman Theatre Change Greek Theatre? ’

The influence of Greek drama on the birth and development of Latin theatre is a well-studied phenomenon. Greek theatrical production is not only behind the creation of the first full play in Latin, as well as many successive tragedies and comedies in this language, but its influence, direct or indirect, also needs to be acknowledged in pieces from the so-called ‘minor’ theatrical genres that were performed in the Roman world. Local, Italic forms of entertainment (such as Atellan plays) also contributed to the Roman theatrical landscape, but it is likely that they had themselves developed and evolved through contact with dramatic activities taking place in Greek-speaking areas. Roman theatre emerged at a time and in places characterized by contacts between several cultures, and Latin drama never quite succeeded in becoming fully independent from its models. Early Roman tragedies and comedies were certainly no mere literal translations of Greek plays. Yet, Latin dramatic literature remained deeply indebted to Greek drama. This is not in itself surprising, since this is arguably the case for the major part of Latin literature, something which has led D. Feeney to call Latin literature ‘a Grecizing literature in the Latin language’. From 240 BC onwards, it took some time for Latin theatre to acquire a status of its own, and even then Greek theatre was still perceptible behind most Roman dramatic productions. In parallel, Greek drama never stopped to exist in itself even in the heart of the Roman empire. As G. Hutchinson has noted: ‘At the very time when Latin literature is sometimes supposed to achieve independence from Greek, Greek literary activity is acknowledged in the capital with more emphasis than

before.' 3 This is also true for dramatic activity. Indeed, despite the (false) idea that Greek and Latin literatures succeeded each other in two distinct chronological series, Greek literature stayed well alive even as Latin literature developed. 4 The division between Greek and Roman literature, as Hutchinson (2013, 12) remarks, 'seems obvious to us, because it is reinforced by our own academic structures. But Romans employed considerable artificiality to maintain it [. . .]'. Scholarship on Greek drama has recently begun to focus on its evolution beyond its flourishing in Classical Athens: Greek theatre did not end with Euripides for tragedy and Menander for comedy. 5 Even in Italy and in Rome itself, the political and cultural 'centre' of the Roman Republic and then Empire, Greek theatre continued to be composed (sometimes even by Romans themselves) and, more importantly, performed. 6 Further East, dramatic competitions and festivals continued well into the Roman Imperial period, offering occasions at which Greek drama, including new plays, was performed. 7 The question that will be discussed in this chapter, therefore, is less odd than it might sound: has Latin drama exerted any kind of influence on late Greek dramatic production? And if so, is there anything in this process or its results that should encourage us to redefine what we encompass under the label 'Ancient Greek theatre/drama'?
The 6 Two successive grants from the Swiss National Science Foundation have allowed me to study the question of the performances of Greek theatre in Rome and Roman Italy during the Republic and Early Imperial times. By carefully collecting the available evidence, both in literary and in documentary sources, for theatrical performances in Greek language in these contexts, it became possible to demonstrate that they were less rare than previously thought. Some preliminary results of this research are available in Paillard (2019a) and Paillard (2021 forthcoming). On Romans who composed poetry and plays in Greek, see Hutchinson (2013) 143-146. On the survival of Greek comedy in the Roman Empire, see Marshall/Hawkins (2016). 7 See Graf (2015) and (2016); Skotheim (2016).
understand more precisely what late Greek literary production may owe to Latin literature, in particular as regards late Greek theatrical production. 8 As mentioned, Greek drama was still alive and performed at least during the first centuries of the Roman Empire. If theatre, as an independent literary genre, seems to be almost entirely excluded from the Second Sophistic as noted by M.-H. Garelli (2007, 209 n. 1), this still does not mean that drama in Greek language was not composed and performed anymore during this period. 9 As such, it was theoretically exposed to the influence of what was composed and performed in Latin during the Imperial period and earlier. Therefore, if this is not the case for theatrical literature as opposed to other literary genres, discussing the reasons for such an absence of influence is still a worthwhile enterprise.
Yet, this project is not an easy one, and there are good reasons why scholars seem to have avoided a systematic analysis of this question. Among the first problems encountered when discussing the possible impact of Roman theatre on late Greek theatre (from the end of the 3 rd century BC to the 3 rd century AD) is the scarcity of extant Greek dramatic texts in this period. No Greek play dating to this time is known to us in its entirety: we must rely on fragments and indirect testimonies about texts that have not survived. In most cases it remains uncertain whether what we have was intended for performance at all. Besides, when we do possess texts (fragments) that can be identified as belonging to the dramatic genre (widely speaking) and that date to the post-third-century BC, it remains difficult to gauge whether they differ from Classical drama (e.g., in style, structure, or metre) because of a possible influence from Roman drama or by reason of the intrinsic evolution of Greek drama after the 5 th century BC. 10 After the 3 rd century BC, our understanding of the evolution of Greek theatrical production becomes patchy at best and makes it almost impossible to pinpoint 'Roman influences' in the meagre fragments we have.
Yet, the problem could be addressed from a range of new, indirect anglessome of which will be examined in the following attempt at opening a path that research has hitherto left underexplored. 11 Rather than providing answers, it will examine examples of elements where the influence of Roman theatrical production and tastes could have weighted into the composition of late Greek language drama. Some speculation will be unavoidable, but it is hoped that this attempt will show that late Greek theatre might not have developed in total independence from Roman theatrical practice, however closely the latter was originally dependent upon the former. New contexts and performative aspects, the mixing of theatrical and non-theatrical genres and the influence of new dramatic (widely speaking) genres, as well as the question of the pantomime, will be investigated as possible ways to tackle this problem. Finally, the question of the definition of 'Greek theatre' will be addressed in light of the elements discussed in this chapter.

Contexts and Performance
Greek drama staged during the Roman period, even when Classical plays were reperformed, was different from what 5 th -century Athenians would have seen. A performance of a tragedy in Classical Greece was different, on many levels, from what it was in Rome, Naples, or even Athens itself, in the first century AD. Centuries of evolution rendered the content, audience expectations in terms of literary genres, and the performative aspects of a dramatic spectacle different from what would have been experienced in the 5 th century BC.
Traces of this dichotomy between what 'Greek theatre' was in the Classical period and what it had become by Roman times can be found in the apparently incoherent attitude of Cicero towards Greek drama, which was no doubt representative of what many of the contemporary élite members of Roman society thought of Greek theatre. On the one hand, citing Classical Greek dramatic authors, reading their plays, studying them at school, trying to imitate them by writing tragic pieces in the Classical style, were all deemed respectful activities for members of the learned élite. On the other hand, contemporary performances of Greek drama were not seen in such a positive light, and Greek actors in particular were criticized openly by the same people who professed admiration for Classical Greek theatre. 12 Such an attitude may well stem from the idealization of the Classical past that manifested itself among some members of the Roman élite. 13 Yet, it remains likely, however biased our (mostly élite) sources might be, that the theatrical performances in Greek which Cicero could have seen in theatres in his time did not correspond with the idea he had (and we still have) of what was going on in 5 th -century Greek theatres.
In this section, we will discuss three aspects of Roman-era dramatic performances that might have influenced how late Greek theatre was composed and performed: the contexts, occasions, and spaces of performances, the blurring between reality and fiction, and the increased spectacularity of what Romans liked to see on stage. For each of these three aspects, it will be important to discuss whether a clear influence of Roman theatrical practice on Greek contemporary production can be pinpointed, or whether Greek theatre had simply followed its own evolution, which had begun much earlier than 240 BC. The two explanations are difficult to disentangle and not mutually exclusive.

Contexts
Classical theatre, too often reduced in our minds to its manifestation as full-scale performances of tragedies, comedies, and satyr plays delivered in a theatre, has been described as an 'institution'. Such performances were following the ' [. . .] conditions and conventions of ancient Greek open-air, communal, religious theatre'. 14 Although including performance in the definition of theatre was an important step away from considering ancient Greek theatre plays as mere texts, this definition excludes a large number of theatrical activities that took place outside of the strict framework of religious festivals where tragedies, comedies, and satyr plays were performed in a contest, in theatre buildings (whether permanent or temporary).
Indeed, there is convincing evidence that less formal theatrical performances took place, for example, in the margins of events that did not officially include theatre performances, such as the Olympics, or at other occasions. 15 There is no good reason to exclude such performances from what we consider to be 'Ancient Greek theatre'. As W. Slater (2007,45) puts it: 'We should not insist on too strict a dichotomy between familiar formal festival categories and the artists on the fringe.' Likewise, we should abandon the idea of an exclusive link between 'proper' Greek theatrical performances and theatre buildings: one can exist without the other. 16 This may well be especially true for Late Classical (4 th -century) and post-Classical times: with the spread of theatre as a marker of Greekness it is not unlikely that theatrical performances began to take place in an increasing number of less formal contexts and at more various occasions than what was the case in Athens in the fifth century. 17 In the Roman period, trends identified in the change of spaces and occasions for theatrical performances in the post-Classical times seem to develop further.
As the aim of this chapter is to discuss the possible influence of Roman/ Latin theatre on late Greek dramatic production, it focuses on Rome and Roman Italy, as the first centres where Latin theatre emerged. 18 Occasions While many theatrical events still took place within the framework of regular public religious festivals comprising dramatic competitions, one sees a proliferation of other theatrical or theatre-like events taking place at other occasions in the Roman world. Following a trend that had already begun earlier, 19 victories, funerals, or other wide-scale public events were considered good opportunities to stage theatrical performances. In such contexts, theatre in Greek language was 17 Suffice it to note that Alexander organized dramatic performances sometimes far away from places that may have offered conventional formal spaces and contexts for such activities. On this topic, see Le Guen (2014). On the variety of contexts and occasions at which fifth-and fourth-century autocrats organized theatrical performances, see Csapo/Wilson (2021 forthcoming). On the spread of Greek tragedy after the Classical period, see Duncan/Liapis (2018) and Le Guen (2018), both with further references. 18 It will not be possible here to survey all theatrical events that included performances in Greek and might have taken place in Rome and its surroundings from 240 BC: only some examples will be discussed. 19 See Csapo/Wilson (2021 forthcoming) for the idea that, already from the fifth century, autocrats tended to use theatre more as a way to celebrate their own life milestones rather than keeping performances as general religious events. sometimes performed in Rome and Roman Italy. The delicate question of whether such occasions were still thought of as religious, thus keeping theatrical performances within a framework not dissimilar from what was known of the Classical period, remains difficult to answer. 20 It could be argued that many of those apparently non-religious occasions retained a fundamentally religious character. Indeed, entertainment organized on the occasion of a military victory, for example, could be understood as a way of thanking the gods for bringing military success to the organizer. 21 Livy (39.22)  About the time that these reports were brought from Spain, the Taurian Games were performed for two days for religious reasons. Then for ten days, with great magnificence, Marcus Fulvius gave the games which he had vowed during the Aetolian war. Many actors too came from Greece to do him honour. Also a contest of athletes was then for the first time made a spectacle for the Romans and a hunt of lions and panthers was given, and the games, in number and variety, were celebrated in a manner almost like that of the present time. 22 The distinction between the clearly religious nature of the Taurian Games and the games subsequently organized by Marcus Fulvius Nobilior should not detract from the fact that a religious component is also present in the latter (see voverat). (Whether dramatic activities took place at this occasion is less easy to deduce from the text itself. The Greek artifices mentioned by Livy need not necessarily have been actors, and nothing clearly indicates that they performed at all.) What remains likely, however, is that the religious character of the contexts in which drama was performed became less and less important in the Roman world. Theatre performance, an activity closely linked to the cult of Dionysos in the Greek world, was increasingly perceived as being comparable to other types of popular mass entertainment in the Roman world. Whether Terence's complaint (The Mother-In-Law, 25-40) that he must compete with boxers for the interest of the Roman audience reflects a historical event or not, his remark still indicates that the success of theatre as mass entertainment was becoming more fragile in the Roman world. Theatre was in direct competition with other types of (more spectacular) entertainment. The religious component of performing or attending theatre was not felt as important enough to draw the exclusive interest of the audience. As A. Petrides (2010, 92) notes: 'Make no mistake: panem et circenses was a phenomenon mainly of Rome and Greece under Rome, and chiefly a development of the Imperial period rather than the Hellenistic period.' Even for theatre, the 'entertainment' component took precedence over its religious component. It is therefore likely that the contexts for performance of Greek theatre had to be adapted to the new Roman framework: beside major religious festivals, theatrical performance in Greek must have taken place at occasions that were felt to be less religious, or at least less directly and closely linked to the cult of Dionysos in particular (despite the close links between Dionysiac cult and the associations of technitai). 23

Spaces
Theatre was also often played outside of theatre buildings or areas (temporary or not) specifically designated for this type of entertainment. While, as mentioned above, theatre could in some ways also be performed outside of theatre buildings in Classical Greece, the variety of places in which it was performed in the Roman world seems to have been larger, and the spread of the habit of building stone theatres throughout the Roman Empire did nothing to stop theatre being performed outside of them.
For centuries after the 'invention' of Roman drama, Rome did not possess any permanent stone theatre building, not even a designated place regularly reserved for theatrical uses. In consequence, what we consider to be truly Roman/ 23 One cannot exclude that the perception of dramatic performances as religious events or not could have been dissimilar for the people actively involved in the event (guilds, performers, theatrical practitioners, organizers) and for the audience. The links between Roman Emperors, ruler-cults, and theatre, consolidated through the associations of technitai, could bring another element to this question. It would be interesting to know whether spectators actually perceived theatre performances as a religious occasion within this framework. Christian authors, at least, regularly insist on the fact that theatrical activities are intrinsically linked to the cult of pagan gods (see, e.g., Garton [1972] 37). However, it remains difficult to tell, for example, whether the presence of statues representing members of the imperial family in some theatres encouraged the audience to feel like they were indeed attending a religious event in honour of a divine being, the Emperor and his family replacing Dionysos. (The same type of statues, however, was also found in non-theatrical and non-religious buildings.) On the relationship between autocrats and theatre in general, see Csapo et al. (forthcoming 2021).
Latin dramatic plays, such as the comedies of Plautus or Terence, were not composed with a performance in a big theatre in mind. Early Roman comedies were performed in smaller-scale settings (e.g., on temporary structures on the Forum Romanum, or on the Palatine outside the temple of Magna Mater), thus also reinforcing a feeling of closeness to their audiences. 24 During the Republic, it was this type of performance context that dominated the theatrical landscape of Rome. Plays could also be performed in temporary structures in the circus, for example (see Plautus' Miles gloriosus, 991), or near other temples, perhaps using their steps as seating spaces. The number of spectators would have been much smaller than what the theatre of Dionysos in 4 th -century Athens, for example, could accommodate. While larger-scale permanent theatre buildings began to be constructed in Rome and in the broader Roman world in greater numbers during Imperial times, the relatively long period of existence of a Roman drama which was intended for performance in different, small-scale contexts left traces in the definition of 'theatre' to the Roman mind: at least for some time, the emphasis moved away from large-scale tragedies and comedies, and what would have barely, until then, made its way into a 'Greek definition' of theatre, began truly to be considered theatrical. Minor genres such as mimes, performances of extracts, 25 more or less Romanized theatrical performances of Oscan dramatic forms, began to make their way into what was understood as 'theatre'. Such 'minor' productions were easier to perform in non-formal contexts and outside of a proper theatre building and might thus have made up a large part of what was performed during the early stages of Roman theatre production in Rome. As W. Slater (2007,45) remarks, this increasing importance of 'fringe-performers' could explain 'the movement away from formal drama to mime and pantomime'. Indeed, even the later Imperial pantomime could easily adapt to various performance settings and could range from wide-scale spectacles including large numbers of performers to much smaller, more intimate performances that could well take place in the streets or in private residences. Theatrical performances in private contexts also became increasingly frequent in the Roman world, although they were not unknown in the Greek world, especially since the times of Philip and Alexander the Great. 26 To come back to public performances, A. Duncan (2006, 215) speaks of a 'developing theatricalization of "ordinary" public space'. In the Roman world, dramatic performances indeed freed themselves from the formal space of the theatre and, conversely, spectacles happening outside of theatre buildings began to be thought of as 'theatre' as well. We know of at least two occasions at which theatre in Greek language was made part of this trend. Suetonius mentions that both Caesar and Augustus organized theatrical spectacles of some sort (ludi performed by histriones), which included spoken/sung parts (since the language of the actors is noted), regionatim urbe tota, i.e. in all the suburbs of the city of Rome. Presumably, such performances took place 'in the streets', so as to be available to everyone, perhaps on temporary structures (Suetonius, Life of Julius Caesar, 39.1): Edidit spectacula varii generis: munus gladiatorium, ludos etiam regionatim urbe tota et quidem per omnium linguarum histriones, item circenses athletas naumachiam.
He gave entertainments of diverse kinds: a combat of gladiators and also stage-plays in every ward all over the city, performed too by actors of all languages, as well as races in the circus, athletic contests, and a sham seafight. 27 The same anecdote is reported about Augustus in Suetonius' Life of Augusutus 43.1 (fecitque nonnumquam etiam vicatim ac pluribus scaenis per omnium linguarum histriones). This shows that drama is no longer limited to the formal space and context of the theatre, where spectators came to attend a performance, but now really pervades the whole social space of the city. In these instances, theatre in Greek (it would be surprising if Greek was not among the languages spoken by the performers) was made part of this Roman evolution. What would have been perceived in the Greek world as something not fitting a Classical definition of 'theatre', also no doubt began to be considered as such. The Roman interest in shorter, less complex, theatrical forms, must have encouraged at least some composers of pieces in Greek to adapt their art to these new contexts, spaces, and tastes. Instead of (or beside) composing new comedies and tragedies for formal contests or occasions, composers of Greek drama might have begun to create scripts for mimes, or, later, for pantomime libretti, as discussed below.

Performance: Blurring the Lines Between Reality and Fiction
Another aspect of Roman theatrical activities is the tendency to play with the borders between reality and fiction, and, for actors, between real identity and performed fictional character. This tendency became especially marked in Imperial times, as has been demonstrated by Duncan (2006, passim but esp. chap. 6). While in the Greek world, and still mainly in Roman Republican times, the gap between reality and fiction was as clearly maintained as possible and any uncontrolled transgression was perceived as dangerous for one's identity and, at a larger level, for society itself, in Imperial Rome, both performers and audience consciously began to create and to appreciate types of entertainments that blurred the distinction between self and other, between reality and fiction. In other words, while the Greco-Roman world seemed mainly to have followed platonic-like views on mimesis (Resp. 595a) until about the first century AD, Imperial times revelled in playing with the interaction between real and fictive worlds.
Absent from the Greek world, some types of mass entertainment became increasingly popular with the development of the Roman Empire. Such spectacles, as, for example, gladiatorial shows, venationes, reenactment of naval battles, were in direct competition with theatrical performances and indirectly influenced them. Indeed, to remain attractive to an audience now used to seeing blood shed before their eyes and other extreme forms of spectacle offered to their gaze, theatrical performances in Imperial Rome might have tended to include more 'real spectacle', more playing with the boundaries between reality and fiction, and less formal, well-defined acting according to a literary script as would have been the case in a (re)performance of a Greek tragedy, for example, or of a traditional Roman comedy.
To a mass of Roman (and non-Roman) spectators used to seeing people being killed in the arena, the Greek tragic habit (not to call it a rule) of never showing a scene of killing on stage, but rather of having a messenger narrate it, must have appeared almost ridiculous. Unfortunately, we remain ill-informed about the staging and performative aspects of plays in Early Imperial Rome. Yet, some evidence that theatrical entertainment began to adapt to the new tastes of its audience, an audience used to witnessing violence and death onstage, may have survived.
Seneca's tragedies would be good places to look for theatrical performances that included the spectacle of real violence (or even perhaps death) on stage, but of course we still do not know whether these were actually performed or not. However, as Duncan (2006, 199) points out, 'even if Seneca's tragedies were never staged, they speak to their historical moment's emphasis on looking at violence'. Two other phenomena (also examined by Duncan 2006, 200-203) seem to lead to the same conclusion: the staging of executions as if they were theatrical performances (e.g., Strabo 6.2.6-7), and the fact that real executions were sometimes embedded within what was defined as a theatrical performance (e.g., Suet. Life of Caligula 67). Such spectacles clearly blurred the borders between reality and fiction in a way that was entirely new.
Conversely, the Early Imperial trend towards a general theatricalization of real life also contributed to this blurring between the world depicted on stage and reality. 28 Theatre was real life as much as real life was theatre. Suffice it to recall here Augustus' alleged question at the end of his life, recorded by Suetonius (Suet. Life of Augustus 99.1), when he asked whether it looked like he had correctly played 'mimum vitae'. 29 This absence of real distinction between being oneself and performing as someone else was no longer fear-inducing, but became a part of theatre and of society itself in the Imperial period.
At a smaller level, this phenomenon also had implications for the performers. Once the anxiety about actors' identity vs their ability to impersonate someone else relaxed itself to turn into a source of fascination, it opened the door for new performance practices which went hand in hand with a new type of questioning of actors' identities. As actors allowed themselves to play on stage between their real identities and the characters they impersonated, the audience itself began in some cases to get confused about the true identities of the performers: was this female mime dancing lasciviously on stage actually a prostitute in real life, rather than an actress impersonating one? 30 Spectators' confusion can be forgiven when one considers some of the most striking examples of performers consciously playing on the distinction between real and fictive identity on stage. In the Greek (at least Classical) world, it seems that it was always clear to the audience, even in the most 'metatheatrical' situation, whether the performers spoke/ sung as themselves or as an enacted character. For Greek language theatre performed in the Roman Imperial period, this might have changed.
The example of pantomime dancers (on which see below), who intentionally changed masks on stage in view of the audience, thus embodying again for a while their 'real' identities as they were still in acting situations, is an indication that Roman audiences were not always exposed to a clear distinction between self and other in the case of theatrical performers. Two anecdotes about pantomime performers also show that the border between reality and fiction was so blurred that it became unclear whether a dancer had not actually become 28 This theatricalization of public life was not a new invention of Imperial Rome. It can be traced back, in certain contexts, at least to the Hellenistic period, as Chaniotis (1997 and2009) has demonstrated. The way in, and the degree to, which such theatricalization interfered with theatrical practice, however, seem to have taken another level altogether in Imperial Rome. 29 Whether or not this anecdote and the words themselves should be taken as reflecting a historical fact, they nonetheless constitute a compelling testimony of the kind of shift that took place at the time in the perception of the borders between reality and fiction. On 'theatrical deaths' in Imperial times and their philosophical implications, see Edwards (2002). 30 See Duncan, this volume, on the question of the identity of female mime dancers. the character being impersonated. Lucian (On Dance 82) reports that a pantomime dancer who was performing the role of Ajax became so involved in the character he was playing that it became unclear whether he had actually become mad. A similar anecdote is told about Pylades by Macrobius (Sat. 2.7.16). The question is not only a matter of exaggerated performance, whether these anecdotes be historical or not, but witnesses a new way of performing theatre, where the pleasure of the audience is not found in the viewing of a fictive world but in the blurring of the border between reality and fiction. Likewise, Suetonius (Life of Nero 21.3) mentions that Nero performed in tragedies wearing a mask representing himself. Nothing could spell out more clearly Roman Imperial fascination with the interplay between real identity and fictive, theatrical(ized) character.
This leads us to the question of the importance of masks for the definition of Greek theatre. 31 When only looking at what we know of formal performances of Greek drama in the Classical period, it would be fair to think that the use of masks should be taken as one of the defining criteria of 'Greek theatre'. 32 Masks seem indeed to be intrinsically linked to another of these possible defining criteria, i.e. the fact that the actor 'becomes someone else' when performing. As Rodríguez Adrados (1975,(45)(46) summarizes: 'The mask signifies that there is no mere representation but that the actor 'is' somebody different, outside the frontiers of time and space, outside the limits of the human, and of course of the individual.' In the Classical Greek world, there is indeed evidence that the idea of acting, in a theatrical or theatrical-like context or activity, without a mask looked disturbing. 33 This fundamental link between mask and theatre seems, however, to have become much less important in Roman times, as we have seen above. With the proliferation of theatrical performances that either took place without the performers being masked (as in mimes) or which consciously used the mask in a way that allowed the actors to play with the border between reality and fiction, between identity and otherness, this mask, instead of a marker of this border, became a tool to transgress it. As such, it could not be taken anymore as a secure element for establishing a definition of 'theatre'. 34 31 I am grateful to Oliver Taplin for a very stimulating discussion on this point. See Taplin (2018) on the fundamental role of the mask for the invention of theatre. See also Goette (2020). 32 For explanations of the use of mask in ancient drama, as well as further references on the topic, see the beginning of Duncan's chapter and Taplin's chapter in this volume. 33 Dem. 19.281. 34 This change in the importance and use (or non-use) of the mask might have been somehow linked to the loosening of the relation between Dionysos and theatre, but this topic would require another discussion altogether.
Greek to Latin and Back: Did Roman Theatre Change Greek Theatre?
Did such a change have any influence on Greek theatrical production of the time? There is indeed good evidence that some theatrical performances in Greek language took place without the actors being masked, and this type of less formal theatrical activity might have become increasingly frequent with the success it encountered in the Roman world. Some inscriptions relating to mime actresses suggest that they might have performed in pieces containing words spoken or sung in Greek. 35 If a mime actress performed in Greek but without wearing a mask, should we exclude her performance from the definition of 'Greek theatre' only because of the absence of a mask? 36 It would have been clear (despite the possible ambiguity discussed above) that the actress was not performing as herself (contrary to a simple dancer, for example, or an acrobat) but as the character of the story she impersonated, in spite of the absence of a mask. I would therefore suggest that, instead of the mask itself, it is rather the mechanism of impersonation, the enacting of someone else, that should lie at the core of a definition of Greek theatre. (The linguistic component of the performance is also obviously crucial. It allows us to define what is Greek theatre as opposed to, for example, Latin theatre. It is also decisive in allowing us to leave aside, among other types of entertainments, staged gladiatorial shows where the performers could be perceived as embodying a persona rather than fighting as themselves.) 37

Mixing and Innovation
One of the most defining features of Roman/Latin theatre is its seemingly constant mixing of different theatrical traditions (Greek, local Italic) to create new (more or less literary) theatrical genres. 38 Borders between those genres also became more porous than might have been the case for Classical Greek theatre, 35 See, e.g., CIL 6.10096, also quoted (with translation) by Duncan, this volume. On female mime performers, see Panayotakis (2006). 36 On mime, see Duncan's chapter in this volume. 37 In some extreme cases, even this mechanism of impersonation could have been played upon by the performer. The discussion below about the various ways in which the word tragoidos could be interpreted hints at cases where it might not have been clear to the audience whether the performer was on stage as himself (e.g., as a virtuoso singer of tragic passages) or as embodying a fictive character (i.e., with a clear dimension of acting present in the performance). 38 This tendency to mix literary genre might, however, not be entirely new, as we already see this phenomenon emerge in Hellenistic times, for example with pieces mixing satyr plays and bucolic traditions: see Krumeich et al. (1999) 11. More generally on the contamination between literary genres in the Hellenistic period, see Fantuzzi (1980). where it is usually easy to classify plays as either tragedies, comedies, or satyr dramas. We know of at least one example of a famous Roman actor, Roscius, who performed in both tragedies and comedies, thus indicating that the two genres were not hermetically separated. Authors, and perhaps actors (although we do not know whether Roscius' case was an exception rather than the rule), seemed not to be exclusively specialized in only one genre anymore. Given the restricted and fragmentary nature of the theatrical texts in Greek that remain from the relevant period, it is almost impossible to tell whether such a phenomenon might have influenced Greek theatrical production of this time at a literary level.
Lucius Anicius, who had been Roman praetor, upon conquering the Illyrians and bringing back as his prisoners Genthius, the king of Illyria, and his children, in celebrating games in honor of his victory, behaved in the most absurd manner, as Polybius tells us in his Thirtieth Book. For having sent for the most celebrated scenic artists from Greece and constructed an enormous stage in the circus, he first brought on all the pipers at once. These were Theodorus of Boeotia, Theopompus, Hermippus and Lysimachus, who were then at the height of their fame. Stationing them on the proscenium, he ordered them to accompany the chorus in unison with their piping. When they went through their performance with the proper rhythmic movements, he sent to them to say they were not playing well and ordered them to show more competitive spirit. They were at a loss to know what he meant, when one of the lictors explained that they should turn and go for each other and make a sort of fight of it. The players soon understood, and having got an order that suited their own appetite for license, made a mighty confusion. Making the central groups of dancers face those on the outside, the pipers blowing loud in unintelligible discord and sounding their pipes discordantly, advanced toward each other in turn, and the dancers, clapping their hands and mounting the stage all together, attacked the adverse party and then faced about and retreated in their turn. And when one of the dancers girt up his robes on the spur of the moment, and turning round lifted up his hands in boxing attitude against the piper who was advancing toward him, there was tremendous applause and cheering on the part of the spectators. And while they were thus engaged in a pitched battle, two dancers with musicians were introduced into the orchestra and four prizefighters mounted the stage accompanied by buglers and clarion players and with all these men struggling together the scene was indescribable. As for the tragic actors Polybius says, 'If I tried to describe them some people would think I was making fun of my readers. ' 39 In this detailed description, theatrical practitioners such as aulos-players, chorus members, and tragoidoi (on which, see below), are asked by the organizer to perform alongside other types of performers/athletes who do not usually appear in theatrical contexts. Moreover, the theatre specialists are at some point required to perform in a very different way from what they are used to and to transform themselves into performers of non-theatrical activities. We will never know what the tragoidoi were actually doing, nor what they were expected to do in such a spectacle, but it is likely that the same was true for them: they were not behaving as tragoidoi usually do. The spectators react positively to this performance, thus confirming Roman tastes for types of entertainment that was different from traditional drama. The much-discussed possible reasons behind Lucius Anicius Gallus' decision of inviting Greek artists and asking them to perform such a spectacle need not concern us here, nor does Polybius' perception of the event. 40 What matters for the purposes of the present chapter is to point out that the 39 Transl. Paton 2012, p. 139-141. 40 On these questions, see Paillard (2021 forthcoming), with further references. organizer of the event had invited the best Greek scenic performers only to deliberately ask them (before the performance, if one follows Franko [2013]) to deliver a spectacle that was more adapted to Roman tastes than what they were no doubt used to do in Greek settings. 41 The Greek performers accepted the invitation and the request to perform in a spectacle that was certainly not in their usual repertoire. 42 We might therefore have here an early example of Greek theatrical practice that actively evolved to adapt to Roman tastes. Greek theatrical artists adapted to a new way of 'doing theatre' suggested by their Roman audience's tastes.
The end of this passage leads to the question of the meaning of tragoidoi. It is notably difficult, and not only in this passage, to decide what exactly was designated by this word: playwrights who composed tragedies? Actors specializing in tragic roles? Singers of extracts of tragedies? People who recited tragedies with a more or less pronounced degree of acting? The problem is best illustrated by what M. Griffith (2018, 216) says about Nero's many ways of 'performing theatre': the Emperor is known to have performed as a singer of pieces accompanied by cithara (citharoidia), to have composed tragedies, and also acted in them. Griffith concludes: '[. . .] such a range of talents and contexts seems to have been somewhat typical of a tragoidos (however we choose to translate this term)and presumably Nero could and did perform some of his pieces in Greek'. Indeed, translating tragoidos remains a problem, and I would suggest that the marked hesitation about the exact meaning of the term has something to do with the Roman interest in types of entertainment that were previously outside of the margins of what could be described as 'theatrical'. This evolution should encourage us to modify the boundaries of what could be encompassed under the term 'theatre'. If the vocabulary employed in Roman contexts does not clearly distinguish between someone who recites an extract of a tragedy without acting and someone who really impersonates a fictional tragic character on stage, this could mean that more types of entertainments had made their way into the category of 'theatre'. 43 The degree to which a reciter accompanied his recitation or his song by some form of acting (gestures, uses of props, tone of the voice) could vary. 44 While 41 Again, whether it was planned from the outset or not does not matter here. 42 For the reasons why these Greek artists might have accepted the invitation, see Günther (2002) 131-132. 43 In formal contexts such as festivals, categories for citharoidia and actors of tragedy remained clearly differentiated, however. 44 On the relationship between recitations and fully acted dramatic performances (mainly in the early Empire) and on the fact that the former could include elements rendering them 'quasi-dramatic', see Bexley (2015).
'semi-theatrical' or 'para-theatrical' performances were not unknown in the Greek world, it is the success they encountered with Roman audiences that changed their status. From a type of entertainment that barely merited the title of 'theatre' compared to traditional tragedies or comedies, they acquired an importance at least comparable to that of formal traditional comedies and tragedies, which warranted their inclusion inside the margins of what was considered as theatre.
These points are important to keep in mind when focusing on the definition of Greek theatre: Greek practitioners quickly adapted to the new contexts and opportunities opened by Roman tastes and practices. Even if it remains difficult or impossible to gauge whether a Roman spectator would think of a somewhat theatricalized recitation (sung or not) on stage of an extract of a Greek tragedy as 'Greek theatre', it would be too reductive to state a definition of the term that excluded a priori all performances of this type. The 'impersonation' aspect may be what remains at the centre of our proposed definition, but even this seems to have been up for alteration in the Roman world.

The Case of Pantomime
Pantomime is often considered to be the Roman theatrical entertainment par excellence. Its immense success began in the early Empire and lasted for six centuries. 45 Its specific form (dancer with closed-mouth mask, accompanied by singer [s] and musician[s]) seems to have been 'invented' in a context that blended Greek, Italic, and Roman theatrical and dance cultures. Evidence suggests that the first (or at least some of the first) pantomime libretti were composed in Greek, and some of the most famous early pantomime dancers seem to have come from a Greek-speaking background, in particular Pylades (sometimes credited with the invention of pantomime). 46 One of the best pieces of evidence for the fact that early pantomime libretti were composed in Greek is found in Macrobius' Saturnalia (2.7.12-13): 12. Sed quia semel ingressus sum scaenam loquendo, nec Pylades histrio nobis omittendus est, qui clarus in opere suo fuit temporibus Augusti et Hylam discipulum usque ad aequalitatis contentionem eruditione provexit.
12. But having once mounted the stage in giving my account, I shouldn't omit the actor Pylades, who flourished in the Augustan age and brought his follower Hylas so far along in his art that they competed on an equal footing. 13. The people then divided its support between the two, and when Hylas was dancing to some song whose concluding rhythm contained the phrase 'the great Agamemnon,' Hylas measured out his stature to give the impression of someone massive and grand: Pylades lost his patience and called out from the audience, 'You're making him big, not great!' 47 Sadly, this τὸν μέγαν Ἀγαμέμνονα is one of the very rare fragments 48 of these early pantomime libretti, some of which were probably soon also composed in Latin. There is good evidence that in the 1 st century AD, pantomime libretti were sometimes composed by well-established Roman poets. 49 Later, in the 2 nd century AD, pantomime became, along with other traditional theatrical genres, part of the contests organized at the occasion of sacred festivals and games in Greece. It would therefore be very surprising if Greek theatrical composers did not, in turn, create new libretti in Greek for pantomimes, no matter what kind of opposition this genre had initially faced in the Greek regions. The famous pantomime Apolaustus, who won an impressive number of victories at the end of the second century AD all around the Eastern Mediterranean and in Greece itself, never seems to have performed in Italy. 50 This suggests that he was dancing to accompanying libretti composed in Greek rather than in Latin, although one cannot affirm this with absolute certainty. Unfortunately, there is little evidence to confirm that the number of pantomime libretti composed in Greek increased when categories for such performers were created in Games taking place in Greece. As mentioned above, almost no written trace of pantomime libretti survives. The best candidate for a (probably late) Latin pantomime libretto may be the so-called 'Barcelona Alcestis', if one follows Hall's (2008) convincing proposal.
Yet, possible indications that the success of pantomime performances in the Roman world had influenced in turn Greek dramatic practice might come from more indirect sources. Quintilian's (11.3.178-180) description of the way Demetrius and Stratocles acted on stage emphasizes the former's exaggerated gestures, poses, his way of playing with the wind in his clothes, and the latter's 'out of character' laugh addressed to the audience from behind the mask. He also notes that both were able to impersonate many very different characters. These performing abilities are not dissimilar from the way we would describe the skills of a pantomime performer. While it is true that Greek actors' performance style became increasingly spectacular over time, would it be completely out of the question to suppose that the success of pantomime, from the beginning of the Imperial period, encouraged Greek theatre performers to go even further in their attempts at providing spectators with a more visually stunning, spectacular performance? The success of pantomime could have influenced the composition of Greek tragedies at a more textual level, too. A. Zanobi (2008 and2014), expanding on an idea expressed by B. Zimmermann (1990[Engl. transl. Zimmermann 2008, has highlighted the possible influence of pantomime on the composition of Seneca's tragedies. I can only agree with Webb (2018, 304) when she states: 'It would not be surprising if this were true also of the Greek tragedies composed for competitions in the second and third centuries AD, when pantomime was becoming more and more widespread in the East; unfortunately, the lack of evidence for the nature of these plays makes it impossible to know with any certainty whether this was the case.'

Synthesis and Conclusion
After this survey of three angles (contexts, mixing and innovation, and pantomime) from which the question of the influence of Latin theatre on late Greek theatrical production might be addressed, it will be useful to summarize some conclusions and to come back to the definition of 'ancient Greek theatre'.
A direct influence of Latin dramatic literature could not be highlighted. A more detailed and exhaustive examination of all dramatic fragmentary remains (both Greek and Latin) might provide different results, but such a task would be beyond the scope of this preliminary attempt at providing a framework for discussing the question, and it remains uncertain whether the evidence available could lead to further conclusions. The problem of the small number of extant texts that can be classified as 'dramatic' from the period between 240 BC and the 2 nd century. AD, the rather short-lived history of Roman comedy and tragedy, as well as the fact that pantomime libretti were apparently thought less worthy of transmission, might explain why the influence of Roman theatrical genres on Greek (literary) theatre remains difficult to track. This is not the place to discuss the whole history of Latin literature, but it is worth remarking that 'formal theatrical genres' such as comedies and tragedies almost completely disappear between the 1 st century BC and the 1 st century AD. Besides, the spread of Latin drama outside of Rome and Italy is still in need of a comprehensive study, as noted above. If Roman theatre spread to the Greek East even in a limited way, such geographical contexts might have provided another place where Latin drama could have influenced Greek drama. However, from the time Rome began truly to expand in the Greek East, and even more clearly during the Early Empire, the use of Greek theatre (rather than drama in Latin) as a mass communication medium able to reach non-Latin-speaking populations and as a political tool was clearly understood by Roman élites and autocrats. 51 This might be one of the reasons which discouraged the production of more formal comedies and tragedies in Latin and which favoured genres for which linguistic/literary components were less important, or which used a language more widely understood. The influence of such dramatic production on Greek pieces (whether comedies or tragedies or other 'minor genres') thus becomes extremely hard to trace.
Yet there are indications that late Greek theatrical production could not entirely escape the preferences of what was quickly becoming the dominant political power around the Mediterranean. The causes of the success of pantomime are no doubt multifarious, but it is clear that it is the genre's success in Rome itself that caused its spread all around the Roman Empire. Its formal admission as a distinct category in the dramatic competitions of the Greek East must have encouraged Greek poets to compose libretti for entries in such prestigious contests. Dramatic poets and performers were not completely isolated from their socio-political and cultural contexts: if the demand of that audience was higher for pantomime than for full tragedies and comedies, they had to adapt somehow to the new tastes in order to survive. The same is likely for the development of other 'minor dramatic genres'. In other words, what might have influenced the composition of late Greek dramatic plays is less Latin theatre in itself than Roman tastes. If Rome's interest in theatre was more towards less complex/less literary forms of entertainment, more oral and visual productions, less welldefined genres, and a fascination for the blurring between reality and fiction, there are good chances that Greek poets and composers of theatrical pieces were encouraged to produce texts/supports for performances along those lines.
How does this relate to the possible need for a redefinition of what we understand by 'ancient Greek theatre'? The need for such a reflexion stemmed from the observation that various research projects devoted to ancient Greek theatre seem to use different definitions of 'Greek theatre'. Some only encompass what was performed in permanent theatre buildings, either including everything or excluding musical and other non-mimetic performances. Traditional literary genres divisions are followed by some, who understand 'Greek theatre' as a label only applying to formal tragedy, comedy, and satyr play, while others include minor mimetic genres or other forms of performances at the margins of the theatrical, wherever they are performed.
Recent research on the evolution of Greek theatre after its life in Athens in the 5 th century has shown that it would be artificial (not to say plainly wrong) to restrict the definition of 'Greek theatre' to include only Classical Athenian tragedies, comedies, and satyr play. Theatre in Greek language had many other manifestations. This becomes even clearer for later periods, during which additional types of performances in Greek language either appeared or acquired a visibility and a status that they had previously not had, possibly through the influence of theatrical production in Latin and certainly influenced by Roman tastes. In a rather paradoxical and unexpected way, Roman theatre thus changed Greek theatre. In the conclusion of his chapter on music and dance in post-Classical tragedy (in Liapis/Petrides 2018), after his survey of the many shapes that tragic performances could take in later periods, Griffith notes: 'Whether we choose nowadays to characterize all these post-Classical performances as "tragedy" is a delicate matter of definition. If we mean by "tragedy" a fully staged production of a complete play, or even a written play-text carefully read in school or in the privacy of one's home or library, then we obviously should exclude many of the phenomena that I have been discussing in this chapter. But if we expand our horizon to include all kinds of adaptations and musically enhanced dramatic scenes, performed by a skilled singer with or without a chorus, then the picture looks quite different.' 52 Indeed, the same question applies to Greek theatre more generally: if we want to be able to include in our definition of 'Greek theatre' the various forms that late theatrical performances could take, we must go beyond the idea that it was limited to fully formal tragedies, comedies, and satyr-plays.
What should remain, then, at the core of a definition of 'Greek theatre' so that it could encompass not only Classical but also later theatrical production?
The first and most straightforward element should probably be language. A performance without words, sung or spoken, even if only as an accompaniment as in the case of pantomime, can hardly qualify as theatrical, and if the language is not Greek, it should obviously not be considered to be Greek theatre.
Should the place of performance be taken as one of the criteria for what qualifies as Greek theatre and what should be excluded from its definition? If for the Classical period it might be safe to assume that what was really considered as theatre always took place in a theatre building, we have seen in this survey that this was not necessarily the case for later periods. Theatrical activities frequently took place in other venues, including spaces not specifically designated for dramatic performances.
Should we, then, revert to more literary criteria to establish a definition of 'Greek theatre'? With the proliferation of less-clearly defined types of theatrical performances, and less-evidently literary/written theatre, it would be difficult to argue that such a clear-cut definition should be used. In the Roman period, theatrical performances in Greek very often existed beyond the borders of welldefined literary genres.
Is the distinction between music, dance, and theatre any more valid? While Greek drama seems always to have encompassed music and dance, it is possible to draw a line between 'theatre' or 'drama' on the one side and purely nonmimetic dance and music (including song) on the other, where the performers do not take on the identities of the characters whom they depict. Likewise, acrobatic performances (even if they happen to be performed in a theatre) do not require that the acrobat perform as someone else: they are themselves, doing a specific (visual) performance.
Classical tragedy, comedy, satyr play, pantomime, and the later proliferation of 'minor' theatrical genres all have in common the fact that the performer 'becomes someone else'. I have suggested here that this dimension of impersonation is the smallest common denominator between Classical Greek theatre and what we see performed in Greek language in later periods. A mere recitation or reading in Greek of a tragedy of Euripides, for example, in a Roman theatre in Italy would not count as 'Greek theatre', if there is no dimension of 'performing the other'. The reciter recites as himself (even when pronouncing the words spoken by a fictive character in a dialogue), he does not become Electra. The actor 'becomes' Electra. The pantomime dancer, with his costumes and masks, becomes the character he impersonates. Citharoidia, on the other hand, if understood as the mere singing of, in at least some cases, dramatic pieces, is not theatrical in itself. However, while it might theoretically be easy to draw a line between performances where the performer performs as him/herself and cases where he/she impersonates a fictive character, the fact that it is in many cases difficult to understand what a 'tragoidos' was exactly thought to do in the Roman period indicated that even such a line might have become blurred in the Roman world. Was there in this period such a clear difference between a tragic actor (performing as Oedipus) and a singer who sung, without acting or with minimal acting, an extract of Oedipus' role in a given tragedy? As we have seen, playing between real and fictional identity became a favourite activity in Imperial times.
To conclude, this leads us again to the question of the mask. Was the mask, in Roman-time theatrical performances in Greek, still as fundamental as it seems to have been in Classical drama? In other words, should we exclude from 'Greek theatre' everything that was not performed with masks? As we have seen, the success of mime and of certain forms of tragic singing/recitation (tragoidia) might encourage us not to draw too tight a line around our definition of 'Greek theatre'. Mime performers no doubt impersonated 'someone else' although they were not wearing masks; conversely, pantomime dancers were often very close to appearing on stage as themselves rather than as fictive figures. It is possible that the wearing of a mask might have looked less and less important as the play between identity and fiction was increasingly perceived as enjoyable. As it seems to be a mark of Roman Imperial tastes, should we perhaps understand such types of performances (when in Greek) as being 'Roman theatre in Greek language', in the same way as Latin plays, at their beginning, could be qualified as 'Greek theatre in Latin language'?