At some point in the early second century CE Suetonius set out to compose biographies of important Roman literary figures. The largest surviving section of this work—known as the De Grammaticis et Rhetoribus—opens with what is generally considered an account of the early beginnings of philology in Rome:Footnote 1
grammatica Romae ne in usu quidem olim, nedum in honore ullo erat, rudi scilicet ac bellicosa etiam tum ciuitate necdum magno opere liberalibus disciplinis uacante. initium quoque eius mediocre exstitit, siquidem antiquissimi doctorum, qui idem et poetae et semigraeci erant—Liuium et Ennium dico, quos utraque lingua domi forisque docuisse adnotatum est—nihil amplius quam Graecos interpretabantur aut si quid ipsi Latine composuissent praelegebant. nam quod nonnulli tradunt duos libros—de litteris syllabisque, item de metris—ab eodem Ennio editos, iure arguit L. Cotta non poetae sed posterioris Enni esse cuius etiam de augurali disciplina uolumina ferantur.
primus igitur, quantum opinamur, studium grammaticae in urbem intulit Crates Mallotes, Aristarchi aequalis, qui missus ad senatum ab Attalo rege inter secundum ac tertium Punicum bellum sub ipsam Enni mortem, cum regione Palati prolapsus in cloacae foramen crus fregisset, per omne legationis simul et ualitudinis tempus plurimas acroasis subinde fecit assidueque disseruit, ac nostris exemplo fuit ad imitandum. hactenus tamen imitati, ut carmina parum adhuc diuulgata uel defunctorum amicorum uel si quorum aliorum probassent, diligentius retractarent ac legendo commentandoque etiam ceteris nota facerent, ut C. Octauius Lampadio Naeui Punicum bellum, quod uno uolumine et continenti scriptura expositum diuisit in septem libros…
(De Gramm. 1f.)The study of ‘grammar’ was not even pursued at Rome in the early days, still less held in any esteem; and naturally enough, since the state was then still uncouth and given to war, and had as yet little leisure for liberal pursuits. The beginnings of the subject, too, were humble, for the earliest teachers, who were also both poets and half-Greeks (I refer to Livius and Ennius, who gave instruction in both tongues at home and abroad, as is well known), did no more than interpret the Greeks or give readings from whatever they themselves had composed in the Latin language. For while some tell us that this same Ennius published a book On Letters and Syllables and another On Metres, Lucius Cotta is right in maintaining that these were not the work of the poet, but of a later Ennius, who is also the author of the volumes On the Science of Augury.
In my opinion then, the first to introduce the study of grammar into our city was Crates of Mallos, a contemporary of Aristarchus. He was sent to the senate by king Attalus between the second and third Punic wars, at about the time when Ennius died; and having fallen into the opening of a sewer in the Palatine quarter and broken his leg, he held numerous and frequent conferences during the whole time both of his embassy and of his convalescence, at which he constantly gave instruction, and thus set an example for our countrymen to imitate. Their imitation, however, was confined to a careful criticism of poems which had as yet but little circulation, either those of deceased friends or others that met with their approval, and to making them known to the public by reading and commenting on them. For example, Gaius Octavius Lampadio thus treated the Punic War of Naevius, which was originally written in a single volume without a break, but was divided by Lampadio into seven books…
(tr. Rolfe, adapted)As often happens with ancient narratives of beginning, Suetonius’ representation is more significant for the ideological framework that informs it than for its historical accuracy. For a start, Suetonius introduces the term grammatica without defining what it consisted of as a distinct discipline. This is because Suetonius is not interested in defining a field of inquiry; his attention is on individual figures and his approach is based upon the centrality that the grammarians occupied in the elite education and culture of his days. And in fact even in the opening passages the focus is on the progressive incorporation of this study into Rome's cultural life matched by the progressively higher standing of its practitioners.Footnote 2 Progressivism also shapes the characterization of Rome ‘before grammatica’ as uncouth and warlike, with its texture betraying topoi that can be traced back to Porcius Licinus in the late second century BCE and find their most famous manifestation in Horace's Ars Poetica. Footnote 3 Embroiled in Suetonius’ narrative are also alleged ancestral assessments of poetry as we read them in Cicero's Tusculan Disputations (sero igitur a nostris poetae uel cogniti uel recepti, ‘therefore poets were either known or welcomed late by our ancestors’, 1.3) and in a fragment attributed to Cato the Censor (poeticae artis honos non erat, ‘there was no honor in poetry’, Gell. 11.2.5). That Suetonius would be affected by existing accounts of poetry's early history makes perfect sense. After all, in his time grammatica—as Quintilian puts it—included teaching and ‘learning speaking properly and the interpretation of the poets’ (recte loquendi scientia et enarratio poetarum, 1.4.2). The relationship between ars grammatica and ars poetica as articulated by Quintilian would have been indisputable for Suetonius and so would have been their common history, by extension. Accordingly, it is not at all surprising to find Suetonius citing, however dismissively, Livius Andronicus and Ennius: widely recognized as some of the initiators of Roman poetry, they are here presented as pioneer grammarians as well. Thus, if, on the one hand, Suetonius refers to activities like reading and reviewing (praelegere, legere, retractare), commenting and expounding (commentari), analyzing and translating (interpretari), and reorganizing (diuidere), on the other hand he alludes to the formation of an environment within which poetic texts were being produced, circulated, and acted on.
James Zetzel has recently noted that Suetonius’ account is in part consistent with what we read in Quintilian: apart from practicing enarratio—the explanation of difficult passages based on corrected reading (emendata lectio), the early grammarians would have put a note on suspect lines almost in the manner of the censors, ranking/canonizing some authors, and excluding others altogether.Footnote 4 It would seem that textual criticism did not go much further than that. Zetzel points out that for the Republican period the evidence in support of the explicit comparison of different readings from different manuscripts is scanty. As such, it would appear that these early practitioners operated by relying upon their own judgement rather than on a systematic approach or archival research.Footnote 5
Although accurate, Zetzel's reconstruction is also somewhat discouraging. This is because in a desire to reclaim Roman philology, we tend to look at select evidence and to assess our findings against an understanding of it as a highly codified and compartmentalized branch of learning. The challenge here is to acknowledge that ‘philology’ is a highly elusive term; the practices that it invokes are varied but also very concrete.Footnote 6 Philology includes conservational activities like collating different manuscripts, noting errors and correcting them, but also the production of books in which textual materials derived from excerpting, glossing, paraphrasing, and so on were gathered. In this article I understand ‘philology’ as the totality of textual practices and approaches adopted in the conservation of exemplars and the making of books related to these exemplars; my focus is on how these practices and approaches were understood and redeployed in the construction of poetic texts in Latin.
This article takes its inspiration from Suetonius’ invocation of the earliest Roman poets in the opening of his De Grammaticis et Rhetoribus. Its goal is to identify the traces that these poets’ handling of texts and books has left in what survives and the effect that this handling had on the development of Roman poetic craftsmanship. I am especially interested in how these poets operated on Greek books and in how their textual sensibility manifested itself. These traces are difficult to recognize and the texts that bear them are sparse and scarce. In recent years much has been written about the Hellenistic libraries and much scholarship has been produced on the early development of Latin literature.Footnote 7 The risk here is to be stuck in between the greater amount of evidence that we have about the work that went on in places like Alexandria and the diverging opinions of those scholars who have paid attention to the formation of a literature in Latin in Rome. To top it off, the heavily translational nature of early Latin poetry poses a methodological conundrum. In past years intertextuality has been the primary model used in the study of (later) Latin poetry. By assigning primacy to the reader as meaning-making agent, intertextuality locates meaning in the relationship of a text with prior texts and in the generic codes which these texts employed. Based on the notion of discourse, intertextuality has opened up the possibility of exploring textual relations through language by deploying generic codes and models as regular and systematic meaning-making patterns.Footnote 8 This article is inevitably driven by an intertextual gaze, as it looks at more than one text at the same time. The difference is that I am not interested in generic codes or models as discourse but in teasing out their materiality and their history. My intertextual gaze is on the early Roman poets’ interventions on books, texts in books, and texts across books. In this article I focus on a sample of texts derived from the works of Livius Andronicus, Ennius, and Terence. Before engaging with them, I will briefly consider what we know about the traffic in Greek books in late third- and second-century BCE Rome and the sort of philological practices that went on in the Hellenistic libraries (with a focus on Alexandria).
Books in Late Third- and Early Second-century BCE Rome
In the late third and early second century BCE great quantities of books were stored in libraries located in the largest Hellenistic cities. During this period repositories of written materials existed in Rome and the rest of Italy as well. As Michael Affleck notes, in Rome temples and other public buildings fulfilled an important archival function for the preservation of texts associated with rituals and regulations, as well as other documents relating to the running of the city (including censorial materials, pontifical tables, and the like).Footnote 9 Although we have no substantial evidence as to where exactly these written materials were stored, one fact is clear: ‘Rome was literate centuries before it developed what we have come to call literature.’Footnote 10 Accordingly, the issue here is to locate in Rome the types of books handled by the early poets, namely and primarily, books that bore Greek literary texts. Again, we possess no secure evidence as to where and how the poets acquired these books; however, with Affleck I find it reasonable to suppose that from the early third century BCE onwards the Romans hauled to Rome book collections from Southern Italy as part of their booty. It is most plausible that these Greek books, once in Rome, were kept in private houses. In this respect, a case has been made for Livius Andronicus who enjoyed the patronage of the Livii and used the book collection that they most probably owned.Footnote 11 The technitai or craftsmen of Dionysus have been invoked as possible conveyers of Greek books too. The term designated writers, actors, and musicians with a tradition based on the reperformance of written texts that went back to the theater of Dionysus and fifth-century BCE Athens. Organized in guilds that did not fall under the jurisdiction of any rulers, they traveled to all parts of the Hellenistic world and negotiated on equal terms with the cities or the royal houses that contracted their services.Footnote 12 Although some have posited a direct influence of these guilds and their members on the development of Latin poetry (especially drama), references to their presence in the city remain scanty.Footnote 13 This fact, however, does not exclude that scripts could have been acquired from these guilds either in Rome or Southern Italy.Footnote 14 In treating the word scriba, Festus points out that the term used to designate public clerks (librarii) and poets (poetae) alike. He also adds that Livius Andronicus composed a hymn of expiation and that after the propitious progress of the Second Punic War, in 206 BCE the Senate proposed to found a collegium on the Aventine in the temple of Minerva.Footnote 15 It is feasible to think that such an establishment would have included a supply of Greek books. The same can be said about other temples built some years later. After his triumph in 187 BCE or during his censorship in 179 BCE, Marcus Fulvius Nobilior rededicated a group of sacred objects plundered from Ambracia in a new temple. They consisted of images of the nine Muses and the temple is known as the Aedes Herculis Musarum. Situated in the Campus Martius, it became the new meeting place for poets.Footnote 16 In the portico of this religious complex Fulvius had also taken care to display his fasti and the year-by-year list of consuls and censors.Footnote 17 It is clear that in Fulvius’ temple writing in its varied forms played a major role; its association with the Muses would call for the presence of Greek books as well. In his Histories, Polybius attacks the approach of Timaeus and, mockingly, adds: ‘you can make inquiries from books without dangers or sufferings, provided one takes care to have access to a town possessing many documents or with a library in the vicinity.’Footnote 18 Polybius does not mention any specific location, and it is possible that he was thinking of Greece; regardless, he appears to imply that libraries were somewhat common, or at least not extremely hard to come by. What we know for sure is that the same Polybius, at some point, took advantage of what must have made an impressive entrance in Rome: the library of Perseus that Aemilius Paullus shipped from Macedonia as part of his spoils and donated to his sons in 168 BCE.Footnote 19 Polybius stages his first encounters with the young members of the powerful Scipiones around this library: ‘I have already explained how this friendship developed from the borrowing of books and from discussions around them.’Footnote 20 I find this passage remarkable: we have people of different backgrounds, Greek books, and social relations forming around and through them. In 160 BCE Terence composed two plays for the funeral of Aemilius Paullus, the Hecyra and the Adelphoi, but only the latter made it on stage.Footnote 21 It is not out of place to imagine that the poet used Greek copies from the Scipiones’ library as well.
The few tesserae I just gathered do not even begin to resemble a satisfactory mosaic; however, they suggest a substantial traffic in books with features of its own. Nothing in them indicates the existence of the kind of book experts or organization of books that characterized the Hellenistic libraries. Contextual differences should not put us off; they should incite us to pursue the matter further. To do so, it is best to consider, first, how books were conceptualised and used in the Hellenistic libraries and, then, to explore how the practices developed there were redeployed in the Roman context. I argue that the early Roman poets acted as a conduit for the transfer of these practices and their works bear clear but idiosyncratic signs of some of them.
Philology and the Hellenistic Libraries
It is well known that the foundations of philology are to be found in the Hellenistic libraries around the beginnings of the third century BCE. The great literature inherited from the classical period was collected, preserved, and used in them at a scale never seen before. In practice, textual criticism was born from the whole range of activities that served to maintain and enrich these repositories of books; in principle, it evolved from the moment in which texts came to be understood as entities living a separate existence from the book: they could be copied from one medium to another in their entirety or in small bits and, therefore, be studied independently from their medium.Footnote 22 The dynamic configuration of material practices and conceptual developments that made up these libraries pushed readers to step into new territories. For a start, reading came to involve more than one book. It included finding relationships among their texts, comparing, criticizing, explaining, investigating, classifying. This expansion, along with the creation and coexistence of multiple copies, evolved into a particular cultural conception of the text which Jacob describes as ‘a fixed and unambiguous apparatus, resulting from an authorial intention and excluding the coexistence of variants’.Footnote 23 Such an understanding revolutionized the micro-dynamics of reading and affected the formation of new types of writing as well. Guided by scepticism and scrutiny, reading became an activity focused on the search for textual errors, corruptions, and flaws; writing, on the other hand, came to be used as a tool for emendation, to correct and explain the text.
Studies on the working methods adopted in the Alexandrian library suggest that from Zenodotus onwards scholars tended to choose an exemplar as the basis for their work. On that they annotated their preferred readings in free spaces or in between lines, and marked their interventions with an obelos, or symbol in the margins. This process of correction, generally labelled as diorthosis, was the defining activity of Alexandrian scholars and could entail a simple deletion of scribal errors, a systematic collation from different exemplars resulting in more radical corrections, or new conjectures.Footnote 24 These interventions occupied an interlinear or marginal position in the exemplar; in some cases they were accompanied by brief explanatory notes.Footnote 25 Needless to say, this way of handling texts was fundamentally conservative and depurative for the aim was to restore the ‘authentic’ form of the text and to produce a ‘model’ exemplar. In accordance with the Aristotelian principle of aprepeia (‘inappropriateness’), suspected corruptions were generally identified and rejected as unsuitable to a particular writer's style and usage.
The general outline that I have just presented brings to the fore the issue of authorship. In its simplest meaning authorship indicates a direct and unambiguous relationship between writer and text. This relationship is what the Hellenistic scholars aimed to restore; however, it is clear that this relationship was complicated by their very interventions. The textual critic was not a reader detached from the text and the author; he was someone who intervened in this relationship by writing on the exemplar in order to re-establish, at least in principle, the author's intentions and modes of expression. Moreover, although textual critics engaged in writing like an author, they remained somewhat marginal just like their interventions on the manuscript.
Sean Gurd has discussed the marginality that philology occupies in relation to the text as a form of ‘self-criticism’ which makes the text ‘apparently independent’.Footnote 26 Aristarchus’ principle of elucidating the text on the basis of what is in the text is perhaps the best illustration of this apparent independence; however, regardless of any sort of philological self-criticism and textual self-determination, the text still owed its existence to the exemplar as an object. In turn, this exemplar bore the traces of the scholar's annotations and conjectures along with his name for purposes of identification.Footnote 27 Finally, the text and the scholar's work encompassed in the exemplar contributed to the formation of an exclusive community demarcated by its own rules and conditions. These included (among others) the acquisition of specialized knowledge, language skills as well as methods of textual interventions. If this were not enough, what used to be contained in the margins or in between lines was soon gathered and expanded into new and independent types of texts and books: commentaries, glossaries, lexicons, and the like.Footnote 28 Parts of texts and notes on the margin were also excerpted, collected, reorganized, and circulated in other forms. All of these writings were developed to assist the work of future textual critics alongside a whole apparatus of instructions for the acquisition of proper expertise and competence; at the same time, they opened up new authorial possibilities and created a competitive environment around parallel cultural activities.Footnote 29
The librarians that worked in Alexandria were not only scholars, some were poets as well. Callimachus was one of them and, like others around and after him (e.g. Posidippus, Asclepiades, and Apollonius of Rhodes), composed poetry of all sorts by exploiting his access to a multitude of literary materials. Some years ago W. Clausen described the changed conditions in the following way:
…now a scholar-poet could con and compare texts; pluck a ἅπαξ out of Homer and define it in a context of his own making, perhaps to spite another scholar-poet; employ an obscure variant of myth or legend, the while deftly signaling to an alert reader his awareness of other variants as well; subtly modify an admired metaphor or simile; set an old word and a new one in an elegant collocation.Footnote 30
Clausen's rather crude description suggests a particular attitude towards books. Books are not simply mobile media meant to archive the works of the great poets from the past; their texts can be scoured, picked apart, and reprocessed into new texts and independent new books.
Philology in Translation: Livius Andronicus
According to Suetonius’ account above, one of the first who engaged in any sort of textual work in Rome was Livius Andronicus. We have no evidence of this outside of Suetonius but we do have a few remains from his translation of Homer's Odyssey. The first line of the Odyssey and its rendition in the Odusia read as following:Footnote 31
It has been traditional to look at this line very narrowly, thus Gian Biagio Conte in his History of Latin Literature comments: ‘Livius Andronicus, in his translation of the Odyssey, shows his acute awareness of the position of the translator: he admits in his first line that his freedom is limited, showing a respect for the original that is almost obsequious.’Footnote 32 Like other commentators before and after, Conte remarks on the archaizing choice of insece for ἔννεπε, emphasizes the semantic parallelism between uersutum and πολύτροπον, notices the alliteration uirum–uersutum, and considers the substitution of the local goddess Camena for the Homeric Μοῦσα. Some time ago, Stephen Hinds emphasized that uersutus does not simply translate the Homeric πολύτροπος; it also refers to ‘translation’ as uertere is the technical verb for it, and translation in the widest sense is exactly what Livius was doing.Footnote 33 Accordingly, the poet here constructs a close relationship, if not an identification, between himself the translator-poet and Odysseus the cunning man (uirum…uersutum). This relationship becomes even more intriguing if we consider that the first part of Livius’ cognomen is Andro- (Andronicus) and ἄνδρα is the first word of the Odyssey.Footnote 34 Since the first words of a work functioned also as titles (titulus), Livius appears to have put his signature on both the text that he produced and the manuscript that bore it.Footnote 35
It is quite clear that Livius had no intention to intervene on a Homeric exemplar and restore original intentions and phrasings through corrections in between lines or at the margins. What he was interested in was to lift the Homeric text from the exemplar at hand and to transfer it into a new one for purposes that had nothing to do with textual conservation as such; the little we have clearly demonstrates that along the way he transformed the Homeric text not only linguistically and graphically, but also metrically, in meaning, and in size.Footnote 36 What I should like to note here is that textual reprocessing from one language into another was not alien to Alexandrian scholarship as the first years of the library had witnessed the beginning of Hebrew to Greek translations at the behest of Ptolemy II Philadelphus with the Pentateuch, the Five Books of Moses.Footnote 37 Strongly influenced by what went on in the library, Bible translations present expansions, suppressions, and all sorts of reminiscence and manipulations across languages and texts.Footnote 38 Moreover, these translations alert us to the predicaments of directionality, that is, of conceiving translation as a transfer from source text to target text whose only aim is to defy linguistic boundaries. As Dries De Crom reminds us, very often translations happen in multilingual and multicultural situations, with translator and readers/audience operating within the same socio-cultural environment. In these environments, manuscripts in one language tend to coexist along with manuscripts bearing their translations.Footnote 39
Livius Andronicus too operated in a multilingual environment with some availability of Greek books and surrounded by people involved with them in similar ways. Just to cite an example: code-switching as well as puns and jokes involving Greek expressions are rife in Plautus’ comedies.Footnote 40 In the same way, the pun ἄνδρα–uirum pointing to Andronicus suggests that his readers’ horizon of expectations included a degree of competence in Greek and knowledge of Homer. Moreover, Plautus stands with Livius towards the beginning of reprocessing Greek drama for the Roman stage: in Plautus’ prologues we find the same sort of self-advertisement (Plautus uertit), although more overt.Footnote 41 Finally, from the comparison between Menander's Dis Exapaton (11–112) and the corresponding lines in Plautus’ Bacchides (494–562) we learn a few things about the impact of uertere on an exemplar text. With less than a hundred lines to work with, scholars have noted that Plautus combined two monologues into one, omitted two dialogues, reversed the order of entry onto the stage, and expanded the conversation.Footnote 42 Cuts and variations characterized the Odusia as well; the approach is the same, the end product generically different. What makes the Odusia somewhat unique compared to Plautus’ comedies are the exegetical materials that it contains.
Some years ago Franco Montanari attempted to define similarities and differences between glossary and paraphrasis; they both present the same exegetical material and refer to a text in a syntagmatic order but they are different in the way they relate to the exemplar:
…the glossary breaks down the text into single words or phrases and organizes it by placing a translation next to each word (often into two columns but not always); the paraphrasis constructs a continuous discourse that can be read autonomously and ‘covers’ completely the ‘translated’ text.Footnote 43
Paraphrases and glossaries tended to be written interlinearly or in the margins of an exemplar; however, paraphrases could easily exist as separate entities from the paraphrased text. This is because the paraphrasis does not need to refer to a specific lemma (word or phrase in the text) like the glossary. As often remarked, Livius frequently engages in paraphrasing. Here are some examples: uersutus and Camena (fr.1 Mariotti) in the first line are clever translations but they are also exegetical comments.Footnote 44 The Camenae were connected with a sacred grove and a spring just outside Porta Camena or Capena, and with King Numa.Footnote 45 In the verse Mercurius cumque eo / filius Latonas (fr.10 Mariotti) Livius does not call Apollo by name; he identifies him as ‘the son of Latona’, therefore offering genealogical information while playing with patronymics (here matronymics). In sancta puer Monetas / filia med docuit (fr.12 Mariotti; med: Flores) Livius refers again to the Camena and calls her ‘daughter of Moneta’.Footnote 46 It is often assumed that Moneta is glossing Mnemosyne; if so, the ultimate source would be Hesiod's Theogony (v.53) and the text would imply that Livius had access to a copy of it. Moneta, however, existed also as a deity in Rome. In 345 BCE L. Furius Camillus dedicated a temple to her as Juno Moneta and built it the year after on the Arx (in the garden of Aracoeli).Footnote 47 The Libri Lintei, which recorded the names of the elected consuls, were stored in this temple (Livy 4.20.8). From 273 BCE onwards it also housed a workshop for the minting of silver coins (Livy 6.20.13).Footnote 48 Accordingly, the translation of Mnemosyne from a Hesiodic exemplar into Livius’ text via Homer incorporates also a paraphrasis that points to an important place in Rome where the memory of the past was preserved in books and other media. In relation to Livius’ text and his Camena, the word Moneta fulfils an in-text exegetical function; in relation to the Homeric exemplar, it is glossing material derived from Hesiod and redeployed inside another text and book.
Exegetical work done by others seems to have played a role as well; in fact, it appears to inform the construction of Livius fragment 11 (Mariotti), which translates a dancing scene in Odyssey 8 (378f.): nexabant multa inter se flexu nodorum dubio (‘they were thoroughly entwined in each other in an alternating winding of loops’). In the early 1930s Hermann Fraenkel noted that Livius’ verse looks to the Homeric text, ταρφέ’ ἀμειβομένω (‘exchanging places often while dancing’. 8.379); however, the first part (nexabant multa inter se) replicates a paraphrasis to be found in the V-scholion: πυκνῶς πλέκοντες εἰς ἀλλήλους (‘they were thoroughly entwined in each other’). We have no clue about the second part of the line; however, it is quite possible that the scholion was at some point much longer and more elaborate.Footnote 49 At the time Fraenkel admitted that the material used by Livius antedates Aristarchus.Footnote 50 Edward Courtney has commented on the issue by suggesting that ‘the school master Livius was using a traditional explanation of the schools later incorporated in scholia.’Footnote 51 Lack of evidence for pre-Aristarchean commentaries does not preclude the fact that exegetical work on Homer already existed; moreover, traditional paraphrases, whether transmitted orally, in the margins of an exemplar, or in separate books, were still dependent on close inspections of manuscripts and traffic in written objects.
As I already pointed out, in the Hellenistic world philologists tended to work closely on an exemplar with an approach that was suspicious and conservative. In the case of the Odusia, the Homeric exemplar is nowhere to be seen or touched, and the Homeric text (including exegetical materials by others) is reworked into a new text preserved in a new exemplar; that said, Homer is far from being overwritten and the Odusia is far from having a completely independent life. Livius’ text begs the reader to turn to a Homeric exemplar over and over again, to be suspicious of Livius’ word choices, explanations, and elaborations of and on meanings. That this was the case can be seen in Horace centuries later. In the Ars Poetica, Horace famously discusses the shortcomings of the ‘faithful translator’ by offering his own version of the Odyssey’s opening lines:
These lines can be viewed as a corrective commentary on Livius’ work. Here we see Horace expunging πολύτροπον and, therefore, erasing Livius’ choice of uersutum and the Homeric epithet in one stroke. Philologically, the deletion is radical; however, Horace compensates for it by invoking the proper deities and by using the Homeric hexameter (adding to Ennius’ emendations of Livius’ text, as we shall see). In all, Horace's extends the editorial tradition by bringing to light hidden aspects of the original text even while obscuring others because viewed as inappropriate. In a way, the chain of individual interventions creates a story revolving around an exemplar that gains eternity through textual emendations and exegetical elaborations carried out in other exemplars. For if each text (Homer's, Livius’, and Horace's) owes its existence to individual interventions and can be attributed to individual authors, taken together they produce a narrative that is coherent and driven by the same effort and goal.Footnote 52 Retrospectively, they appear as a sequence of texts which, like sketches in a storyboard, pre-visualize continuities and changes leading to the same end: the survival of previous works, the creation of authorial multiplicity, and the proliferation of books. This is the disposition towards literature and authorship to which Livius contributed alongside with specific ideas about the process of reading, writing, and intervening.
The Poet Philologist? Ennius
It is generally accepted that Ennius was born in 239 BCE, either before or just after the death of Callimachus. Coming from Apulia, Ennius was educated in a Greek-speaking environment, and the remains of the Annales suggest that he had some exposure to the Aetia.Footnote 53 Here are some of the key fragmentary lines used by scholars in their assessments: after his invocation of the Olympian Muses and their beating feet (Musae, quae pedibus magnum pulsatis Olympum, 1.1 Sk), Ennius describes his poetic investiture by drawing on a framework familiar to us from Hesiod and reworked by Callimachus.Footnote 54 As opposed to the Greek precedents, in his dream Ennius encounters Homer (uisus Homerus adesse poeta, 1.3 Sk) and, during their conversation, finds out that Homer's soul has migrated into his body.Footnote 55
Book 7 included other programmatic materials. In a famous fragment Ennius shuns the Saturnian by associating its rhythm with pre-poetic figures of song and by refusing to write about the war treated by Naevius (scripsere alii rem / uorsibus quos olim Faunei uatesque canebant, ‘others wrote on the subject by using verses [turns of language] which were once sung by fauns and uates’, 7.206f. Sk). It is generally assumed that Ennius looked to convey a sense of superiority in these lines by situating his predecessors in an uncouth past.Footnote 56 Although their attribution to specific books is highly contested, other fragments add to our view on Ennius’ positioning.
The first, preserved by Servius (ad Aen. 8.361), runs as follows:Footnote 57
contra carinantes uerba atque obscena profatus
(Ann. Sedis Incertae fr.576 Sk)Some years before Skutsch's edition, Alberto Grilli proposed to situate the fragment at the opening of Book 7. In his view, profatus relates to the religious-divinatory sphere and would take Ennius as its subject. Consequently, in this line the poet would have cursed his detractors as a way of concluding his attack on Naevius and other poets writing in Saturnian. Grilli supported his proposition by turning to Callimachus’ virulent attack on the Telchines:
I know that often the Telchines, who are ignorant and are not friends of the Muse, grumble at my poetry…Footnote 58
Rather than using a verb equivalent to epitrúzein to define the abusive language of his rivals, Ennius would have chosen a visual image drawn from the agricultural world as carinare would be related to carere (‘to card wool’). Accordingly, Grilli interpreted the meaning of this verse along these lines: ‘having spoken abusive and obscene words against those who tear apart with insults.’ This is all good and well, except that it is hard to see how Ennius could be claiming to be prophetically speaking obscena. It would make more sense to accept profatos (the accusative plural) and read this in parallel with carinantes:
Interestingly, Servius preserves the following line in the same context:
In his edition of the Annales, Enrico Flores inserts this fragment into book 7 going against the tendency of placing it elsewhere (i.e. the Satires).Footnote 59 Regardless of the text to which it belongs, the line expands on the position that Ennius takes in the fragment previously quoted. carinare ties the two fragments together, except that in the first one it seems to encapsulate the notion of adopting improper diction, while in the second one, it appears to refer to the written word. Finally, Grilli found a cohesive framework to book 7 suggesting that, just like in Callimachus’ Aetia, so too Ennius’ attack is lodged in a dream setting that looks both back to the beginning of the poem and his encounter with Homer and slightly ahead to his claim to specialized knowledge:Footnote 60
Thomas Habinek has argued that these two lines anticipate the reevaluation of Roman traditions in accordance with Greek philosophy, and testify to Ennius’ intervention in ‘a roiling debate over the nature, meaning, and class identification of sapientia.’Footnote 61 In his view, the dream of sophia would have had the same relationship with the Pythagorean tradition as the process of metempsychosis that opens up the poem. In the Greek context, however, sophia did not refer solely to philosophy; it also incorporated into its semantic area the meaning of poetic techne, i.e. the ability to create poetry that can only be judged through the poetic object.Footnote 62 If the equation of sophia with sapientia is not at as culturally straightforward as one would like to assume, its construction relies on the way glossaries were organized. In the line sophia works as a lemma and sits right before sapientia which operates as an explanation. Inserted inside the poetic text, lemma and gloss are locked together almost as an undividable unity.
In another fragment Ennius makes yet another competitive claim:
Nor to the Muses’ crags…nor anyone was a ‘specialist of the word’ before this man…we dared open…
It has long been tempting to read the phrase dicti studiosus as a direct translation of philologos.Footnote 64 Used for the first time in Plato's Phaedrus, the idiom is attributed to Socrates and meant to refer to himself (236e5). Rudolf Pfeiffer interpreted the word as someone ‘fond of talk, dispute, dialectic in a wide and rather vague or ironical sense’.Footnote 65 In the De Grammaticis et Rhetoribus Suetonius reports that Eratosthenes adopted the ‘nickname’ of philologos to qualify his possession of ‘varied and multifaceted knowledge’ (multiplici uariaque doctrina, Gramm. 10.4).Footnote 66 As far as we know, the same Eratosthenes defined grammatiké as ‘complete mastery (héxis) in letters with letters meaning written compositions (sungrámmata)’.Footnote 67 The relationship between the two testimonies has been variably discussed; most probably, Eratosthenes saw grammatiké as a field of specialized activities focused on texts which demanded philologia, understood as encompassing universal knowledge.Footnote 68 The claim seems bold. Some years ago Sander Goldberg commented that the word may not have been ‘an Alexandrian catch-word in Ennius’ day’ or that it may not necessarily suggest that ‘Ennius means to associate himself here with Alexandrian literary canons’.Footnote 69 Regardless of this, the fact is that even in this pitifully damaged fragment Ennius presents himself as a new kind of poet. The crags of the Muses and the opening up of either ‘gates’ or ‘sources’ are definitely Callimachean.Footnote 70
As it stands, most scholars remain skeptical about singling out Callimachus’ poetry as Ennius’ main model for it is clear that a great variety of cultural materials went into the composition of the Annales.Footnote 71 In a case like this, it is wise to go back to the distinction between ‘code/genre model’ and ‘example model’ elaborated for the study of Latin poetry by intertextualists like Gian Biagio Conte and Alessandro Barchiesi in the 1980s.Footnote 72 The first type of model has to do with a situation in which a text betrays the use of formulae and features that are drawn from previous texts and that are characteristic of a particular genre (say, epic, lyric, etc.); in the second case a text is used as an exemplar, it is as if copied into another text inside another book with individual words or passages cited and transformed. In going back to Callimachus what we see is that Ennius extracted particular scenarios (i.e. the dream, the Muses’ rocks) or scenes (i.e. attacks on predecessors and competitors) from Callimachean passages, and reprocessed them to become part of a new narrative. The point here is not so much to establish whether Ennius espoused Alexandrian or Callimachean canons as such but to acknowledge that his way of handling books and texts is in line with contemporary practices. In relation to Callimachus (here the exemplar), Ennius’ approach is not conservative; it is transformative. Transformation presupposes the exemplar's reduction into smaller units at different levels of the text (lexical, thematic, structural, and so on). Somewhat ironically, breaking down the text into parts is foundational to philology since the attempt to emend alterations and errors requires the close inspection of discrete elements of the text.
Ennius offers a fine example of how the conservative principle meets with the transformational principle in another fragment:
These two lines presuppose Livius’ text and his invocation to the Camena in the first line of the Odusia.Footnote 73 As if working from Livius’ exemplar, Ennius emends the text and restores the original Homeric deity, the Muse; at the same time, he pushes forward the meaning of insece from ‘sing’ to ‘pursue’, perhaps going so far as to correct graphically insece into inseque.Footnote 74 The semantic shift (along with the graphic emendation if this was indeed carried out) obscures the idea of singing as being improper to Ennius’ poetic project without erasing the Livian precedent. With the Livian insece kept or assumed, the line beckons to a loaded history, calling for further interventions and multiple authorships. Horace's choice of insector in Epistles 2.1.71 speaks precisely of these:
I am not in pursuit of Livius’ songs, and think that they ought to be destroyed, which I remember Orbilius to dictate to me when a boy with his rod.
In relation to the Annales, Horace's insector works as an external gloss: it makes explicit the variant of ‘pursuing’ as proper to Ennius’ poetry and, at the same time, records the emendation performed by Ennius on Livius’ text. In the history of insece across the three texts, we can see very clearly the materiality of philological practices in Rome. In the lines just quoted, Horace captures this history while denouncing the influence of grammarians like Orbilius (a well-known interpreter of archaic Latin poetry) and their narrowly authoritarian readings.Footnote 75
Dramatic Philology
To measure fully the impact of philological practices on the creation of a poetic tradition in Latin it is important to have a look at the early dramatic tradition as well. The Prologues to Terence's comedies provide us with some compelling clues. Let us turn first to the Adelphoi:
Since the poet has found that his writings are carped at by unfair critics, and that his adversaries represent in a bad light the play that we are about to perform, he shall give information about himself; you shall be the judges whether this ought to be esteemed to be to his praise or to his discredit. The Synapothnescontes is a Comedy of Diphilus; Plautus made it into a play called the Commorientes. In the Greek, there is a young man, who, at the early part of the play, carried off a courtesan from a procurer; that part Plautus has entirely left out. The poet [i.e. Terence] has adopted this in the Adelphoi, and has transferred it, translated word for word. This new play we are about to perform; determine then whether you think a theft has been committed, or a passage has been restored to notice which has been passed over in neglect.
(tr. Riley, slightly adapted)The prologue of the Adelphoi unfolds before our eyes how early Roman comedy was composed. Plautus—who belonged to the generation before Terence—wrote his plays by cutting out portions of Greek texts, translating, and recomposing them into new poetic outputs. By the time of Terence the rules of composition seem to have changed: once used, a passage from a Greek original can no longer be used again; reusing a part that has already been used is associated with stealing. The Prologue argues that the poet has excerpted a section left untouched by Plautus and translated it word by word into his own composition.
The process of composition outlined here makes explicit the practical handling of manuscripts and texts that we surmised from Livius and Ennius. For one thing, the exemplar is unambiguously represented as an object that can be dissected and parsed into smaller parts; second, some of these parts are said to have been excerpted and transferred into another text.
I would like to zero in on excerpting as this is shown to be crucial in the passage. In his contribution ‘Excerpting as a reading practice’, David Konstan points out that excerpting is a reading practice which presupposes access to texts and choice of passages to be jotted down. In its simplest form, the practice responds to the need to extract from exemplars the most notable passages or verses in order for them to be used in other contexts, a speech, a poem, a dramatic text, to give substance and weight.Footnote 76 Evidence for this practice goes back to Aristophanes through Plato and Xenophon, and includes important compilations like those ascribed to Tyrteus and Theognis. Surprisingly, perhaps, reading and excerpting were far from being solitary activities. In the Memorabilia, for example, Xenophon's Socrates talks about it in these terms: ‘I go about reading [unrolling] the treasures of wise men, which they have written and left behind in books, in common with my friends, and if we see something good, we excerpt it.’Footnote 77 If excerpting as such was indeed a communal practice, clearly the collection and circulation of these excerpts played a key role in the transmission of texts and in shaping the cultural horizon of future readers throughout antiquity.
The Prologue's speech in the Adelphoi implies the formation in Rome of a community of readers engaged in reading and excerpting from Greek dramatic texts, and in transferring and assembling the cut-outs into collages in Latin. Interestingly, the complaints of Terence's rivals place authorship in the realm of textual plucking rather than in the composition of new texts. Not only that: authorship is figured as a claim of ownership over the excised passages. In turn, proof of ownership is based on the new text and a systematic collation of this new text with both Greek originals and contemporary Latin texts of the same genre. In philology, collation relates strictly to the arrangement and ordering of a number of texts in order to establish their relationship, to compare, and classify their variants with the goal of producing an authoritative archetype. At its most basic level, collation implies reading two texts placed side by side and noting differences. In Terence's case, collating is figured as the means by which to record and police the appropriation of portions of Greek dramatic texts by poets in Rome.
The adaptation of philological practices entailed here is, in many ways, extraordinary. In line with them, Terence defends himself by arguing on pure philological principles and by adopting terminology derived from philology: he has picked up ‘a passage left untouched’ by Plautus (locum integrum) and he has ‘expressed it word by word’ (uerbum de uerbo expressum). Rather than claiming ownership over the passage excised from the Greek original, Terence places himself in the position of the philologist whose mandate is to recover and preserve. Recovering and preserving is here figured as the relocation of materials from the original exemplar into another text written in a new book and into another language by adopting strict linguistic equivalences and correspondences. What used to be hidden by neglect is now ‘brought back to life’ (reprehensum). What stands out in this defense is the transformation of philological tenets into the promotion of poetry as the means by which transmission and conservation can be upheld in practice.
The prologue to the Andria allows us to expand our view:
Menander composed the Andrian and the Perinthian. He who knows either of them well, will know them both; they are in plot not very different, and yet they have been composed in different language and style. He confesses that he has transferred what was suited from the Perinthian into the Andrian, and has employed that as his own. These parties censure this process; and on this point they differ from him, that plays ought not to be mixed up together. By being thus knowing, do they not show that they know nothing at all? For while they are censuring him, they are censuring Naevius, Plautus, and Ennius, whom our poet has for his precedents; whose carelessness he prefers to emulate, rather than the mystifying carefulness of those parties.
(tr. Riley, slightly adapted)In this passage the poet makes a claim of ownership which, at first sight, seems to replicate the logic of Terence's detractors in the prologue of the Adelphoi; at a closer look, however, we notice that he is not at all claiming for himself the passages extracted from the Greek plays; he is only alleging to make use of them as if they were his own. The regimen implied in this assertion is one of usufruct. Put simply, usufruct is the right of a person to use another's property and to enjoy its fruits for a certain period of time. In Roman law, the usufructuary treated property as if his own and had the duty of maintaining it in the state in which it was received. The standard of care required was that of the bonus paterfamilias so he was liable for any deterioration or negligence.Footnote 78
Declaring to be using excerpts as in a condition of usufruct allows Terence to remain within parameters of ownership that are acceptable for his detractors; however, another issue comes up in this context, that is, the practice of mixing excerpts from different exemplars. To evade this further charge, Terence uses knowledge as a weapon: it is only by reading a multitude of Greek originals and knowing the Roman dramatic tradition well that any judgement regarding language and style can be made. The tension between philological conservatism and poetic innovation is very clear here, and it is in a passage like this that we can also see how and when the split between philology and poetry in Rome started to emerge. Terence claims to prefer the sloppiness of Plautus to the obscura diligentia (‘the mystifying diligence’) of his enemies, who, in his view, use their philological knowledge only to chastise and denigrate.
Using portions of texts that have been already used by other poets or deploying two or three originals in order to create a new play in Latin, this seems to be the bone of contention. In the prologue to the Heautontimoroumenos, Terence discloses the technical term for the latter process:
As for the malevolent rumors spreading abroad that he has mixed together many Greek plays while writing a few Latin ones, he does not deny that this is the case, and that he does not repent of so doing; and he affirms that he will do so again. He has the example of good poets; after which example he thinks it is allowable for him to do what they have done.
(tr. Riley, slightly adapted)contaminatio, this is how Terence's detractors define the transformational work inherent in the poet's combination and recomposition of cut-outs from different dramas. The word has strong religious connotations, and it does denote the gravity with which texts (Greek or Roman) had begun to be handled and scrutinized. Two decades or so ago Goldberg remarked that the charges of contaminatio lodged against Terence had less to do with aesthetic issues than with the practical fact that, by making many plays out of a few, he was unfairly reducing the store of Greek originals available to the poets operating in Rome.Footnote 79 In another way, however, Terence's prologues alert us to the fact that by the mid-second century BCE Rome enjoyed a well-established philological and literary community with its own rules of engagement and reasons for contention.
Conclusion
In The Powers of Philology: Dynamics of Textual Scholarship, Hans Gumbrecht suggests that ‘all philological practices generate desire for presence, desire for a physical and space-mediated relationship to the things of the world (including texts).’Footnote 80 Among the activities that make up philology, the most intriguing is perhaps textual editing. This is because—as Gumbrecht points out—it unleashes the imagination, ‘conjuring up the desire of embodying the text which can transform itself into the desire of also embodying the author of the text embodied.’Footnote 81 There is no doubt that these appetites and urgencies informed the development of philological practices in Rome, and found a material expression in the creation of a poetic tradition in Latin. The early Roman poets embodied Greek authors along with the texts that embodied them, and they did so in a variety of ways: Livius Andronicus fashioned himself as Odysseus, facilitating a textually mediated relationship with Homer through his own work. Ennius presented himself as Homer reincarnated, giving materiality to his philological desire of authorial embodiment in a most conspicuous way. In elaborating on Callimachus’ precedent and realigning Livius with Homer, Ennius locked his philological work into a new text to be passed on for further interventions and corrections. Excerpting and composing collages from Greek dramatic texts ‘faithfully translated’ is at the basis of poetic drama in Rome. Authors like Terence imagined this creative process as a method for preserving texts and for ensuring their regenerative power. Philology fuelled these early poets’ imagination, their craving to touch and handle exemplars and texts, to add, correct, comment, explain, delete, correct, and rewrite.
The survey I present in this article is only a partial snapshot of the impact of philology as a practice in Mid-Republican Rome. Texts in books stimulate and increase the capacity to imagine the past, to resurrect those who no longer exist. Philological writing in the ancient world was moved by the impulse to make the world of the past present, to have an immediate connection with that world, and to incarnate over and over the authors of the past.
