Sometime in the later part of the nineteenth century, from her family home in Amherst, Massachusetts, Emily Dickinson wrote the following poem on a scrap of lined paper, which she folded as an enclosure into a letter to her cousins in Boston:

A famous recluse who rarely left Amherst, Dickinson composed these lines a two hours’ drive from where I live, just across the Charles River from Boston, in Cambridge. They were not published until after the poet’s death, in a collection of poems and letters issued in 1894, when they entered the American canon. When I was a child growing up outside Boston a century later, they were perhaps the first lines of poetry I ever memorised.

The phenomenon described by Dickinson here – that reading a book is a means for the mind to freely travel – is one familiar to all of us who love reading and one that has appealed to me since girlhood. But as I myself have come to identify as a writer and an author in adulthood, I have come to recognise another facet of Dickinson’s verses. When she describes the book as a ‘frigate’, a quick little boat, that can ‘take us lands away’, she is thinking both as author and as reader. For the reader, the book is a vessel for the imagination and a device for intellectual mobility. But for the author, too, the book is an object that travels, a ship that sails.

***

The phrase that is the focus of this article, vade, liber, ‘Go, book’, is one that reflects this second reading of the meaning of Dickinson’s frigate. Although, as we will see, vade, liber first appears in Ovid’s poems from exile – a series of works that often frame themselves as letters – the instinct to address the poetic text as it is sent away has its origins earlier, in the late Republican lyric productions of Horace and Catullus, as well as a long tail of repetition and reception in Latin poetry of the first millennium CE and in vernacular European literature of the Middle Ages and after. This article joins the many other scholarly discussions of this fascinating, transhistorical material by highlighting the imperial context of this literary gesture and placing it within the larger phenomenon of what we are calling here the Roman spatial turn.

Ovid’s Tristia were written in the years following the poet’s exile to Tomis, on the Black Sea, some six or seven months’ journey from Rome.Footnote ² In the first poem of the first book of these poems, the author bids farewell to his newly written liber, a papyrus book roll, portrayed as an enslaved person, sent abroad with instructions to find his way back to the owner’s house in Rome.Footnote ³

To scholars of the ancient book, Ovid’s poem is best known for the rich trove of references it gives to the shapes that ancient books took, negatively adduced as the trappings that Ovid’s book does not have: no purple cover, no fancy bosses, no rubricated title, no fragrant oil.Footnote ⁵ Texts like this one have been vital sources for helping us to reimagine the ancient Latin poetry book as it must have looked in its own day, filling in the remarkably restricted picture afforded by papyrological survivals.Footnote ⁶ But in the context of the present discussion of Roman studies’ spatial turn, we may return to Ovid’s farewell to his book – and his original coinage of the vade, liber refrain – as an entry point into the discourse about place and space with which Roman literature engaged so dramatically in the years of imperial expansion.Footnote ⁷

Text and space are interconnected concepts in the Ovidian imaginary: like letters and petitions, the book moves where the poet cannot go, both metaphorically as the enslaved man making a journey on his enslaver’s behalf to the capital and literally as a text that circulated beyond the author’s place of exile in Moesia, at the edge of the Roman Empire.Footnote ⁸ Words written by Ovid and others were one of the poet’s principal obsessions in his exile poems, and Ovid often used texts and their representations within his poetry to negotiate his relationships with the imperial regime and with the literary past and future.Footnote ⁹ Moreover, as Sarah Lindheim has written, the very project of Roman elegy was informed by the phenomenon of cartographical thinking that was part and parcel of Rome’s imperial expansion in these decades.Footnote ¹⁰ For Ovid in exile, the mobility of texts was not a fantasy but a reality.

As a primary example of what I am going to call poetic envoi, Ovid’s apostrophe to his book in Tristia 1.1 serves as a model for later authors and readers.Footnote ¹¹ And it opens several interpretative doors that I would like to enter here:

1. (1) It calls on the notion that the book, as an object, has a physical identity independent of its author, one that is activated in the very reality of its different and independent mobility in the space of empire: that is, Ovid’s book can and will travel where its author cannot.

2. (2) It invokes the equal and opposite notion that the book, as an object, is if not physically then metaphysically connected to and dependent upon the identity of its author, notionally ‘belonging’ to the author as a child to his father or an enslaved person to their enslaver, to name two of the most active metaphors for book production in antiquity, which are animated by the wordplay between Latin liber, ‘book’, and līber, ‘child’ in Ovid’s poem. That is, like a child or a slave, insofar as the book is seen to be an extension of its author, by sending it away, Ovid may notionally move through Roman imperial spaces – discursive, legal, geographic – by bibliographic proxy.Footnote ¹²

3. (3) Finally, by attending to these dynamics of mobility, we see how the specific phrase, ‘vade, liber’, emerges as a textual crux: a formulation that pointed to tensions within the ‘discursive’ and coercive systems of imperial space that would itself come to be repeated and reworked in various ways throughout the literature of the period and after. As an emergent ‘meme’ in Latin literature, the vade, liber refrain is one that gives us a way in to thinking about mobility as a phenomenon underlying many of our own notions of textual existence, transmission, and survival.Footnote ¹³

With this in mind, I begin briefly by revisiting some of the background about the culture of writing in the Roman period, arguing that writing itself emerged as a translocal and proto-global phenomenon in the ancient Mediterranean. Next, I introduce possible theoretical frameworks for considering the nature of the book as an object that travels. And ultimately, in the last section of the article, I return to the formula of the vade, liber refrain and the trope of the author’s envoi to the book to track its emergence in the Roman literary canon and to posit some theories about how it relates to questions about the dynamics of authorship and reception that emerge from the spatial turn.

***

In Empire of Letters, I argued that writing at Rome was always a translocal and transhistorical activity: like many of the most important cultural institutions at Rome, the alphabet itself was recognised to be a borrowing of a borrowing (from Phoenicians to Greeks to Etruscans to Romans), and the use of writing was seen to put Romans on equal footing with more ancient writing cultures around the Mediterranean.Footnote ¹⁴ Moreover, when we consider the particular materialities of Roman writing practices, we find further evidence for their spatialised characteristics. In local contexts, local materials were everywhere used for everyday note-taking: archaeological remains and literary testimony show us that ostraca, waxed tablets, folded leaf tablets, papyrus sheets, pieces of animal skin, and apartment walls were all enlisted as supports for inscription.Footnote ¹⁵ Despite the scarce early references to books made from membrana (‘parchment’) in Martial’s epigrams, the papyrus scroll (liber, volumen) was everywhere the preferred medium for books, and everywhere but Egypt, papyrus (charta, papyrus) was as a rule imported.Footnote ¹⁶ High-quality ink pens (calami) might also have been trade goods.Footnote ¹⁷ Moving beyond the traditional poetics of gift-giving and writerly friendship, as a more developed book trade emerged in the very period we are discussing, books themselves came to be seen as consumer products to be exported and traded, much to the anxiety of their authors.Footnote ¹⁸ The materiality of books as much as their contents linked them to systems of movement and exchange. Thus, even before they were written, Roman books were objects that travelled.

In addition to the established frameworks of book history and bibliography, the critical approaches of mobility studies provide a useful lens through which to consider how Roman books both travel through and create spaces of connection within the Roman empire, as we shall see below.Footnote ¹⁹ When it comes to textual studies, too, mobility has been an important metaphor informing several of the paradigms by which we understand texts’ transmission and survival, including translation (as a ‘carrying over’ from one language to another), mouvance (a term coined in the 1970s by Paul Zumthor to describe the characteristic variability of medieval poetry in the vernacular performance tradition), variance (Bernard Cerquiglini’s extension of mouvance into the scribal and manuscript context), and even – a term already mentioned – meme (Richard Dawkins’s coinage for a unit of cultural information spread by imitation).Footnote ²⁰ It is perhaps the last of these that is most pertinent to the topic which will occupy the remainder of the article, where we return to the vade, liber refrain and the book sent away by its author.

***

It is the case that the formula of author’s envoi to the book is, within the classical tradition, an innovation of Roman poetry in the first century BCE, without direct Hellenophone precedents.Footnote ²¹ And, as we shall see here, it is a direct response to the opening of imperial space in that period. If we take the formula to consist of an instruction to the objectified text and a vocative address, it appears as early as Catullus 35, where the poet asks a papyrus letter to fetch his friend Caecilius from Comum to Verona:Footnote ²²

Strictly speaking, this model falls slightly short of the vade, liber formula: although the movement of the text is implied, it is the addressee of the letter who is told to come (veniat) to hear Catullus’s thoughts, rather than the poem that is told to go to express them on his behalf.Footnote ²³

For a more immediate precedent of Ovid, we look to Horace, who idiosyncratically frames the authorial sphragis at the end of the first volume of the Epistles not as a literary letter, but as a direct address to book itself:Footnote ²⁴

Using the occasion of bibliographic closure to set the book loose into the space of the ancient Roman book trade,Footnote ²⁵ Horace makes reference to the Vicus Tuscus, the major street of bookshops in the ancient city of Rome close by the temples of Vertumnus and Janus, and to the Sosii brothers, a family of booksellers who operated there.Footnote ²⁶ As Ovid will likewise do in the Tristia, Horace speaks to the book as if it were an enslaved boy, although here it is one wishing for freedom from the author-master that will only be found in prostitution at the bookseller’s stall.Footnote ²⁷ By this formulation, Horace exploits the tension between openness and closedness, the public and the private, that is characteristic of Roman treatments of book circulation in this period.Footnote ²⁸ With promiscuity standing in for fame, the slave-book is made to accept Horace’s displaced desire for wide publication while author is allowed to espouse his usual preference for the notional privacy of circulation among his coterie, characterised here by protection in a private library, under lock and key (odisti clavis …), and by copies passed directly from one friend to another, signed and sealed (… et grata sigilla).Footnote ²⁹

At the end of the poem, the address to the book becomes an inversion of the common epigraphic formula of the speaking object, in which the funerary monument addresses the passer-by to eulogise the honorand. Echoing the request made by the poet in Catullus 68 that the Muses ‘let his paper speak this in its old age’ (facite haec charta loquatur anus, 46),Footnote ³⁰ Horace now asks the book to remember him, its former enslaver, to far-away readers in Tunisia and Spain:

In Horace already we can uncover several impulses that will stay with the vade, liber trope through its later codification, ranging from the precise syntactical structure of the imperative address to the localisation of author, book, and reader across space and time and the description of the personified book as ventriloquising author’s voice.

At this point, we are able to outline what will emerge as the standard form of the trope characterised by four features: (1) an imperative verb of motion with a vocative address of the book itself; (2) an elaborated fiction that the book, thus personified, may speak to others on the author’s behalf; (3) use of the occasion of sending to characterise author, book, and readers (both real and ideal) as occupying distinct geographical places and chronological moments; and (4) a tendency to exploit these dynamics of real and imagined presence and absence to negotiate authorial anxiety about the violation of privacy that literary publication and circulation entail.Footnote ³¹

***

When Ovid borrows from Horace, in the passage of Tristia 1.1 with which we began, the dynamics of space are obviously very different, inverted by the conditions of exile, but these same features appear:Footnote ³²

Here we find repeated the characteristics outlined above: the direct address to the book, named as such, with an imperative verb of motion (i.e. vade, liber); the request that the book speak on behalf of the author (verbisque meis … saluta, … dices, … negabis); the fact of the distance between author and audience, mediated by the book that moves between them; and finally, a sense of author’s distinction and even alienation from both audience and text, who follow different paths than he does and have futures that are ultimately unknown to him (siquis … siquis …).

What in Horace was articulated as the author’s desire for the book to stay when it wished to leave, in Ovid is transformed into the author’s wish for the book to leave when it is the poet himself who wishes to go: possum nunc meus esse liber! (‘Would that I could be my book!’, 58). Where Horace’s book is sent away from the city to the provinces, Ovid’s book is sent from the periphery to the capital with instructions for how to find the master’s house and to navigate the monumental city safely.Footnote ³³ Horace’s fuge, ‘flee!’, becomes Ovid’s vade, ‘go!’ and the book, the bibliographic proxy by which Ovid hopes to see and be seen again at Rome (i.e. with his poetic ‘feet’). Rather than having the book tell of the author after his death as Horace’s is asked do, enacting the inverted temporality and spatiality of exile, Ovid’s talking book-object is rather made to say of the absent author that he still lives: vivere me dices.Footnote ³⁴

A closely connected example from the third book of the Tristia repeats the phases from Tr. 1.1 (vade … vivere me dices), but reworks the bibliographic envoi as epistolary letter:Footnote ³⁵

As Stephen Hinds has observed, the striking personification of the book-as-traveller is somewhat tempered here as Ovid incorporates the imperative refrain (vade … littera) into an epistolary formula (vade salutatum … Perillam), effectively glossing the salutation Naso Perillae suae salutem dicit in verse.Footnote ³⁶ Though it may have been hastily written (subito perarata), in the remainder of the poem, the letter is made to speak with Ovid’s voice, and in so doing recaptures the grandeur of his claims to literary immortality familiar from the end of the Metamorphoses: me tamen extincto fama superstes erit, | dumque suis victrix septem de montibus orbem | prospiciet domitum Martia Roma, legar (‘Even when I am dead my fame shall survive. As long as Martian Rome gazes forth victorious from her seven hills over the conquered world, I shall be read’, Tr. 3.7.50–52). Ironically, the extension of imperial space and time is the source of both Ovid’s punishment and his imagined triumph.Footnote ³⁷

A final instance of the formulaic envoi in Ovid appears in the last book of the Epistulae ex Ponto, effectively rewriting the fantasy of urban publication that played out in Tristia 1.1 as a personal and private dedication of poems to then-consul Sextus Pompeius.Footnote ³⁸

Whereas in Tristia 1.1, the book was given the itinerary of the city of Rome with directions for finding safety at home at the poet’s house and instructions to make haste (longa via est, propera!, ‘The road is long. Make haste!’, 127), Pont. 4.5 is a more mature (resigned?) product of Ovid’s exile, given a more extended – though still imaginary – itinerary of the long ‘return trip’ from Tomis to Rome: crossing of the Thracian land mass, passage over the Aegean and Greece, and arrival to Italy by way of the Ionian Sea, at which point the journey to Rome requires only ten more days. Where the road is long(er), the visit to Rome is short(er): now, the ‘light couplets’ are told to eschew the crowd and go directly to the safekeeping of their patron. The neoteric aesthetics of slightness are again put to material ends: for such a journey, it is apparently better to pack light.

***

Though Ovid used the vade, liber formula multiple times in the exilic works, by reworking both Horace’s and Ovid’s versions of book apostrophe many times throughout his writing, it is arguably Martial who ‘memeified’ the formula, incorporating the imperative verb form and the now-familiar dynamics of surrogate speech and travel-by-proxy.Footnote ³⁹ Near the very beginning of the first volume of Epigrams, Martial explicitly copies the language and theme of Horace Epist. 1.20 while also borrowing the parve liber phrasing from Tristia 1.1:Footnote ⁴⁰

In 1.70, Martial reworks Ovid’s Perilla poem (Tr. 3.7), excusing himself for not coming in person and sending his book instead along the route from the author’s home on the Quirinal to his friend Proculus’ on the Palatine:Footnote ⁴¹

Recalling the journey of Ovid’s Tristia to Rome, Martial addresses his book again in a pair of poems in Epigrams 3 which introduce for the theme of Martial’s retreat from Rome to Forum Cornelii (modern Imola) in Cisalpine Gaul. Along with Book 12, composed in the poet’s hometown of Bilbilis in Roman Hispania, because of the thicket of allusions to the Tristia, these are often described as charting Martial’s ‘exile’.Footnote ⁴² Unlike Ovid, though, Martial left the city by choice, and epigram 3.4 celebrates rather than bemoans his distance:

In the pair poem, 3.5, the book is again a traveller, now arriving in the capital covered in dust and, as in Ov. Pont. 4.5, given precise instructions for how to find its way to the friend’s house in the new town:

Drawing attention to the two models of book circulation evoked in the formula of the book apostrophe – to be consumed by the few or by the many (vis commendari … multis, an satis unus erit? 1–2, ‘do you wish to be recommended to many or will one suffice?’) – Martial directs this book to find satisfaction in private readership and enforces the distinction between letter (epistula) and book (liber) as two differently coded literary forms.

With certain variations of language, the trope of the author’s envoi appeared at least two more times in Martial: i nostro comes, i, libelle, Flavo […] vale, libelle (10.104.1, 18) and, from the final ‘exile’ in Bilbilis, ibis, io, Romam nunc peregrine liber (12.2). On the whole, the original examples remain representative. The dynamics of movement in and out of public and private space invoked by the authorial envois were made possible by of the reality of imperial space and the imagined global networks that came with it. But each author responded to these opportunities differently: Ovid wishes to be his book and to go back to Rome while Martial escapes the city that confounded him, relishing the disembodied access he still has to an audience of readers at Rome and abroad by way of the book trade.

***

Having thus emerged in the context of imperial and bibliographic expansion in the first century BCE and CE, the vade, liber trope was taken up throughout the literature of imperial Latin, appearing again at Statius’s Silvae 4.4 curre … epistola (‘run, letter’ 1), Ausonius’s Eclogues 1 volate, versus (‘fly, poems’, 14) and Epistles 12 Perge, o libelle (‘Go forth, little book’, 1), and Sidonius Apollinaris’s Carmina 3 i, liber (‘go, book’, 9) and 24 Egressus foribus meis, libelle, | hanc servare viam, precor, memento (‘When you pass out by my door, little book, pray remember to keep this route’, 1–2).Footnote ⁴³ In (probably) the second century CE we find it for the first time in Greek in an epigram in Strato’s Musa puerilis, closely following the Horatian model:

In the early Middle Ages, the Ovidian model was followed by the likes of Alcuin of York, in Carolingian France, and Eugenius of Toledo, in Gothic Spain.Footnote ⁴⁴ Even more, in the emergent literature of the Romance vernaculars in the Middle Ages and the Renaissance, the vade, liber formula would become a commonplace, appearing perhaps most iconically in the epilogue of Geoffrey Chaucer’s classically themed Troilus and Criseyde. Here, the poet knowingly translates Ovid’s words, sending his ‘litel book’ to pay homage to a parade of great ancient authors before it goes to the living dedicatees ‘moral Gower’ and ‘philosophical Strode’ (5.1856–7) for corrections: Footnote ⁴⁵

According to John Tatlock, Ovid’s special status as a poet of love meant that his refrain was taken up with particular interest in courtly song in Provençal and, by its influence, Old French and Italian, including lyrics of Guido Cavalcanti, Dante Alighieri, Francesco Petrarch, and Giovanni Boccaccio.Footnote ⁴⁷ Because it was so easily repeated and repurposed, and because it so succinctly captures the inherent tension in writing that has been known since Plato that to make one’s works known is to let them go, as a text, vade, liber itself travelled, as an incipient meme in the classical tradition and an emblem for the mobility of both authors and books.Footnote ⁴⁸

***

Even more than the aesthetics of intertextuality, the book’s send-off offered Roman authors particular opportunities to map places and networks within the city and beyond. If we return to the third of four characteristics of the motif outlined previously – that is, questions of relative space and time – we can see the particular ways in which imperial geography were originally implicated. Horace, in Epistles 1.20, points not only to the location of the bookshop of the Sosii Brothers on the Vicus Tuscus but also to the provincial centres of Ilerda and Utica, in Spain and Tunisia, both newly charged as sites of important Caesarian victories in the First Civil War. Ovid, in the Tristia, has his book roll enter Rome as a tourist, enacting his journey of exile in return. And Martial, variously, emphasises the distance between author and addressee not only as a barrier to reunion, but as an occasion for poetry itself: whether that is the distance between two neighbourhoods in Rome, as in 1.70 and 1.108, or, in 3.4, the distance between the capital and Martial’s temporary place of retirement in Forum Cornelii. So too do 10.104 and 12.2 similarly explore the dynamics of imperial space, with 10.104 describing the book’s trip to Hispania Tarraconensis (i, libelle, … longum per mare, 10.104.1–2) and 12.2 enacting that journey in reverse, sent from Martial’s home in Spain on yet another journey of the book to Rome (ibis io Romam … auriferi de gente Tagi tetricique Salonis, 12.2.1–2). In the context of Rome’s growing dominion in Italy and abroad, for readers not in Rome – or Imola or Bilbilis or Verona – the pleasurable fictions of these poems relied on the mobility of texts.

In Latin, the idea of mobilitas is associated with ‘quickness’, in both our senses of that word in English: the capacity for movement and rapidity, but also the quality of being alive. No stranger to the idea of treating texts as people, in a discussion of the heavenly bodies in the dialogue De natura deorum (2.42), Cicero describes the Stoic view that stars are living creatures born amidst the most rarefied and vital part of the sky, imbued with ‘the keenest senses’ (sensu acerrimo) and the ‘swiftest mobility’ (mobilitate celerrima):

It is for this reason, he goes on to say, that they must be thought to possess sensation and intelligence and to be counted among the gods. The ability to move is an external sign of their psychic animation: literally, the quality of having an anima and thus of being an animal.

In my readings of Roman poetry, I have described the vade, liber refrain as a kind of an emblem for the reality that authors, books, and texts all travel. And like all literary apostrophes, the authorial envois I have highlighted are a reminder that there is not a fast distinction between the quick and the dead, the animate and the inanimate, the present and the absent, the near and the far.Footnote ⁴⁹ More than simple apostrophe, however, the personification of books speaks to the inherent paradox that inanimate books carry the traces of their animate authors and, even more, that writing itself has the transcendental ability to transform our animate states of mind.

Books are the instruments and the sites of those transformations and those entanglements. As things that exist on geographical and chronological scales much greater than human, able to survive for centuries if not millennia and mobile in ways that people are not, books are, perhaps, even more than objects – or what Timothy Morton, thinking in ecological terms, has called ‘hyperobjects’, that is, things that are vastly distributed in time and space, beyond the measure of the human experience and the gauge of the human gaze.Footnote ⁵⁰

We are accustomed to thinking about the material instantiations of text as instilling fixity. Once the text is copied, once the book is printed, it is thought to undergo changes in what traditional bibliography would count as ‘accidental’ and not ‘substantive’ ways.Footnote ⁵¹ Marginalia, cutting and pasting, copying and excerpting, quotation and reworking, are imagined to be intrusions into the notional ‘end state’ of the published work. And they can be seen as aspects of the mobility of texts in two dimensions: made possible by the fact that texts, as objects, do travel independently of their makers; moreover, that readers can do whatever they want to texts because they can do whatever they want with books. The old adage verba volant, scripta manent – ‘Words fly; writing remains’ – is turned on its head. By considering as fundamental to the dynamics of books their physical and metaphysical mobility, we see that fixity is the very thing that they defy: they are movers. But ultimately, like Dickinson’s frigate, what is most remarkable about the spaces of books is that to enter into them, we have but to remain where we are.