Travel in the Roman World

I see you know that I arrived in Tusculum on November 14th. Dionysius was there waiting for me. I want to be in Rome on the 18th. I say “want,” but really I have to be there—it’s Milo’s wedding.

(Cic. ad Att. 4.13.1 = no. 87 SB; trans. Shackleton Bailey)

So Cicero writes to Atticus in the middle of November 55 BCE, asking for an update on the political situation in the capital. His tongue-in-cheek tone—he has (cogimur) to be in town for a wedding—belies both his urgent request for information and the complex network of roads and technologies of travel on which his upcoming trip depends. Neither Cicero’s letter nor his mini-break away from the city are out of place in his correspondence from the mid-50s. On Shackleton Bailey’s dating, twenty-nine letters survive from the years 56 and 55; nearly half were written outside Rome. We learn from those letters that Cicero traveled to Tusculum, Cumae, Napes, Pompeii, and Antium. In other years he ventured even farther afield. When he was reluctantly made governor of Cilicia in 51, he recounted his journey there in a series of plaintive missives to Atticus, griping, for instance, about “waiting for a passage” after being held up for twelve days in Brundisium by an “indisposition” (ad Att. 5.8.1 = 101 SB; trans. Shackleton Bailey). The only thing worse than waiting for a ship was being a passenger on one; a month later he grumbles from Delos: “Travelling by sea is no light matter, even in July…. You know these Rhodian open ships—nothing makes heavier weather” (ad Att. 5.12.1 = 105 SB; trans. Shackleton Bailey).

As detailed as Cicero’s correspondence is, it reveals just one individual’s experience of travel in the late Republic. For all the frustrations of his journey to Cilicia or the panic of his flight into exile—narrated in another series of letters from 58 (ad Att. 3.1–8 = 46–53 SB)—he had much less to complain about than those who died in shipwrecks (e.g., CIL III 1899 = CLE 00826 = ILS 8516 [Blaška Voda; 201–300 CE], CIL XIII 2718 [Autun; second or third century CE]) or in pirate attacks (e.g., AE 1982.512 [La Muela; 71–130 CE]; CIL III 1559 = III 8009 [Slatina; 118–138 CE]). Indeed, Cicero can hardly be said to have lived an average life, even for an elite Roman; nevertheless, his Letters show how deeply travel was woven into the texture of the ancient world and how many aspects of the Roman experience relate to it. Rather than pretend to total coverage of the topic, in this chapter I offer some ways of thinking about travel and its place in the Roman world, exploring the practice, ideology, and imagination of travel from many different directions. Moving from a top-down perspective on Roman infrastructure and the role of travel in the practice of Roman power, I turn to a bottom-up view of the experiences of individuals traveling, concluding with a survey of the imaginative representation of travel in the literature of the Roman world.

Cicero might complain about Rhodian ships, but his Letters take for granted the roads, harbors, and organization that make his travels possible. The effect that this infrastructure had on the ability of information, people, and goods to travel through the Roman world might best be described in terms of the concept of “connectivity.” Originating in the fields of graph theory and network analysis, connectivity expresses the relationship between places in terms of the costs and time required to move between them rather than their geographical distance. The wide-spread adoption of the term in the study of the ancient Mediterranean is due in large part to the pioneering work of Horden and Purcell (2000, 123–72, 562–71; many responses by, e.g., Harris 2005 and Knapp 2008; see further Pettegrew 2013). Connectivity also informs Walter Scheidel’s and Elijah Meeks’s Stanford ORBIS project (http://orbis.stanford.edu), which uses interactive geospatial modeling to plot land, sea, and fluvial routes between major centers in the Mediterranean in terms of the average temporal and monetary costs for each route. By expressing travel as a function of connectivity, the ORBIS model redraws our map of the Roman world and highlights the effects that infrastructure has on shaping its accessibility (Scheidel 2014). Physical infrastructure is just one aspect of a region’s connectivity; in order to be used, it has to be known to and understood by individual travelers. A second aspect of infrastructure is what might be called its “legibility.” This concept, which originates in Urban Studies, applies the metaphor of reading to the experience of navigating a city or, in this case, the world (see, e.g., Lynch 1960, 2–3). Linking these two approaches to understanding space, I construe connectivity more broadly to include all the intellectual and cultural processes involved in making infrastructure legible and comprehensible to travelers. While such processes of collecting, ordering, mapping, and communicating knowledge to the Roman traveler are harder to capture through mathematical models, such as ORBIS, they are no less important for understanding how the world is connected.

Roman connectivity was enabled in the first instance by a highly sophisticated and extensive transport network that consisted of roads as well as maritime and fluvial routes. Roads, which are still visible throughout much of the Roman world, could be traversed on foot; on the backs of horses, donkeys, or camels; and by a variety of vehicles powered by human beings and draught animals (Chevallier 1976, 178–81; Bagnall 1985; Laurence 1999, esp. 78–94; Kolb 2000, 308–32; Adams 2001; Matthews 2006; Adams 2007; Adams 2012, 229–31; Bekker-Nielsen 2013; Owens 2012; see also Land Transport, Part 1 and Land Transport, Part 2). Built by Appius Claudius Caecus in 312 BCE (cf. CIL XI 1827 = ILS 54 [Arezzo; copy of CIL VI 31606 from the Forum Augusti]) the Via Appia, which Statius called the “queen of long roads” (Silvae 2.2.11), was the first long-distance link in the Roman transportation network. Due to the Roman practice of extramural burial, the Via Appia, like many roads, was lined with tombs; visitors to Rome, for instance, would be greeted first and foremost by its dead (see Coarelli 2014, 365–400). As Rome’s power grew, so did its road system; Romans either constructed new roads, as they did in many of the western provinces, or improved upon already existing ones, as in the eastern provinces. Roads’ importance to Roman travel lay in their support of Roman troop movements, in making inland territory easily legible and accessible in the wider network of the Roman state, and in enabling economic activity.

Travel by road was assisted by a secondary infrastructure of private hostels and state-sponsored stopping places to house emperors and Roman officials (Lat. mansio), which developed during the Roman imperial period, gradually replacing the older practice of requisitioning room and board for official travels (Casson 1974, 197–218; Mitchell 1976; Millar 1977, 32–33; Adams 2001, 142–45). Information, goods, and services could also be transported along Roman roads by the cursus publicus, an express imperial messenger service founded by Augustus (Ramsay 1925; Stoffel 1994; Kolb 2000). Although there is far less archaeological evidence for roads in Egypt, there too land transport played a central role in the movement of people and goods throughout the province. In addition to archaeologically attested routes that connected the Nile Valley to the Western Desert and the Red Sea coast, Colin Adams collects evidence for a wider network of more temporary roads closer to the Nile, long since destroyed by the river’s annual flooding, which ran from the Mediterranean coast to as far as Syene (Western Desert—Darnell 2002; cf. Adams 2007, 30–33; Western Oases; Red Sea—Sidebotham and Wendrich 2001–2002; Gates 2005; Adams 2007, 33–42; Sidebotham 2011, 125–74; cf. Strabo 17.1.45; Pliny NH 6.33.165; IGR I 1142 = OGIS 701 = SEG 51 2123 [Antinooupolis, modern Sheikh ‘Ibāda; 137 CE]; The Eastern Desert and the Red Sea Ports; Nile—Adams 2007, 22–29; cf. P.Oxy. LX 4087, 4088 [first half of the fourth century CE] on way stations in Tacona and Oxyrhynchus).

Roads were by far the costliest means of transporting goods and traveling; according to calculations made by applying the ORBIS model to data from Diocletian’s Edict on Maximum Prices of 301 CE, transportation by wagon cost between five and fifty-two times more than travel by boat for equivalent distances (Scheidel 2014, 9–10). Although the Edict has proven notoriously difficult to interpret, Scheidel’s predictions about the extremely high costs of travel by road are broadly consistent with data from early eighteenth-century England, where transport by land was between five and twenty-three times more expensive than that by boat (Scheidel 2014, 9–10). According to these calculations, land transport would have been so costly as to be prohibitively expensive for moving many inexpensive, bulky goods, yet archaeological and documentary evidence attests to roads as an important mode of transport in the Roman world. In response to this apparent paradox, Adams (2001, 2007, 2012) has suggested that travelers and traders could employ alternative strategies that undercut the prices implied by the Edict. For instance, farmers might transport goods (and people) by using their own animals or rent them out less expensively when they were not needed elsewhere. Even these mitigatory strategies had their limitations, however, and they would not be effective for long-distance travel and transport. For these journeys, boats were an indispensible link.

While roads have left the most enduring mark on the archaeological record, maritime and fluvial routes were equally if not more important for the connectivity of the Roman world. Mariners traversed both short coastal journeys and longer sea lanes (Arnaud 1992, 2005; Sea Transport, Part 1). Sea travel was highly seasonal, and it is therefore likely that temporal and monetary costs varied throughout the year (cf. Casson 1995; Veg. Mil. 4.39; Cod. Theod. 13.9.3), but despite the variability in speed, cost, and route, maritime travel was on average the most efficient mode of transportation for goods and people available in antiquity, especially over long distances (as calculated by ORBIS “Sea Transport” according to wind speed; for historically reported sailing times, see Ramsay 1904, 379; Braudel 1995, 358–63; Casson 1995, 282–88; Arnaud 2005, 102; on the cost of maritime transport, see Arnaud 2007 and Scheidel 2013, who argue that freight charges in Edict on Maximum Prices correlate closely to average sailing time).

At the interface between land and sea was an extensive network of ports and harbors. Roman quays, which usually reached to the shore, were ordinarily backed by large warehouses, most typically a horrea for grain storage. Their basic form has long been well known from Vitruvius (De arch. 5.12; cf. Cass. Dio 40.11.1–5), as well as from visual depictions, which are especially well represented at Herculaneum and Pompeii. In addition, Roman harbors are an active area of research; an increasing number of these sites have been excavated thanks to advances in underwater archaeology, which has added substantially to our knowledge about the details of their construction (see further the gazetteer of sites in Sea Transport, Part 2).

In addition to sea lanes, many larger rivers were also navigable, but there is no comprehensive treatment of fluvial routes in the Roman world (the best overview is by Campbell 2012, esp. 200–245). In Egypt, where documentary papyri offer a fuller picture of economic activity, the Nile was a particularly important route (Adams 2001; Cooper 2011). Contracts cover every detail of the journey, mandating at which harbors the captain could moor (only the safest), the hours during which he could sail (the proper ones, presumably only during the day), and what to do in case of foul weather (see, e.g., P.Hib. II 198 (third century BCE), with Bagnall 1979; P.Ross.Georg. II 18 [140 CE]). These fluvial cargo ships also carried passengers: a letter from 7 BCE, for instance, records one Athenodorus’s efforts to secure passage for his sister (BGU XVI 2604). Even if a captain managed to land in the right harbors, sail at the right time, and avoid foul weather and pirates, travel by river was not always smooth sailing. On the Nile, the swift rapids of the cataracts required particular care and navigational skill. Ships had to be offloaded, and their passengers and cargo were transported overland to the east of the river, while the boat was guided down with cables from the shore (cf. Pomp. 1.9.51; Jaritz and Rodziewicz 1993; Between Egypt and Meroitic Nubia: The Southern Frontier Region). Such challenging waters also invited daredevils. According to several sources, some locals took advantage of the swift currents at the first cataract to put on displays of their water rafting prowess (Strabo 17.1.19; Seneca Nat. 4A2.6–7). Visitors were also known to try their hands at the rapids; the Greek intellectual Aelius Aristides, for instance, proudly claimed to have taken part in such extreme watersports (Aristid. Or. 36.47–51 [Egyptian Discourse]).

Elsewhere in the Roman world, the role of the Tiber in transporting bulk goods to Rome, especially grain, is particularly well attested (cf. Strabo 5.2.5; see further Tuck 2013, 237–44). After the fire of Rome, Nero used grain ships to transport debris away from the city to the marshes of Ostia (Tac. Ann. 15.43). Bridges built over the river presumably restricted the size of vessels that could reach the capital, but large, wide-bodied ships could still travel upriver to Rome, as depicted in a variety of reliefs, including the so-called Tiber sarcophagus and sections of Trajan’s column (Tuck 2013, 238). As a result of all this mercantile activity, the river was often thick with smaller boats (Lat. lintres, lembi, lenuculi, and lusoriae), which helped to guide larger ships and to ferry goods to shore. This activity was such a central feature of Roman life that it was reflected by Propertius, who used the Tiber’s congestion as a metaphor for his love for Cynthia in his first books of poems (1.14.3-4). A century or so later, Pliny the Elder was more direct, calling the river the “merchant of all things produced in the whole world” (HN 3.54). Together with the Nile, the Tiber attests to the importance of rivers for transporting goods and people in the Roman world.

In order to be properly used, the physical infrastructure for travel in the Roman world had to be made legible to Roman officials, merchants, and private travelers. In this section I examine some of the means through which this legibility was achieved: namely, labeling roads, collecting route itineraries, describing journeys, and making maps. These processes of ordering, collecting, and presenting information in textual or graphic form were at once functional and ideologically charged. Although this chapter’s focus is on the organization of knowledge related to travel, the intellectual and cultural processes involved can be understood as related to the practices of ordering and presenting other categories of knowledge throughout the Roman empire (see König and Whitmarsh 2007, esp. 3–39; and König, Oikonomopoulou, and Woolf 2013 on libraries).

Roads were regularly marked with milestones (Lat. miliarium), which recorded the distance from each terminus and the name of the magistrate or emperor who supervised their construction (Chevallier 1976, 39–34; Kolb 2004; cf. CIL I² 21 [Pontinia; 253 BCE]). In this way, roads themselves expressed their place in the wider Roman transportation network. At least some roads also displayed longer and more detailed epigraphic itineraries, such as the fragmentary example from Augustodunum listing routes in northern Italy and southern Gaul (CIL XIII 2681a, b [= ILS 5838], c = CIL XVII² 490a–c [before 200 CE]; see also Chevallier 1976, 52–53, and below). These markers served twin functions, simultaneously orienting travelers and expressing space as Roman; the process of building a road and linking it to the road network of the empire represented a significant centralized intervention in the landscape. The ideological function implicit in all milestones takes precedence in the Miliarium Aureum (Golden Milestone), which Augustus set up in the Roman forum in 20 BCE. By listing the principal cities of the empire and the distances to them as measured from the Severan Wall, the monumental mile marker both established the city as the center of the vast imperial road system and commemorated Augustus’s role in overseeing the Roman transporation network (Pliny HN 3.66; Suet. Otho 6.2; Cass. Dio 54.8.4; LTUR 3 [1996], 250–51; Brodersen 1995, 254–60). In this sense, the Miliarium Aureum can be seen as an expression of the extent of Roman power and the connection between Roman roads and Roman control. Similarly, the Stadiasmus Provinciae Lyciae, a large monumental pillar from Patara, Lycia, dedicated to the emperor Claudius in 45/46 CE, records distances along Lycian roads (SEG 44 1205 = SEG 60 1550; first publication: Şahın 1994; editio princeps: Işık, İşkan, and Çevik 1998/1999). The Stadiasmus, which was originally topped by an architrave and probably an imperial statue, provides a further link between the measurement of the Roman world and ideological expressions of Roman power and administration. Given that the Stadiamus was completed only a few years after the annexation of Lycia in 43 CE, it is likely that it did not reflect Roman improvements to the road network in Lycia, but rather that it reappropriated Lycian roads as Roman ones. Both the Stadiasmus and the Golden Milestone, therefore, serve an ideological function of expressing imperial control through infrastructures of knowledge, even if they only functioned as signposts in “a very extended sense” (Jones 2001, 168; see further Lebreton 2010, esp. 67–76).

In addition to epigraphic routes, ancient geographic knowledge was also expressed through the rich tradition of textual itineraries (see Elter 1908; Kubitschek 1916; Dilke 1985, 112–29; Elsner 2001; Salway 2001, 2012). These texts make few pretenses to literature—they are usually little more than lists of places and the distances between them, although sometimes they added more colorful material directed at the tourist, pilgrim, or merchant—but they represent one way of making Rome’s network of land and sea routes legible and functional for their readers (thus Veg. Mil. 3.6 recommends using interaria for traveling; see further Adams 2001; Brodersen 2001). Some, such as the Itinerarium Burdigalense (see Elsner 2001), recorded actual routes; others are almost certainly compilations, such as the third-century CE Antonine Itineraries, which begin with the Pillars of Hercules and continue counterclockwise around the Roman empire. The sources from which they were compiled are unknown, but they may derive from epigraphic itineraries, such as the Augustodunum fragment discussed above. As functional as these texts seem, they also express a viewpoint about the world, its connections, and its center(s). In his analysis of the Itinerarium Burdigalense, for instance, Jaś Elsner shows that it was “the first Roman Christian text to present Jerusalem as the centre of the world” (2001, 195). The recognition that itineraries reflect ideologies is particularly significant because of their wide influence on historical and geographic works, such as Strabo’s Geography and Pliny’s Natural History (see Clarke 1999, 197–210); larger monumental representations of space, such as the so-called Map of Agrippa; and more personal objects, such as the Dura Europos shield, a third-century CE parchment fragment, which graphically represents cities on the coast of the Black Sea and their distances (Cumont 1925, 1–15, 1926, 323–37; Dilke 1985, 121–23). Itineraries and their ideologies lay at the heart of Roman conceptualizations of their world.

Itineraries are only capable of encoding a single journey at a time. Maps, which graphically represent physical geography and routes in a two-dimensional space, have a greater density of information, and a single map could, in principle, express every possible path through the Roman world. Strict cartographic constructionists, such as Brodersen (1995, 2001), take a consistent concept of scale as a key element of mapping, and they emphasize the differences between modern maps and Greek and Roman representations of geographical space. If we allow for a slightly broader definition of a map, which places more emphasis on perspective than on adherence to a consistent scale, however, it is possible to identify at least two distinct Roman cartographic genres: monumental maps, whose function is closely linked with their context, and smaller maps, usually on papyrus. Like itineraries, even these apparently very functional objects express strong views about the shape of the world and the meaning of travel through it.

Although there is evidence for several monumental maps in the Roman world (see Talbert 2012), only two survive, and both are fragmentary: the Forma Urbis Romae, an early third-century CE marble plan of the city of Rome, originally displayed in the Temple of Peace (extant fragments available in an online edition: http://formaurbis.stanford.edu), and the cadastral plan of Roman centuriation, which was executed in 77 CE in Arausio (modern Orange) under the auspices of Vespasian (edict of Vespasian: AE 1952.44 = AE 1963.197 = AE 1999.1023; see further Piganiol 1962; Dilke 1973; Jung 2009). While the cadastral plan of Arausio imposes Roman order on territory appropriated by the state, it seems unlikely that the Forma Urbis Romae served a similarly practical purpose. In its original context, it was mounted more than four meters above the ground on an interior wall of the Temple of Peace; it would have been impossible for a viewer to get close enough to appreciate its detailed depiction of the layout of the city of Rome (Trimble 2007, 369–70). Even though much of it would have been illegible to an ancient viewer, the marble plan served an important ideological purpose by monumentalizing the meticulously ordered cartographic knowledge that it represented (see further Trimble 2007). If the connection between cartographic representation and state control is implicit in the Orange Cadasters and the Forma Urbis Romae, both of which depict space from the top-down perspective of Roman power (see further Nicolet 1991, 95–122; Laurence 1998), it appears to have been explicit in another monument to imperial space, which does not survive, Agrippa’s “map” of the world. According to Pliny the Elder, Agrippa commissioned a “representation of the whole world for the city to look upon” (HN 3.17) in the Porticus Vipsania, which Augustus completed after Agrippa’s death in 12 BCE (Plin. HN 3.17; Cass. Dio 55.8.3–4; Arnaud 2007–2008; Talbert 2010, 136–37, 2012; see further GLM, 1–8 for testimony and LTUR 4 [1999], 151–53 on the Porticus Vipsania). The details of the map remain murky, but it is clear that whether it was a graphical representation of the world or a monumental inscription, like Augustus’s own Res Gestae, it connected the traversibility of the Roman empire with the princeps and his power.

These monumental maps were highly localized and tied to specific archaeological contexts; neither could have aided the traveler on the road. A second genre of maps, which either survive on papyrus or are shaped to resemble a papyrus roll, suggests the possibility of more portable maps that contained detailed route information. Two such maps are extant, although both seem too luxurious to have been used as a road atlas. An incomplete map, produced in the first century BCE or CE, appears on the front of the Artemidorus Papyrus, alongside a geographical text and highly skilled sketches of human faces, feet, and hands (MP³ 168.02; editio princeps: Gallazzi, Kramer, and Settis 2008; the reordering of the fragments first proposed by Nisbet 2009 and supported by D’Alessio 2009, 36–41 has now been accepted by most scholars). A second, the so-called Peutinger Map, a medieval copy of what is likely a fourth-century CE original, survives as eleven parchment segments, which can be joined together to form a long and thin map that closely corresponds to the dimensions of a papyrus roll (cf. Talbert 2010, 143; average dimensions of papyrus rolls are calculated by Johnson 2004). Although the Peutinger’s routes, distances, and place names show a clear dependence on itineraries—in some cases the map retains a place name’s oblique case, as if copying directly from an itinerary—this information was either out of date, inaccurate, or otherwise insufficient to be of practical value to the traveler in the fourth century (Talbert 2010, 108–17).

While neither of these maps seems likely to have been used to enable travel, both appear to share a set of cartographic conventions for depicting settlements, waterways, and land routes, and both seem to have distorted their dimensions in order to fit within the constraints of their medium (see Gallazzi, Kramer, and Settis 2008, 275–308). The maps’ shared representational language suggests that there was a genre of cartography on papyrus that potentially could have extended to portable guides for travelers, even though neither the Artemidorus Papyrus nor the Peutinger Map was likely to have been used for this purpose. What, then, were these two maps used for? It is possible that both of these deluxe maps, like the Forma Urbis Romae, represented space and routes purely for intellectual and/or ideological purposes. For instance, Richard Talbert has persuasively argued that the Peutinger map does much more than represent land routes; in particular, he shows that the outlines of the land were drawn before roads, that Rome is positioned to occupy the map’s center, and that the map uses a variable scale that focuses attention on Italy (Talbert 2010, 2012). All these details suggest that the map’s composition does more than simply graphically represent information from itineraries. Rather, it reflects Romans’ knowledge about the connectedness of their world and how they imagined the shape of their empire.

In this section I turn from the physical and intellectual infrastructure of the Roman world to show how travel was integral to the construction, maintenance, and regulation of Roman power in both political and economic terms. Rome sat at the notional and, roughly, the geographic center of an enormous empire, which at its height occupied almost the entirety of the oikoumenē (inhabitable world). In both practical and ideological terms, the physical and intellectual infrastructure I have described in the preceding section defined the extent of the Roman world and enabled Rome to exercise centralized control over it: by transforming the landscape of the Mediterranean, it made even the farthest reaches of the empire accessible to agents of the state, to the army, and to individuals for personal and economic purposes; it marked that landscape as Roman through buildings and inscriptions; and by communicating and disseminating such information, it enabled Romans to make use of this infrastructure to move through their world.

A high proportion of travel in the Roman world can be attributed to either military or commercial interests; the needs of the Roman army and the desire for economic opportunities contributed to motivating and financing the infrastructure on which other forms of travel depended (see further Roman Army and Roman Economy). The Roman road system, for instance, was essential to the movements of the army, whose maritime activities appear to have been limited to the transportation of supplies (Roth 1999, 189–95; see Bremer 2001 on the use of rivers to supply military camps). It is often argued that roads were constructed primarily to support military activities, and that commercial activities were only an indirect benefit (Schlippschuh 1974, 87; Kissel 1995, 54–77; Roth 1999, 214–17). While this may be an overstatement, it is clear that the construction of roads contributed to Rome’s economic and military hegemony throughout the Mediterranean in several ways (Laurence 1999, 11–26). As Ray Laurence’s analysis of the construction of the Via Appia shows, not only did the road serve a strategic function against the Samnites, but as a large public works project, it employed as laborers Roman citizens and their allies (Lopes Penga 1950; Coarelli 1988, 42; see Humm 1996 and Laurence 1999, 15–16, on the construction of the Via Appia).

The same infrastructure that enabled the movement of the Roman army also enabled travel for economic purposes. Literary and documentary sources, especially papyri from Egypt; goods recovered from ancient shipwrecks; and the distribution of finds of specialized products such as Italian wine and snails attest to the quantity and variety of merchandise exchanged throughout the Roman world (Scheidel 2011a; Wilson 2011a, 2011b, 2012, 287–88; see Tchernia 1983, 1986; Panella and Tchernia 1994 on Italian wines; Marzano 2011 discusses the winter snail trade in Roman Egypt; traders themselves often formed associations; see P.Mich. 5 245 [47 CE], CIL III 14165 [Beirut; third century CE], Dig. 50.6.5.3–6, discussed by Meijer and van Nijf 1992, 75–78). The degree to which trade was central to the Roman economy, however, remains a subject of debate (Scheidel 2012a, 7–10, offers a concise summary). Economic formalists argue that the Roman Empire was governed by a market-based economy that integrated local economies on an empire-wide scale (argued for in detail by Temin 2013; Temin 2012 provides an overview; see Erdkamp 2013 and Manning 2014 for important limitations in our evidence for ancient economic activity). Economic substantivists, on the other hand, focus on the role of the state government in price-setting, requisitioning, and distributing goods; in Keith Hopkins’s “tax and trade model,” governmental and elite demands for taxes and rent drove market forces, creating reciprocal flows of taxable resources (1980, 1995/1996, 2009; Silver 2008). While these models differ in the degree to which the wider Roman market and, therefore, trade and infrastructure were important for influencing local prices and production, both models predict significant amounts of commerce, which would entail corresponding travel by merchants (for examples, see Meijer and van Nijf 1992).

Private individuals also traveled for education and employment, as the fourth-century CE curriculum vitae of a certain Conon discovered in modern Ayasofya in Eastern Pamphylia attests: he studied Latin and Roman law in Beirut, got his first job in Palestine, then moved to Antioch and later to Nicomedia; he died while abroad in Egypt, and his father retrieved his body (SEG 26.1456 = 57.2120; Bean and Mitford 1970, no. 49; see Gilliam 1974 on the date). Conon spent his life on the move; travel had penetrated the conditions and values by which he lived his life.

Slaves represent a final group of individuals whose travel can be attributed to the exercise of Roman economic and political power. It is a sobering fact that they likely represent the largest group of economic travelers. A central principle of slavery is the abnegation of basic elements of human agency, such as freedom of movement. Because many slaves were captured in war or by pirates, the sale of slaves often involved their transport (on the “Roman slave supply,” see Bradley 1994, 31–56; Schumacher 2001, 34–43; Scheidel 2005, 2011b; see Matthews 1984 on the importation of slaves to Palmyra). When slaves traveled, it was almost always at the behest of their owners (cf. Bradley 2000, 118). It is very difficult to estimate the ratio of free persons to slaves in the Roman world, but it is safe to say that at many times and in many places in Rome’s history it was a slave society, and that the slave trade was a significant part of the Roman economy (for an overview of Roman slavery, see Bradley, Freedom and Slavery;  Hopkins 1978; Bradley 1994; Schumacher 2001; Scheidel 2012b).

Travel by officeholders and the emperor himself helped to articulate the contours of Roman power, playing an important role in governing and maintaining a geographically, culturally, and ethnically diverse territory. Because movement by Roman state officials was only experienced by those on the road, profectio (departure) and adventus (arrival) rituals played a central role in making their travel manifest to a wider public. Taken together, these rites constitute a form of Roman spatial religion. Making entrance into and exit from a city at crucially marked moments, they made the populace of each city fellow participants in officials’ movements. Both adventus and profectio rituals were highly choreographed, transactional performances between the governing and the governed. Through ritual practice the conceptual geography of the Roman world was temporarily inverted: the center was invited to imagine travel to the periphery, and the periphery was made into a temporary center. Like many formal characteristics of Roman religious and civic life, rites of arrival and departure had their origins in the Roman republic and were continued in the Roman imperial period, but the use of rituals to demarcate official movement is in fact part of a much wider Mediterranean phenomenon. They were an important feature of Seleucid kingship (Kosmin 2014, 142–80) and of their Achaemenid predecessors (Briant 1988; Tuplin 1998).

Adventus rituals are best attested for the city of Rome itself, but they were equally significant outside of the capital (Lehnen 1997; see also Brilliant 1963, 173–77; Koeppel 1969; Alföldi 1970; Castritius 1971; MacCormack 1971, 17–61, 1972; Millar 1977, 28–30; Vitiello 2000; Roddy 2001). Like greeting (apantēsis) ceremonies for Hellenistic monarchs, Roman adventus rituals consisted of two distinct movements. Dignitaries were first met extramurally by crowds, speeches, and acclamations (Dion. Hal. Ant. Rom. 2.34.2; Cic. Mur. 68 [L. Murena], Sest. 68 [Cicero], ad Att. 4.1.5 = 73 SB [Cicero]; Livy 5.23.4 [Camillus], 27.50–51 [L. Nero], 45.35.3 [Paulus]; App. BC 1.33 [P. Furius]; Dio 51.19.2 [Augustus]; Suet. Cal. 4, 13 [Caligula]; Tac. Ann.14.13.2–3 [Nero]; Joseph. BJ 7.68–74 [Vespasian]; see further Pearce 1970; Lehnen 1997, 105–56; Kuhn 2012, esp. 308 n58). Outside the walls an ordered citizenry presented itself to welcome and be viewed; ancient accounts suggest that the crowds were arranged by age, gender, and social status (cf. Livy 5.23.2; Menander Rhetor fr. 381 [Russell and Wilson]). Cicero boasts, presumably not without hyperbole, that he received an extreme version of this treatment on his return from exile: “Almost all of Italy carried me back on its shoulders” (Cic. Red. sen. 39).

In the second movement, the city gate was opened; once inside, dignitaries could expect to be greeted by a large crowd and a second set of acclamations and festivities (gate—Lehnen 1997, 167–69; Suet. Nero 25.1; crowds—Lehnen 1997, 156; Cass. Dio 63.4.1; Pan. Lat. 7[6].8.7; acclamations and festivities—RAC I 1950 s.v. “Akklamation” [Th. Klauser]; Lehnen 1997, 169; Herod. Hist. 1.7.6, Amm. Marc. 16.10.9; Cass. Dio 63.20.5–6; Hist. Aug. SA 57.5). The city itself was decorated for the occasion: a procession carried branches, usually of laurel and palm; torches; incense; and statues of local gods (laurel and palm—Lehnen 1997, 121–22; Plin. HN 15.137; Mart. 8.65; Cass. Dio 63.4.2; torches—Alföldi 1970, 113; Lehnen 1997, 122–23; Tert. Apol. 35.4; Amm. Marc. 21.10.1; Pan. Lat. 2[12].37.1–4; incense—Lehnen 1997, 120; Martial 8.15.3; Pan. Lat. 11[3].10.5, 10[2].6.4; Cass. Dio 63.4.3; statues of local gods—MacCormack 1972, 723; Josephus BJ 7.68–74). If the first movement of the adventus ritual presented a welcoming human presence, the second movement was significant for its spatial dynamics. The city was laid open and presented to its visitors. The arriviste also had a part to play; coinage from the second century CE shows a stereotypical adventus gesture, a raised right hand in a “gesture of benevolent greeting and power” (Brilliant 1963, 173, esp. figs. 4.24 [AU of Elagabalus], 4.26 [AR of Gordian III], and 4.28 [AR of Septimius Serverus]). Failure to welcome with sufficient pomp and circumstance might be taken as an offense (Livy 42.1.7).

Implicit in the process of arrival is a complex social, cultural, and religious negotiation between host and visitor; in exchange for his presence, the local community bestows on the visitor honors, and in some cases treats his arrival as the epiphany of a deity. The language of “visiting” (epidēmia) and “presence” (parousia), frequently used in Greek inscriptions to describe imperial visits, can refer as easily to divinities as monarchs (epidēmia—e.g., Xen. Eph. 1.2.1, Panamara 176; parousia—e.g., Diod. Sic. 3.66.2, IG IV², 1 122, line 34 [Epidaurus; fourth century BCE]; see further Pelling 1988, 179; Koenen 1993, 65; Burkert 2004, 16; Platt 2011, 142–43). Mark Antony’s visit to Ephesus in 41 BCE provides a particularly baroque illustration of both the thin line between arrival and epiphany as well as the transactional nature of an adventus ritual. According to Plutarch, the Roman general was met by “women dressed as Bacchants, men and boys dressed like Satyrs and Pans,” who “led the way into the city, which was full of ivy, thyrsus wands, harps, pipes, and flutes” (Plut. Ant. 24.3). Such an adventus could only be for Antony; the Ephesian ceremony, which greeted his arrival as if it were an epiphany of Dionysus, presciently responded to what would become an important feature of Antony’s Dionsyiac self-presentation just a few years later, when in 39/38 BCE he was celebrated in Athens as “The New Divine Dionysus” (Theos Neos Dionusos; IG II² 1043 lines 22–23 = SEG 130 [Athens; 38/37 BCE]; cf. Plut. Ant. 33.6–34.1, Cass. Dio 48.39.2, Socrates of Rhodes BNJ 192 F2).

Although literary sources generally present adventus ceremonies as spontaneous displays of affection, documentary sources reveal that extensive preparations were required (Lehnen 1997, 105–20). In addition to the cost of the ritual itself, dignitaries had to be fed and housed (Millar 1977, 32–35; cf. Philo Leg. 252–53 [Gaius], IGR III 1054 [Hadrian; from Palmyra, 130–131 CE], Cass. Dio 77.9.5–7 [Caracalla]). Avoiding such spontaneous displays could become a badge of honor. In a letter written to Atticus in August of 51 BCE, Cicero boasts about how little he and his retinue cost the people of Cilicia: “I tell you that … except for four beds and a roof no one takes anything—in many places they don’t even take a roof; they usually sleep in a tent” (Cic. ad Att. 5.16.3 = 109 SB; trans. Shackleton Bailey). His care not to burden the locals appears to have taken its toll on his own finances; in a letter written a few days earlier, he confesses that he is spending “a fortune” and that he is considering taking out a loan to cover his expenses (Cic. ad Att. 5.15.2 = 108 SB). Such parsimony was in good company—Suetonius reports that Augustus frequently made his arrivals and departures at night “in order that he not disturb anyone for the sake of their duty [to pay him respect]” (Suet. Aug. 53.2)—and it reflects how a Roman official could use or avoid ritual to shape his own public image.

The valences of adventus differed depending on where the ritual occurred. Outside of Rome, the emperor’s arrival temporarily transformed a city into a capital; embassies from throughout the empire came to visit him (Millar 1977, 38–39; IGR IV 1693 [Aezani; 4 CE]; IGR IV 349 [Pergamon; 117 CE]), and according to the Digest, a man barred from living in Rome could not live in any city in which the emperor was residing (Millar 1977, 39; Dig. 48.22.18 pr. 1). In addition to raising a city’s prestige, the visit of a Roman dignitary could also confer tangible benefits in the form of gifts. This eugertistic practice seems to have become so well-established by the fourth century that a panegyric for Constantine includes anticipated donations as part of an invitation to visit Autun (Pan. Lat. 7[6].22.3–4; cf. Millar 1977, 37).

Adventus at Rome worked differently. Understood in terms of Arnold van Gennep’s (1960) tripartite schema for rites of passage, it marked the reincorporation into the capital of the emperor or Roman official and his retinue, as well as the celebration of his return home and his military successes. The Roman triumph may be interpreted as representing a particularly elaborate version of this ritual. Both triumph and adventus helped to make travel and military success visible and comprehensible to an immobile population; according to Livy’s narrative for 167 BCE, Paulus’s voyage up the Tiber to Rome in a ship loaded “with Macedonian spoils” (Livy 45.35.3) was witnessed by crowds lining the riverbanks. Despite such similarities, there are significant differences, not least in that triumphs were formally granted by the Senate and the triumphator was the primary bearer of the costs (see further Beard 2007, esp. 187–96 on the costs). While highly successful generals could use triumphs to display their largesse and their exotic conquests, Cicero’s unsuccessful campaign for a triumph during the early 40s BCE shows the perils of triumphing on the cheap. One of the obstacles that he faced was the need to divert resources away from his potential triumph to repay a loan he owed to Caesar (cf. ad Att. 7.8.5 = SB 131; December 25 or 26, 50 BCE).

When it came time to leave, departures (profectiones) were also marked by ritual. Rites of departure often receive only the barest of mentions in Roman sources, but they performed an important function in publicizing and displaying official travel and military campaigns (see Livy 42.49.1–6 [discussed below] and Herod. 6.4.2; the most comprehensive scholarly treatment is by Lehnen 2001; see also Koeppel 1969; Rüpke 1990, esp. 125–51). In one of the fullest extant descriptions of a profectio ritual, Livy describes the departure of the Roman expedition against the Macedonian king Perseus in 171 BCE:

During those days the consul Publius Licinius, after the pronouncement of vows on the Capitol, departed in military dress from the city. This ritual (res) was always conducted with great dignity and grandeur, but it especially attracted the citizens’ gaze and attention, when they followed a consul about to face an enemy great and noble in virtue or fortune. Not just concern for their duty, but also a desire for spectacle (spectaculum) drew them to see their leader, in whose authority and judgment they entrusted the preservation of the entire republic. They thought about the many contingencies of war, how uncertain the outcome of fortune was … would they soon see him with his victorious army ascending the Capitol to the same gods from whom he set out or would they soon offer this pleasure to their enemies? (Livy 42.49.1)

Livy’s narrative emphasizes the civic participation engendered by the people’s gaze; as they look on the spectacle of the profectio ritual, they contemplate the realities of war and anticipate Licinius’s return (cf. Livy 44.22.17). Andrew Feldherr (1998, 11) has suggested that the profectio connects past and present, center and periphery, offering “the same vista that Livy’s monumentum provides to the audience who gaze upon it.” In Feldherr’s reading, this ceremonial departure helps Romans to visualize the events on the battlefield far from their city and connects the populace to the expedition as they watch.

While Romans hoped for and expected state officials to return, commemoration of imperial visits was even more important outside of Rome, where they were rarer. An anonymous panegyric for Constantine from 311 CE, for instance, imagines how the citizens of Augustodonum will react to his depature: “When you depart, the community will hold onto you” (Paneg. 5[8].14.4). Joachim Lehnen has plausibly suggested that the immediate context for this comment is Constantine’s restoration of the city (2001, 16), but the language expresses more broadly its aims to commemorate and extend the imperial presence. The process of monumentalizing an imperial visit took many forms. Several communities held festivals commemorating the day on which the emperor entered the city (P.Oxy. 31 2553 [late second or early third century CE], line 11; Didyma 1, 254, lines 9–13 [Didyma; dated to 212 CE or ca. 230 CE]; cf. FD III 4 [1970], no. 307, col. iii [undated]). Others commemorated the emperor’s absence through dating formulae, which expressed the time elapsed from the moment of the imperial “visit” (epidēmia) or “presence” (parousia), an apparent adaptation of the standard practice of dating by regnal year (cf. Woodhead 1981, 59–60; Ma 1999, 53–54 describes the practice of commemorating a Hellenistic monarch’s “absent authority and his having-been-there”). Such dating procedures were highly local—the date of an imperial visit holds most significance for the community in which it occurred—but they also expressed a city’s place and importance in the Roman world more broadly. In both cases, the emperor’s visit becomes a significant moment in the community, encoded into its religious and civic calendar.

While rituals of arrival and departure make an immobile community fellow participants in official movements, members of those communities also experienced travel themselves. Travel looks rather different from the perspective of individuals, such as the pair of Roman citizens, in all likelihood soldiers (cf. Bernand 1969, 101–3), who left one of the few Latin inscriptions in Philae on the South Pylon of the Temple of Isis:

I, Lucius Trebonius Oricula, was here. I, Gaius Numonius Vala, was here in the thirteenth consulship of the emperor Caesar, eight days before the Calends of April. (I.Philae 147 = ILS 2.2 8758 = CIL III 74; 2 BCE)

Lucius’s and Gaius’s rather superficial engagement with Philae stands in contrast to the deep religious significance that the site held for the numerous pilgrims who left dedications at the temple of Isis in Greek and demotic Egyptian. The contrast between these religious pilgrims and Lucius and Gaius, unlikely forbearers of the genre of narcissistic travel graffiti, raises important questions about the nature of two closely linked categories of travel, tourism and pilgrimage, which can be difficult to distinguish from one another in an ancient context (cf. Dillon 1997; Rutherford and Elsner 2005; Rutherford 2013, 12–14; Morris 1992 discusses pilgrimage more broadly). Both types of journey are undertaken by individuals, groups, and community representatives; both focus on visiting and viewing specific sites, objects, and spectacles; and finally, both sets of travelers search for memorable experiences, which are often commemorated in text, inscription, and votive offerings. They differ primarily in degrees of religiosity and intensity of experience: pilgrimage implies religious engagement with the site and an expectation that the journey there will facilitate contact with the divine (although some, such as Scullion 2005, downplay the difference in religiosity between the two categories). The term pilgrimage often carries with it a series of associations related to Christian tradition, but there are clear differences between pagan and Christian religious travel, even if there are also points of overlap (see Elsner and Rutherford 2005: esp. 3). For Christians, pilgrimage carried with it a distinct set of conventions and expectations, that need not apply to the pagan case: namely that the journey was “always performed by an individual, often as a penance, sometimes with a deeply spiritual significance” (Rutherford 2013: 12; on Christian pilgrimage, see Pilgrimage).

Further distinctions are made by both modern scholars and ancient sources between individual, personally motivated religious journeys and travel undertaken on behalf of a community. State-sponsored pilgrims were regularly referred to in Greek as theōroi, a term that ancient sources associated with the word for god (theos; cf. Lysimachus in BNJ 336 F 9; Harpocration Θ 19 s.v. theōrika = 154–55 Dindorf, Pollux II 55; both are cited and discussed by Rutherford 2013, 145 n14–16), but may actually be related to the Greek verb for viewing (theaomai; see Rutherford 2000, 136–38, and 2013, 4–6, 144–46, on the meanings and etymology of the term). In Greek and Roman contexts, state-sponsored travel to religious sites and spectacles is well-attested from the sixth century BCE through the third or fourth centuries CE, but there is evidence for forms of pilgrimage in the Mediterranean from at least the second millennium BCE (Elsner and Rutherford 2005, 10–11).

The evidence for pilgrimage and tourism is primarily textual, although some material evidence survives, especially from votive offerings (Rutherford 2013, 17–34, reviews the sources of evidence for theōria in the Greek and Roman worlds). Pilgrims worshipping Isis, for instance, traveled to shrines in Germany, Italy, Sicily, Greece, Asia Minor, and Egypt, recording their travel in inscriptions often adorned with representations of feet, which can be understood as votive offerings or commemorations, or even as expressing a wish for a safe homecoming (these inscriptions are collected and analyzed by Takács 2005). In addition, several other literary texts of the Roman period register sacred journeys (notably Aelius Aristides’s Sacred Tales and Lucian’s De Dea Syria; see Galli 2005; Lightfoot 2005; Petsalis-Diomidis 2005). Pausanias’s Tour of Greece is often read as a series of pilgrimage narratives to each of the sacred sites that the author visits (e.g., Elsner 1992; Rutherford 2001). The yearly journey (theōria) to the tomb of Achilles on the west coast of the Troad described in Philostratus’s Heoricus is even more explicitly a pilgrimage (Rutherford 2009; Heroicus 53.4–23); likewise, the discourse of pilgrimage is central to Philostratus’s Life of Apollonius of Tyana, which describes the sage’s visits to temples, tombs, oracles, and healing sites (Elsner 1997).

Apart from intellectuals for whom sacred travel was part of their identity as sophists, sages, and wise men, pilgrimage was an integral part of the reign of the emperor Hadrian, who spent more than half of his rule on the road, making frequent sacred journeys to significant religious sites, especially in Greece (Millar 1977, 36; Holum 1990; Rutherford 2001, 49–50). Having been initiated into the Eleusian mysteries, either as a student or later as Archon of Athens in 112 CE, he made several visits to Eleusis, presumably during his travels to Athens as emperor, in 124, 128, and 131 CE (Halfmann 1986, 116–17; Clinton 1989, 1516–25); he retraced Trajan’s journey to Mount Casius (Jebel al-Akra) on the Syrian coast (Anthologia Graeca 6.332; Spartianus, Vita Hadriani 14.3); and, finally, he traveled to Troy twice, first in 124 CE (Philostr. Her. 8.1) and again in 132 CE, to visit Ajax’s tomb (Halfmann 1986, 199; Philostr. Her. 67–70; see Halfmann 1986, 188–210 for further examples of Hadrian’s travels during his reign). Troy was a long-standing site of pilgrimage in the ancient world, which had been visited by Xerxes (Hdt 7.43), Alexander the Great (Arr. Anab. 1.11.5–7, Plut. Vit. Alex. 15.7–8), L. Cornelius Scipio (Livy 37.37.1–3), Caesar, (Strabo 13.1.27), and Augustus (Cass. Dio 54.7), among others (see Minchin 2012). Through his many pilgrimages, Hadrian reanimated the sacred sites of the Greek past with his presence and codified Greece as a site of religious and cultural memory.

An important aspect of pilgrimage in the Greco-Roman world was the role it played in the performance and formation of the identity of individuals and communities (see esp. Kowalzig 2005; Hutton 2005a; Lightfoot 2005; Rutherford 2013, 217–22). Whereas participation in international festivals, spectacles, and rituals expressed a community’s membership in a larger group (for the significance of athletic festivals for communities’ expressions of their own identities, see van Nijf 2001), pilgrims themselves had the opportunity to encounter other people, cities, and religious communities in the course of their travels. Inscriptions throughout the Roman world hint at the potential diversity of languages, cultures, and motivations that travelers might encounter; the cultural interaction promoted by travel; and individuals’ own strategies of self-representation. Many of these aspects of cultural interaction in modern travel are central to the anthropology of tourism (see further Bruner 2005 and Comaroff and Comaroff 2009, who emphasize the commodification of ethnicity in the tourist trade).

A prime site for studying such cultural interactions is Egypt, a popular destination for tourists and religious pilgrims during the Roman period, whose arid climate has helped to preserve a rich record of their visits (Hohlwein 1940; Foertmeyer 1989; Frankfurter 1998, esp. 166–67, 218–19; Tourism and Pilgrimage). Egypt’s pyramids were included in the canonical list of the seven wonders of the world, codified in the late Hellenistic period, and the Pharus lighthouse was frequently added to subsequent lists (Brodersen 1992; Höcker 2010). In addition, travelers visited and often left their marks in the form of graffiti at a range of Egyptian sites, such as the Memnonion at Abydos (originally the mortuary temple of Seti I and subsequently a temple of Osiris, which later housed an oracle of Bes; see Perdrizet and Lefebvre 1919; Rutherford 2003), the Sphinx at Gaza, on which visitors inscribed poems (IMEGR 127–28 [first or second century CE], 130 [second or third century CE]; cf. É. Bernand 1983), and, most famously, the two colossi of Memnon at Thebes (see Travel and Pilgrimage for a further list of sites). The pair of colossi—in fact two monumental statues of Amenhotep III—were identified by Greek and Roman travelers as depictions of Memnon, the mythological king of Ethiopia. Evocative Latin and Greek graffiti record visitors’ firsthand experience of the mysterious dawn “singing” for which the monument was famous (Strabo 17.1.46; Paus. 1.42.3; Letronne 1833; Bernand and Bernand 1960; Bowersock 1984; Théodoridès 1989; see Travel and Pilgrimage). Through their celebrity, the colossi became sites for literary and cultural interaction. In a cycle of four elegiac poems, which are closely modeled on Sappho, Julia Balbilla, an elite Roman woman of Syrian origin, commemorates Hadrian’s visit in November 130 CE (Memnon 28–31; Bowie 1990). Patricia Rosenmeyer shows how the poems balance encomium for the emperor with Balbilla’s own lyric voice and desire for poetic fame (Rosenmeyer 2008). Balbilla’s composition therefore both looks inward, toward recording and commemorating their experience, and also outward, toward subsequent visitors, readers, and poets.

Balbilla’s lyric voice, which inscribes official movements in an individual’s subjective experience, can be read as participating in a wider genre of literature that incorporates travel within its narrative and poetic imagination. Movement—of people, goods, and ideas—was deeply embedded in the lived experience of the Roman world in its infrastructure; ritual practice; and individuals’ journeys as tourists, pilgrims, and economic agents. This final section treats the place of travel in the Roman cultural imaginary, considering how literature represents and responds to the Roman impetus for exploration and how literary texts participate in shaping Roman conceptualizations of the travel, movement, and space of their world. It is impossible to separate the practice of travel from the representation of it, and in a sense this final section presents a false dichotomy between praxis and representation. And yet many Roman texts can give us insight into how people at different places and times conceived of travel and geography. My focus in this section reflects a growing emphasis on reading travel, space, and place in ancient literary texts through the lens of a nexus of critical approaches collectively termed the “spatial turn,” which represent the translation of the methodologies of cultural geography, anthropology, and urban studies to the analysis of literary, artistic, and cultural practice.

The study of narrative has taught readers to be alert to the relationship between a text’s fabula (the underlying sequence of events) and its sujet (the narrative arrangement and presentation of those events), to use two terms that originated in Russian formalism (e.g., Tomaševskij 1965) and have been widely applied in the field of narratology (see de Jong 2014, esp. 5–6). In narrative, the relationship between fabula and sujet is mediated by one or more “focalizers,” whose point of view is either implicitly or explicitly reflected in the way a story is told. When considering texts’ presentation of space and travel, there is a parallel set of distinctions at work: namely, the relationship between movement, real or imagined, and the narrative representation of it. Movement looks very different depending on where the observers stand; the narratology of space provides a framework and vocabulary for describing the contribution of narrative to the presentation and depiction of movement in literature. Two such narrative modes were outlined by Pietro Janni (1984), who drew on the earlier work on spatial perception and presentation by the psychologist Kurt Lewin (1934): (1) hodological perspectives, which present movement from the perspective of the traveler, and (2) cartographic perspectives, in which travel is described top-down, from a bird’s-eye-view. Michel Rambaud (1966, 1974) posited a tripartite model of space in Caesar’s works that distinguished among “geographic,” “strategic,” and “tactical” space (see further Riggsby 2006, 21–45), which I discuss below. Rambaud’s geographic space corresponds roughly to Janni’s cartographic perspective; his categories of strategic and tactical space, however, divide hodological perspectives into the linear progression of, say, columns of soldiers (strategic space) and the surveying gaze (tactical space). These perspectives—either in Janni’s bipartite or Rambaud’s tripartite division—need not be mutually exclusive, and narratives often switch seamlessly between them, but they represent very different narrative and descriptive logics, which cannot simply be scaled up or down to change spatial perspectives.

In his classic essay, “Forms of Time and of the Chronotope in the Novel,” Mikhail Bakhtin explores the historical development in the novelistic genre of what he calls the chronotope, “the intrinsic connectedness of temporal and spatial relationships that are artistically expressed” (1981, 84). At the conclusion of a sweeping analysis, ranging from the first (Chariton and Petronius) to the sixteenth (Rabelais) centuries CE, he articulates a chronotope of the road, in which “time, as it were, fuses together with space and flows in it (forming the road); … it often happens that all the events of a novel either take place on the road or are concentrated along the road (distributed on both sides of it)” (Bakhtin 1981, 244). The chronotope of the road represents a hodological view of travel, that is, a highly restricted narrative perspective, which reflects the perceptions of the traveler without orienting the reader in a broader geographic and cartographic context (cf. Janni 1984, 79–158, esp. 79–90; Purves 2010, 145–48; Lightfoot 2014, 22–26). It is a powerful structuring device in Roman literature. To offer one example of this narrative strategy in action, Horace’s “Journey to Brundisium” (Hor. serm. 1.5; 36/5 BCE) adopts a strict hodological perspective, which seems to model itself on the form and style of an itinerary or travel guide. As Emily Gowers notes, the poem’s ego-narrator is occupied with “peevish, diary-style entries on the minutiae of travel: dyspepsia, nocturnal disturbances, sore eyes, low-level incidents” (2012, 182). Its hodological first-person narration resists translation onto a cartographic perspective: although the poem begins as a journey south along the Via Appia, the narrator somehow finds himself on the Via Minucia (Hor. serm. 1.5.94; cf. Gowers 2012, 205). Such resistance appears to be part of a larger strategy by which the poem’s tightly focused perspective obscures the larger diplomatic and political implications of the journey, substituting the mundane and personal for the lofty and the civic (cf. Reckford 1999); geographic perspective thereby becomes an index for other registers of content. If Horace’s companion in the poem, Heliodorus, is to be identified with the author of the Theamata Italia, as Gowers suggests (2012, 183), this connection becomes even more explicit. More broadly, Satires 1.5 might be seen as part of a larger genre of Latin travel poetry, including Lucilius’s Iter Siculum, which Porphyrio names as Horace’s model (ad Sat. 1.5.1), as well as several poems that survive either only as titles or in fragments: Varro’s Menippean satires, such as Marcipor “Varro on the Road,” Periplous “Circumnavigation,” and Sesculixes “Half-Ulysses”; Caesar’s poem Iter (cf. Suet. Jul. 56); Valgius (Morel, FLP 106); Ausonius’s Mosella; and Rutilius Namatinaus’s De reditu suo siue Iterum Gallicum (as collected by Gowers 2012, 183–84).

Bakhtin himself connected the chronotope of the road with Roman fiction, namely Petronius’s Satyrica and Apuleius’s Metamorphoses, both of which revel in the first-person restricted perspective of travel. The extant fragments of Petronius’s novel, which was likely composed during the reign of Nero, open with the protagonist Encolpius and his traveling companion Ascyltus lost and going in circles. “Wherever I went,” he says, “I kept coming back to the same place” (6.4). After they each independently get directions from the same old woman, they find that they have been led into a brothel. This episode makes the novel’s restricted hodological perspective part of its joke: they both follow the same directions and end up at the same unsavory place because they cannot imagine their movement cartographically. The chronotope of the road, punctuated by digressions and diversions, organizes the fragments of the Satyrica as we have them: the pair travel throughout Magna Grecia, visiting cities such as Massilia, Croton, and, possibly, Puteoli (if this is the “Greek city” mentioned at 81.3; see further Rose 1962). Apuleius’s Metamorphoses, written in the latter half of the second century CE, takes the hodological perspective further; like its Greek models, such as the (Ps.-)Lucianic Onos (second century CE[?]), travel is narrated through the experiences and perspectives of Lucius, a human who has been transformed into an ass. (P.Oxy. 4762 also contains a Greek fragment of an ass novel, but the papyrus, which was copied in the third century CE, postdates the Metamorphoses, even though the text it contains might predate Apuleius’s novel.) Lucius is barred from leaving the road, and because he is treated as an animal, rather than a human, he lacks agency to determine his own movement and direction. Thus narrative form responds to and reflects the protagonist’s restricted movements.

The chronotope of the road is a narrative strategy and organizational principle for other Greek texts written under the Roman empire, such as Achilles Tatius’s novel, Leucippe and Clitophon (second century CE), which departs extremely rarely from its highly restricted, first-person narrative perspective. (Perhaps the most striking departure comes in an ethnographical description of the habits and lifestyles of Egyptian outlaws living in the Nile Delta 4.12.) In each of these texts, the experience of reading mimics that of the traveler (cf. de Temmerman 2012, 521–22). The road is also a central organizing principle in two quasi-technical genres of Roman-era prose: the itinerary and the periplous (circumnavigation). Pausanias’s Tour of Greece (third quarter of the second century CE) can be read as a transformation and elevation of these genres, further augmenting them with long ethnographical narratives of the places he visits. In other words, his text combines narration of his first-person experiences of travel with detailed description and reporting (see further Elsner 2001).

In contrast to the tightly restrictive, hodological mode of presentation I have just described, narratives can also “zoom out” to view the world and movement within it as if gazing down from above. This impossible, cartographic view is a highly desirable one, and it is often correlated with increased knowledge and an omniscient narrative voice, particularly in ethnographic descriptions. Thus, the island of Britain is imagined as triangular by the geographers and ethnographers of the Roman world, such as Caesar (BGall 5.13), Strabo (4.5.1), Pomponius Mela (3.50), and Tacitus (Agr. 10.3, although the shape to which he compares the island may be hopelessly corrupted: Is a scutula an eye-patch, an elongated dish, a small shield? See Woodman 2014, 133–34 with Ogilvie and Richmond 1967, 168–170).

Few works, if any, can consistently maintain such distant cartographic perspectives. To look more carefully at Caesar’s De bello gallico (written either during or shortly after Caesar’s campaign in Gaul from 58 to 51 BCE), Andrew Riggsby characterizes the function of each of its narrative perspectives as follows: “People fight in tactical space, move through strategic space, and inhabit geographic space” (Riggsby 2006, 25, cf. Rambaud 1974). Caesar’s autobiographical narrative of travel for the purpose of military conquest, therefore, combines the chronotope of the road—exemplified by his descriptions of “forced marches” (magna/maxima itinera, e.g. BGall 1.7, 10, 37, 38, etc.), bridges (BGall 1.13), and other technologies of travel—with strategic and ethnographical views of the landscape. These points of view are highly significant for the meaning of each instance of travel. Thus Riggsby argues that Caesar’s tactical space, a localized representation of the field of battle articulated by its natural features (e.g., his description of the town of Vesontio, BGall. 1.38.4–7), should be understood in the context of Roman agrimensores, whose surveying activities for Roman centuriation described and depicted localized environments from a bird’s-eye-view (Thulin 1971; Riggsby 2006, 32–45). For Caesar and the agrimensores, the process of describing a landscape from the strategic perspective is an important precursor to possessing it. By contrast, Caesar’s cartographic gaze, which is perhaps most obvious in the famous opening description of the tripartite division of Gaul (1.1), falls into a different category of spatial representation. By virtue of its extreme distance, this point of view severs the intimate connection between description and possession of territory; in its place, it gives the narrator an omniscient ethnographical vista from which to describe territories and peoples. From this vantage, the Gauls’ courageousness (fortis) can be correlated with their geographic position vis-à-vis the Germans and Rome: the closer one is to Germany, the braver one must be (Caes. BGall. 1.1; Riggsby 2006, 28–32; cf. Dench 2005, 49–54 for a reading of Caesar’s manipulation of ethnography). By varying its spatial perspectives, the text’s narrative economy is able to reveal and conceal elements of the Roman experience of Gaul.

Scholarship has been particularly attuned to the relationship between geography, the depiction of movement in Roman poetry, and the ideology of the Augustan empire (e.g., Myers 2008; Lindheim 2010, 2011). To take a well-known example, Virgil’s Aeneid (19 BCE), which takes the wandering Homeric Odyssey as a central model, encodes travel into the founding of the Roman state: the poem’s narrative of Aeneas’s journey, its descriptions of the cities he visits, and its ethnography of the Italian landscape write the process of traveling and its complications into Roman identity in the Augustan age (see Thomas 1982, esp. 93–107 on ethnography; Horsfall 1989 on Aeneas as colonist; Quint 1993). The Virgilian connection between travel and Roman cultural identity is mirrored in Livy’s account of Aeneas’s travels to Italy (Livy 1.1); the two genres—epic and historiography—present these origin myths differently, but both emphasize travel as essential to the foundation of the Roman state.

The teleology of the Aeneid’s travels establishes a clear sense of center, but, as Philip Hardie (1986) has argued, the poem imagines Rome’s periphery to be a (literally) boundless one, “an empire without an end” (imperium sine fine, Verg. Aen. 1.279). The connection between travel and state foundation in the Aeneid is indicative of a broader trend for the poetics of travel and space to invite and frame discussion of the political, the social, and the moral in the poetry of Virgil and his contemporaries: for example, the countryside of the Eclogues and the ethnographic descriptions in Georgics 2 (Virg. Geo. 2.83–176; see further Thomas 1988, 170–90), Horace’s Sabine Villa (Epist. 1.7.1–2, 1.16.1–16, Odes 1.17.1–2) and other country retreats (Epist. 1.10.15–21), Horace’s Odes 1.3 (Virgil’s journey), Tibullus’s country retreat (1.5.21–34, where the countryside represents an inversion of the poet’s relationship with his beloved), and the dynamics of city and country in Propertius (see, e.g., Lindheim 2010, 2011). In Roman poetry, travel therefore offers a space to discuss a range of other issues, such as political power, imperial ambitions, and Roman urbanism. This tendency becomes a central subject of Ovid’s exile poetry, in which his rupture from Augustus consigns him to the “edges of the Earth” on the Black Sea (e.g., Trist 1.1; see Williams 1994 on Ovid’s construction of his exiled voice).

The focus on travel is not limited to the poetry of the Augustan era. Thus, Statius’s Silvae (Books 1–3, 93 or 94 CE; Book 4, 95 CE) has been shown to implicate travel in the political and social world of the poet and his patrons (Parrott 2013). Identity and empire are even more explicitly linked to travel in other genres, such as geography, historiography, encyclopedic and scientific literature, oratory and oratorical theory, and prose fiction. Just as physical travel encouraged the contemplation of empire and the confrontation with other peoples and places, literary representations of movement offered space for discourses about the relationship of ethnicity, identity, and imperial power.

For my final pair of case studies, I return to the personal experience of travel with a form of travel story often called the narrative of the quest. In a subset of the Greek and Latin prose fiction of the second and third centuries CE, travel narratives take on deeper significance, and movement through space becomes a kind of pilgrimage, signaling a journey toward greater enlightenment and knowledge of the self. Insofar as it is read as a narrative of pilgrimage, Pausanias’s Tour of Greece might be understood as an example of this type of travel narrative (see further the essays in Alcock, Cherry, and Elsner 2001; Hutton 2005b; Pretzler 2007; Travel and Travel Writing).

Philostratus’s Life of Apollonius of Tyana, composed in the first half of the third century CE (Jones 2005, 2–3), narrates the peregrinations of the wise man to the four corners of the Roman empire (cf. Elsner 1997, 29): north as far the Caucasus (2.2), east through Babylon (1.21–2.1) to India (2.17–3.50), south to Egypt and Ethiopia (5.43–6.28), and west to Spain and the Pillars of Hercules (4.47–5.10). Apollonius’s travels are intimately connected to his own discovery of his philosophic and sagacious self. From the Brahmans in India, for instance, Apollonius learns both the seemingly impossible, levitation (meteōroporein, 3.15, 17) and foreknowledge (prognōsis, 3.16), as well as what lies at the core of their powers, self-knowledge (eidenai heauton, 3.18). In perhaps the most famous passage of the Life, he announces that he has traveled to Egypt in order to find the sources of the Nile: “My own purpose in coming here was to visit you and the sources of the Nile. It would be pardonable to miss these in one who got only as far as Egypt, but for one who has advanced into Ethiopia, as I have, it would be a disgrace to pass them by, and not draw some learning from them” (6.22.1; trans. Jones). Apollonius expresses his desire for knowledge of the river’s sources in terms of the language and metaphors of travel and physical movement (cf. Elsner 1997: 32); it is not enough to hear about them from his experienced (ethas, 6.22.2) guide. As Elsner (1997) shows, Apollonius’s journeys, which emphasize the extent of the Roman empire, become a central source of his own wisdom and sagacity as well as a potential refuge from the tyranny that resides in the center. Thus, when Apollonius is tried before Domitian in Rome, he repeats a saying of the Brahmans in his own self-defense (3.18). His travels also provide him with authority, especially in religious matters. In book 5, for instance, he uses his travels to India to justify “correcting the rites of the Egyptians” (5.25).

Apuleius’s Metamorphoses can be read as a reinterpretation of the narrative of the quest, in which the links between pilgrimage and self-discovery outlined by Apollonius take the form of physical changes (see further Harrison 2001–2002). Like Philostratus’s Life of Apollonius of Tyana, the Golden Ass intersects with the world of the Roman empire, although the protagonist Lucius never encounters the highest echelons of Roman power (Millar 1981). Early in the text Lucius, an overly curious sophist, is transformed through a magical mishap into an ass (3.24); once he is transformed, the narrative is in the simplest terms an account of Lucius’s quest for roses, the antidote to his magical transformation. During his life as an ass, Lucius is passed from owner to owner; as a beast of burden, his freedom of movement is restricted, and he is forced to travel or face beating (in a persuasive reading, Bradley 2000 argues that the novel explores Roman slavery through Lucius’s transformation and forced migration). His salvation comes in the eleventh and final book, when he is told by the goddess Isis, who appears in a dream vision, that he will find a garland of roses at a procession dedicated to her (11.6). Whether it is read as an entertaining parody or a serious religious experience (see further Winkler 1985), book 11 narrates Lucius’s return to human form and his conversion to Isiac religion. He finds and eats the roses at the festival of Isis (11.13) and is restored to his former state. As a result of his journey and his experience, he becomes an initiate in Isiac cult (11.23–24) and a devotee of Isiac religion, eventually moving to Rome (11.26). The novel closes with Lucius preparing to become an initiate of Osiris (11.30). Like Philostratus’s Life of Apollonius of Tyana, the Golden Ass connects travel to a process of pilgrimage, enlightenment, and self-knowledge; Apuleius turns this paradigm on its head, externalizing in Lucius’s grotesque form as an ass an experience that remains internal and, presumably, deeply personal for other sagacious pilgrims.

Lucius’s experience in the Metamorphoses is a test case for the extent to which travel can be a transformational experience for individuals in the Roman world. Yet for all the changes that he undergoes, he concludes the novel in more or less the same physical form that he began it. (He now has a shaved head and has changed his wardrobe.) Such an ending suggests that the experience of travel leaves few physical traces on individuals and landscapes. Travel is an ephemeral process, one that must be continually commemorated and maintained or cease to exist. Yet despite, or perhaps because of, the challenges inherent in preserving travel, it remains a central organizing principle of many literary texts from the Roman world. Its centrality further suggests how important the movement of people and goods was to all facets of the Roman experience. From emperors to slaves, orators to merchants, travel through the world and the infrastructure that enabled it were defining features of the practicalities, realities, and imaginations of Roman life.

Papyrological abbreviations follow http://library.duke.edu/rubenstein/scriptorium/papyrus/texts/clist.html.

Abbreviations for Greek inscriptions follow http://epigraphy.packhum.org/biblio.html.