Imperial Geography and the Medieval Peutinger Map

Abstract

The Peutinger map is an extraordinary world map drawn c.1200 and long considered a copy of a Roman road map made for late antique travellers. This paper presents arguments against these assumptions and concludes that the lost original was more likely to be a Carolingian display map. Ninth‐century scribes had the expertise and resources necessary for creating an antiquarian work based on Roman itinerary lists, while Carolingian rulers had ample motivation for commissioning a map to display their Roman imperial ambitions.

Keywords:

• Peutinger map
• Roman itineraries
• mappaemundi
• Carolingian maps
• Roman geography
• Charlemagne
• Roman roads
• Jerusalem
• Hermann von Reichenau
• Beatus maps

Acknowledgements

I am grateful for the suggestions made by Richard Talbert and Michael Maas, leaders of the National Endowment for the Humanities seminar in Rome in the summer of 2000, and by the participants, including Brenda Fineberg, Natalia Lozovsky, Frank Romer and Miriam Vivian. I also benefited from the advice of Marcia Kupfer, Benet Salway and participants at other meetings where I gave papers on various aspects of the map. These conferences include Shifting Frontiers in Late Antiquity IV (San Francisco, 2001), and international congresses on medieval studies (Kalamazoo, 2001, 2002 and 2004). Colleagues in my reading group at the University of California, Davis, offered welcome commentary on a rough draft. Conversations with Peter Onuf and critiques of earlier versions, by Ingrid Baumgärtner, Evelyn Edson, Marcia Kupfer, Natalia Lozovsky, Benet Salway, David Traill and the editor and anonymous reviewers of Imago Mundi improved my argument. Any errors that remain are of my own making.

Notes

Emily Albu is associate professor of classics at University of California, Davis.

I first saw a facsimile of the entire Peutinger Map (Vienna, Österreichische Nationalbibliothek, Codex Vindobonensis 324), in summer 2000 at the American Academy in Rome, where I was participating in a National Endowment for the Humanities seminar on Roman mapping and ethnography. To that group I presented my view that this was a medieval map intended for display and serving some of the same functions as the Bayeux Tapestry.

On the map's geographical coverage as well as its exact dimensions, missing leaf or leaves, editions and reproductions, see Richard Talbert, ‘Cartography and taste in Peutinger's Roman map’, in Space in the Roman World: Its Perception and Presentation, ed. Richard Talbert and Kai Brodersen (Münster, Lit Verlag, 2004), 113–41.

C. R. Whittaker, ‘Mental maps: seeing like a Roman’, in Thinking Like a Lawyer: Essays on Legal History and General History for John Crook on His Eightieth Birthday, ed. Paul McKechnie (Leiden, Boston and Cologne, Brill, 2002), 81–112, reference at 82.

O. A. W. Dilke, ‘Itineraries and geographical maps in the early and late Roman empires’, in The History of Cartography, Vol. I: Cartography in Prehistoric, Ancient, and Medieval Europe and the Mediterranean, ed. J. B. Harley and David Woodward (Chicago, University of Chicago Press, 1987), 234–57, reference at 238. Dilke thought that the archetype dated from between ad 335 and 366, though the presence of sites destroyed by the eruption of Vesuvius in ad 79, he believed, might argue for a source much earlier still. This suggestion echoes Leo Bagrow's assumption that the original was a Roman product of the first century (Leo Bagrow, History of Cartography, 2nd ed., rev. R. A. Skelton (Chicago, Precedent Pub., 1985), 37–38). Ekkehard Weber, ‘Zur Datierung der Tabula Peutingeriana’, in Labor omnibus unus: Gerold Walser zum 70. Geburtstag dargebracht von Freunden, Kollegen und Schülern, ed. Heinz E. Herzig and Regula Frei‐Stolba; Historia Einzelschrift 60 (Stuttgart, F. Steiner Verlag Wiesbaden, 1989), 113–17, presented the argument for an early fifth‐century date. Patrick Gautier Dalché, ‘La trasmissione medievale e rinascimentale della Tabula Peutingeriana’, in Tabula Peutingeriana: Le antiche vie del mondo, ed. Francesco Prontera (Florence, Leo S. Olschki, 2003), 43–52, reference at 45, note 13 , has cited the sources that establish the dating of the existing map to no later than the beginning of the thirteenth century.

Ekkehard Weber, Tabula Peutingeriana. Codex Vindobonensis 324. Kommentar (Graz, Austria, Akademische Druck‐ und Vorlagsanstalt, 1976), 22, theorized that the transmission of an ancient world map included a Carolingian copy. On this see Gautier Dalché, ‘La trasmissione medievale’ (note 4), 46–47.

Benet Salway, ‘Travel, itineraria and tabellaria’, in Travel and Geography in the Roman Empire, ed. Colin Adams and Ray Laurence (London and New York, Routledge, 2001), 22–66, reference at 58.

Dilke, ‘Itineraries and geographical maps’ (see note 4), 238. This amounts to some 104,000 km of marked roads.

Dilke, ‘Itineraries and geographical maps’ (see note 4), 238–39.

O. A. W. Dilke, Greek and Roman Maps (London, Thames and Hudson, 1985; reprinted, Baltimore, Johns Hopkins University Press, 1998), 114.

Pascal Arnaud, ‘L'origine, la date de rédaction et la diffusion de l'archetype de la Table de Peutinger’, Bulletin de la Société nationale des antiquaires de France (1988): 302–21. Arnaud located this private commission in the early fifth century, when relative peace encouraged civilian travel and religious fervour induced Christians to embark on lengthy pilgrimages. Dilke also recognized this as a civilian map, since no military stations are indicated (Greek and Roman Maps (see note 9), 115). See also Salway, ‘Travel, itineraria and tabellaria’ (note 6), 44–47. I follow the practice of using the convenient singular ‘mapmaker’, while recognizing that a team might well have constructed the map.

Emil Schweder, ‘Über den Ursprung und die ältere Forme der Peutingerschen Tafel’, Jahrbuch für Philologie 39 (1893): 485–512.

David Woodward, ‘Medieval mappaemundi’, in Harley and Woodward, The History of Cartography (see note 4), 286–370, reference at 335. From among the many allusions to these now lost wall maps, Woodward here cites the reference to three by Matthew Paris (c.1200–1259), on the ‘world’ map in his Chronica majora (British Library, Cotton MS. Nero D.V., fol. 1v).

The Vicarello Goblets list 106 mansiones between Gades and Rome, with the mileage from one to the next (Raymond Chevallier, Roman Roads, transl. N. H. Field (Berkeley, University of California Press, 1976), 47–50). For a summary of some other lists, see Salway, ‘Travel, itineraria and tabellaria’ (note 6), 26. Sencer Şahin, ‘Ein Vorbericht über den Stadiasmus Provinciae Lyciae in Patara’, Lykia: Anadolu‐akdeniz arkeolojisi 1 (1994): 130–37, has added another itinerary to the listing of R. Capelli and F. Pesando, ‘Gli itinerari romani: repertorio bibliografico’, in Viae Publicae Romanae: X Mostra europea del Turismo, artigianato e delle tradizioni culturali, Roma, Castel Sant'Angelo, 11–25 aprile 1991, ed. Rosanna Capelli (Rome, Leonardo–De Luca, 1991), 41–44. On the ubiquity of these itineraries, see Ray Laurence, ‘Afterword: travel and empire’, in Adams and Laurence, Travel and Geography in the Roman Empire (note 6), 167–76, reference at 168.

To envision the routes listed in the Antonine and Bordeaux itineraries, see Figure 3.1 in Salway, ‘Travel, itineraria and tabellaria’ (note 6), 23–25.

Ravenna cosmography 1.18.10–15, translation in O. A. W. Dilke, ‘Cartography in the Byzantine Empire’, in Harley and Woodward, The History of Cartography (see note 4), 258–75, reference at 260. Dilke concludes that the cosmographer had access to a collection of Greek and Roman maps. But true itineraries seem a more likely source if the compiler really could have added mileage as the cosmographer suggests. A few of these itineraries were evidently similar to some used by the creator of the Peutinger prototype. We will be able to compare the two more easily when we have the full critical edition of the Peutinger Map, in the form of an electronic database, which Richard Talbert is producing. A companion volume by Emily Albu and R. W. Benet Salway, in preparation for Liverpool University Press, will provide indices and commentary. For the full text of the Ravenna Cosmography, see Ravennatis anonymi Cosmographia et Guidonis Geographica, ed. Moritz Pinder and G. Parthey (Berlin, Fridericus Nicolaus, 1860; reprinted Aalen, Otto Zeller Verlagsbuchhandlung, 1962). For other itineraries, including that of Theophanes (an Egyptian official travelling in the 320s), see Colin Adams, ‘“There and back again”: getting around in Roman Egypt’, in Adams and Laurence, Travel and Geography in the Roman Empire (see note 6), 138–66, reference at 160–63.

The Pithoeanus, now Paris Lat. 4808. Dilke, ‘Itineraries and geographical maps’ (see note 4), 237.

Dilke, ‘Itineraries and geographical maps’ (see note 4), 238.

Dilke (ibid., 239), counted fifty‐two pictographs for spas alone, including twenty‐eight called Aquae, that is, presumably featuring baths. Kai Brodersen has presented the evidence that itineraria picta (illustrated itineraries) contained such decorative images (‘The presentation of geographical knowledge for travel and transport in the Roman world: Itineraria non tantum adnotata sed etiam picta’, in Adams and Laurence, Travel and Geography in the Roman Empire (see note 6), 7–21, reference at 14–16).

Chevallier, Roman Roads (see note 13), 29, noted that the imprecision extended even to the drawing and relative placement of the roads: ‘The fact is, the road routes themselves are looked upon as being more or less rectilinear and are drawn independently of each other. There is no guide to the orientation of the roads nor to the significance of their bends and twists, so the road stations are not sited in their correct relative positions’. See also the analysis of Weber (Tabula Peutingeriana (note 5), 12), concerning the similarly imprecise positioning of rivers and mountain ranges.

Marcia Kupfer, ‘Medieval world maps: embedded images, interpretive frames’, Word and Image 10 (1994): 262–88, reference at 264–69, has collected and analysed examples.

For a wealth of evidence, see Natalia Lozovsky, The Earth Is Our Book: Geographical Knowledge in the Latin West ca. 400–1000 (Ann Arbor, University of Michigan Press, 2000). She presents further evidence of Carolingian interest in Roman geography, and the intimately related connections between imperial power and conquest as understood by Carolingian rulers and scholars, in her article, ‘Roman geography and ethnography in the Carolingian Empire’ (forthcoming in Speculum). I thank the author for sending me an advance copy of this valuable essay.

Lozovsky, The Earth Is Our Book (see note 21), 139. My brief discussion here draws its argument and particulars from this work. Its author has reminded me that while Carolingian ‘geographers’ kept Roman place‐names and presented an essentially Roman view of the world, they also modified the arrangement of their Roman pieces by shifting the focus from the Mediterranean to northern Europe.

We possess, in many copies, the commentaries of Remigius of Auxerre, John Scottus Eriugena and Martin of Laon.

De nuptiis 6.588–end. This encompasses nearly all of book 6 (of 9 books). Martianus Capella, ed. Adolf Dick, 2nd ed. (Stuttgart, Teubner, 1978).

Dicuili liber de mensura orbis terrae, ed. J. J. Tierney and Ludwig Bieler (Dublin, Dublin Institute for Advanced Studies, 1967). For the argument that Dicuil and the Anonymous Leidensis introduced some innovations into the Roman picture of the world, see Patrick Gautier Dalché, ‘Tradition et renouvellement dans la représentation d'espace géographique au IXe siècle’, Studi medievali, ser. 3, 24 (1983): 121–65.

Lozovsky, The Earth Is Our Book (see note 21), 139.

Charlemagne sent embassies to Baghdad in 797, 802 and 807. Richard Hodges, Towns and Trade in the Age of Charlemagne (London, Duckworth, 2000), 36. Two of these journeys lasted four years each, prolonged by frequent absences of the caliph Harun al'Rashid from Baghdad as he pursued military raids and pilgrimages (Michael McCormick, Origins of the European Economy: Communications and Commerce ad 300–900 (Cambridge, Cambridge University Press, 2001), 471). Ambassadors whom Charlemagne had dispatched to Jerusalem returned in 800 bearing gifts from the Patriarch of that city. Charlemagne undertook a vast building campaign in Jerusalem, where he founded monasteries and constructed ‘a hostel, a library and a marketplace’. He even bought land in the Valley of Jehoshaphat, hoping to support his establishments in the Holy Land with the revenues from these estates (Moshe Gil, ‘The political history of Jerusalem during the Early Muslim Period’, in The History of Jerusalem: The Early Muslim Period, 638–1099, ed. Joshua Prawer and Haggai Ben‐Shammai (Jerusalem and New York, Yad Izhak Ben‐Zvi and New York University Press, 1996), 1–37, reference at 14.

Hodges, Towns and Trade (see note 27), 36. The book's second chapter, ‘Charlemagne's elephant’, demonstrates Charlemagne's ties to Muslim lands, India and China. ‘The Carolingian king’, concludes Hodges, ‘was clearly intrigued by the East’.

Einhard, Vita Karoli Magni, ed. and transl. Evelyn Scherabon Firchow and Edwin H. Zeydel (Coral Gables, University of Florida Press, 1972), chap. 33. Einhard calls the three mensae (tables, but also table tops, slabs, planks). He is unlikely to have known the new word for map, which was entering the Latin language as he wrote. In any case, that term (mappa, originally meaning napkin), would not suit a rigid silver map. For centuries tabula (writing tablet, counter, public list, painted tablet), served to designate a map. (So the parchment Peutinger map is more commonly called the Peutinger Table from the Latin Peutingeriana tabula.) In his catalogue of Charlemagne's bequests, Einhard includes a fourth mensa, described only as ‘made of gold and extraordinarily large and heavy’. Concerning the value of these objects as ‘transmitters of symbolic meaning’, see Matthew Innes, ‘Charlemagne's will: piety, politics and the imperial succession’, English Historical Review 112 (1997): 833–55, reference at 848–49.

‘Tertiam, quae ceteris et operis pulchritudine et ponderis gravitate multum excellit, quae ex tribus orbibus conexa totius mundi descriptionem subtili ac minuta figuratione conplectitur’ (Einhard, Vita Karoli (see note 29), chap. 33).

For a summary of this episode, see Warren Treadgold, A History of the Byzantine State and Society (Stanford, Stanford University Press, 1997), 423–24. On the negotiations between Irene and Charlemagne, Treadgold concludes, ‘The marriage would settle their conflicting claims to be emperor and would legally unite the Byzantine and Frankish empires. What this could have meant in practice was and is quite unclear, but the empress was willing to consider it’.

Pascal Arnaud, ‘L'affaire Mettius Pompusianus, ou le crime de cartographie’, Mélanges de l'Ecole française de Rome: Antiquité 95 (1983): 677–99. Arnaud uses Dio Cassius and Zonaras as his primary sources. The Roman historian Suetonius (Domitian 10.3, in his De vita Caesarum), reported that Pompusianus carried the parchment map on his person. We cannot tell what sort of lists or drawings constituted this ‘map’.

Heather Hyde Minor, ‘Mapping Mussolini: ritual and cartography in public art during the second Roman Empire’, Imago Mundi 51 (1999): 147–62.

Liber Pontificalis, ed. L. Duchesne, 3 vols (Paris, E. Thorin, 1886–1957), 1: 432.

Kupfer, ‘Medieval world maps’ (see note 20), 267–69.

Thegan (d. c.848), biographer of Louis the Pious, wrote that of all Charlemagne's treasures, Louis kept only this silver map, ‘triple‐formed like three overlapping disks’. According to Thegan, Louis held this sole keepsake as a fond remembrance of his father (ob amorem patris), paying a handsome price to retain this memento (Thegan, Vita Hludowici Imperatoris, chap. 8, ed. Radolf Buchner, Quellen zur Karolingischen Reichsgeschichte (Berlin 1955), 1: 222). Einhard (Vita Karoli (see note 29), chap. 33), wrote that the map belonged to a part of Charlemagne's estate to be divided among his heirs and those designated to receive alms. In 842 Lothar, son of Louis, destroyed this precious artefact and distributed the pieces to aristocrats in a desperate attempt to buy their support in the civil war against his brothers. It is difficult to overestimate the symbolic weight of this act, by which Lothar signalled his offer to share universal power with loyal nobles. The destruction of the map presaged the dissolution of the empire of Charlemagne in the following year.

On Charlemagne's eagerness to reclaim a Roman heritage, see Erwin Panofsky, Renaissance and Renascences in Western Art (Stockholm, Almqvist & Wiksell, 1960), 44; and the critique of Panofsky by Lawrence Nees, A Tainted Mantle: Hercules and the Classical Tradition at the Carolingian Court (Philadelphia, University of Philadelphia Press, 1991), 5–10. Like Nees, recent historians have emphasized Charlemagne's Christian revival over his classicizing interests. Yet as earlier generations of commentators noted, Carolingian rhetoric revealed a keen interest in proclaiming ties to classical antiquity, whose Latin texts the Carolingians preserved from oblivion. Carolingian scribes would have been able to observe that their most ancient pagan texts survived on papyrus rolls, while early Christian writers had moved more quickly to the codex form. On the persistence in the use of papyrus rolls by pagan writers while their Christian contemporaries were choosing the book format for religious texts, see L. D. Reynolds and N. G. Wilson, Scribes and Scholars: A Guide to the Transmission of Greek and Latin Literature, 3rd ed. (Oxford, Clarendon Press, 1991), 34–35.

For the representation of Jerusalem on medieval maps, see Ingrid Baumgärtner, ‘Die Wahrnehmung Jerusalems auf mittelalterlichen Weltkarten’, in Jerusalem im Hoch‐ und Spätmittelalter: Konflikte und Konfliktbewältigung—Vorstellungen und Vergegenwärtigungen, ed. Dieter Bauer, Klaus Herbers and Nikolas Jaspert (Frankfurt and New York, Campus Verlag, 2001), 271–334.

Twelfth‐century crusaders remembered this name and understood its history. So, for instance, Fulcher of Chartres wrote in his account of the crusaders, Historia Hierosolymitana (c.1130), ‘hanc urbem etiam Aelius Hadrianus imperator mirifice decoravit et vicos et plateas decenter pavimentis exornavit. de cuius nomine Iherusalem Aelia vocata est’ [This city, indeed, the emperor Aelius Hadrian marvellously beautified, and he suitably paved its streets and piazzas. From his name, Jerusalem is called Aelia] (Fulcheri Carnotensis Historia Hierosolymitana (1095–1127), ed. Heinrich Hagenmeyer (Heidelberg, C. Winter, 1913), 1.26.12).

Gautier Dalché, ‘La Descriptio mappe mundi de Hughes de Saint‐Victor: retractio et additamenta’, in L'abbaye parisienne de Saint‐Victor au moyen age, ed. Jean Longère (Paris, Brepols, 1991), 143–79, reference at 93–94; Kupfer, ‘Medieval world maps’ (see note 20), 264.

Gautier Dalché, ‘La trasmissione medievale’ (see note 4), 44–46.

Hans Lieb, ‘Zur Herkunft der Tabula Peutingeriana’, in Die Abtei Reichenau: Neue Beiträge zur Geschichte und Kultur des Inselklosters, ed. Helmut Maurer (Sigmaringen, Thorbecke, 1974), 31–33. Gautier Dalché, ‘La trasmissione medievale’ (see note 4), 46–47, considers it ‘not improbable’—though ‘not a proven fact’—that Lieb's map is the Peutinger prototype. The map's compatibility with ‘an ideology of renovatio imperii’ makes it quite feasible, in his view, that a version of the map existed in the Carolingian age.

The notation ‘a world map on two rolls’ could simply mean that the adhesive between two sheets failed, and the map was preserved in two pieces—a possibility for a long Peutinger archetype.

Pliny the Elder, Naturalis Historia, 4.79; Tacitus, Germania, 1.2.

Lieb knew this to be a fourth‐century name from its presence in the history of Ammianus Marcellinus (per Marcianas siluas; 21.8.2). Lieb thought it extremely unlikely, however, that Hermann got the term from Ammianus. First of all, Reichenau had no copy of that text, which indeed found few readers throughout the middle ages. Ammianus, furthermore, used the plural form, while Hermann copied the singular, as on the Peutinger Map. Hermann's text has circa silvam Martianam at the year 1030 (Hermann Contractus, Chronicon, in Patrologia Latina 143, ed. Jacques‐Paul Migne (Paris, Editions Garnier frères, 1882), 235). The Peutinger Map has SILUA MARCIANA, ‐ci being a common orthographic variant of ‐ti in Latin from the second century ad onward. Lieb also ruled out the likelihood that the rare term had survived in the oral tradition around Reichenau from the fourth century to the eleventh, to surface only in Hermann's history. He did assume, following conventional wisdom on the Peutinger map, that Hermann's source, the ninth‐century Reichenau map and presumed proto‐Peutinger, was a fourth‐century production. The fourth‐century term, silua Marciana, fits nicely with that supposition. Yet its presumed presence on the Reichenau mappa mundi no more marks that map as fourth century than the presence of Pompeii on the Peutinger makes that a first‐century map.

Brodersen, ‘The presentation of geographical knowledge’ (see note 18), 20–21. For a forceful restatement of the way that ‘itineraries dominated and infiltrated all the other categories of ancient representations and perceptions of space’, see Whittaker (‘Mental maps’ (note 3), 83 and 93), who succinctly explains why Romans saw their world this way: ‘Itineraries, like the roads they documented, were the instruments of military conquest and provincial administration’.

For the Italian example, see Whittaker, ‘Mental maps’ (note 3), 81. For Gaul, see A. C. Bertrand, ‘Stumbling through Gaul: maps, intelligence, and Caesar's Bellum Gallicum’, The Ancient History Bulletin 11:4 (1997): 107–22.

Charlemagne's palace library held a particularly rich collection of Roman texts. On the Carolingian rescue of classical authors, see Reynolds and Wilson, Scribes and Scholars (note 37), 97–101.

Ingrid Baumgärtner, ‘Die Welt im kartographischen Blick: Zur Veränderbarkeit mittelalterlicher Weltkarten am Beispiel der Beatustradition vom 10. bis 13. Jahrhundert’, in Der weite Blick des Historikers: Einsichten in Kultur‐, Landes‐ und Stadtgeschichte. Peter Johanek zum 65. Geburtstag, ed. Wilfried Ehbrecht, Angelika Lampen, Franz‐Joseph Post and Mechthild Siekmann (Cologne, Weimar and Vienna, Böhlau Verlag, 2002), 527–49. Dr Baumgärtner sent me several helpful articles, and I owe a special debt to her for her corrections and generous responses to my questions.

Woodward, ‘Medieval mappaemundi’ (see note 12), 303–4, gives a brief description of these so‐called Beatus maps, which occur in manuscripts of the Commentary on the Apocalypse of Saint John by the Benedictine abbot Beatus of Liébana (fl. 776–86). Woodward traces all these to ‘the now‐lost prototype of 776–86’. The definitive edition is John Williams, ed., The Illustrated Beatus: A Corpus of the Illustrations of the Commentary on the Apocalypse, 5 vols. (London, Harvey Miller Publishers, 1994–2003).

Evelyn Edson has reminded me of Beatus manuscripts with a blank where the map is supposed to be. Some of the extreme variants may result from a copyist's decision to create his own map or to copy another available one when his manuscript had no map to follow.

On the shape of the first Beatus map, and its probable source, see John Williams, ‘Isidore, Orosius and the Beatus map’, Imago Mundi 49 (1997): 7–32.

Popular culture, too, continues to elevate the skills of antiquity and repeat the old story of medieval decline, contrary to the evidence of maps. Consider, for instance, the beautifully illustrated coffee‐table book by Nathaniel Harris, Mapping the World: Maps and Their History (San Diego, Thunder Bay Press, 2002), 22. Under a nineteenth‐century drawing of barbarians scaling the walls of Rome as ancient statues fall and shatter all around, the caption reads: ‘Roman mapmaking skills disappeared with the civilization that had created them’.

The map calls attention to Alexander's imperial ambitions with an inscription in India: ‘Hic Alexander responsum accepit usq[ue] quo Alexander’ [Here Alexander got the oracle's answer: ‘How far, Alexander?’].

On this Roman connection between geography and imperial rule, see Whittaker, ‘Mental maps’ (note 3), especially 84. As Whittaker notes here, ‘a global setting was one reason why Strabo thought geographia was “relevant to the practice of provincial governors” by stressing “the whole oikoumene under one rule”’.

On the contemplative journey effected by the maps of Matthew Paris's Chronica majora, see Daniel K. Connolly, ‘Imagined pilgrimage in the itinerary maps of Matthew Paris’, Art Bulletin 81 (1999): 598–622.

Additional information

Notes on contributors

Emily Albu

Emily Albu is associate professor of classics at University of California, Davis.